Introduction: What is self-efficacy and how might it relate to cardiovascular health?
Self-efficacy refers to the confidence in one’s ability to behave in such a way as to produce a desirable outcome (Bandura, 1977). Self-efficacy makes a difference in how people feel, think, and act. In terms of feeling, a low sense of self-efficacy for a particular situation is positively related to depression and anxiety. High self-efficacy for a specific situation allows one to deal better with uncertainty, distress, and conflict. In terms of thinking, the strong sense of competence resulting from high self-efficacy facilitates enhanced cognitive processes and academic performance. Finally, in terms of action, self-related cognitions are a major ingredient of the motivation process. Self-efficacy levels can enhance or impede motivation. People with high self-efficacy in a particular domain of human functioning choose to perform more challenging tasks. They set higher goals and stick to them. Actions are preshaped in thought, and people anticipate either optimistic or pessimistic expected outcomes of a specific task in line with their level of self-efficacy. Once an action has been taken, high self-efficacious persons invest more effort and persist longer than those low in self-efficacy to accomplish a specific task. When setbacks occur, those with high self-efficacy recover more quickly and maintain commitment to their goals (Schwarzer, 1992). Self-efficacy levels for specific cardiovascular health-related behaviors may be an important determinant of future cardiovascular health. Dietary self-efficacy, physical activity self-efficacy, and cessation of smoking self-efficacy are among the examples that will be discussed where self-efficacy for specific health-related behaviors likely plays a large role in future cardiovascular risk factor profiles. The following sections will review the evidence supporting links between self-efficacy of specific cardiovascular health-related behaviors and specific well-established cardiovascular risk factors.
Assessing the multiple dimensions of self-efficacy
Bandura (1986) argued that self-efficacy expectations consist of three dimensions: magnitude, generality, and strength. Each of these dimensions implies different measurement procedures. Magnitude refers to the ordering of tasks by difficulty level. Generality concerns the extent to which efficacy expectations about a specific situation can be generalized to other situations. Finally, strength refers to a judgment of how certain one is of being able to succeed at a particular task (Mudde et al., 1995).
There is no standard measurement for self-efficacy. Self-efficacy, unlike dimensions of personality, must be considered in terms of a specific situation (Gerin et al., 1995). Therefore, different measures are used to assess self-efficacy for each particular health-related behavior studied. In addition, a different measure is often used to assess each of the three dimensions of self-efficacy for a particular health-related behavior.
Self-efficacy and Smoking Cessation
Cigarette smoking has been identified as a risk factor for cardiovascular disease. Furthermore, studies indicate that patients with coronary artery disease who stop smoking have a lower mortality from all causes of death and a less frequent occurrence of myocardial infarction in particular than those who continue to smoke (Vlieststra et al., 1986). Self-efficacy in the area of smoking cession has been thoroughly studied. Thus, this is an excellent first example of how self-efficacy has been shown to predict behavioral change for an important cardiovascular risk factor. i.e., smoking. “Several studies have emphasized the predictive value of self-efficacy for behavior change among smoking cessation treatment participants (for review see: Strecher et al., 1986; Crey et al., 1989). The consensus of these studies is as follows: pre-treatment self-efficacy was generally not predictive of smoking status after treatment. However, post-treatment self-efficacy of subjects who were abstinent after treatment was significantly higher than self-efficacy of those who were not successful. Post-treatment self-efficacy expectations were significant predictors for short-term maintenance of smoking cession (3-6 months after treatment) (Pederson et al., 1991; Haaga, 1989; Coelho, 1984; McIntyre et al., 1983). This relationship held true even when post-treatment smoking status was controlled for and only subjects, who were abstinent after the treatment were included (Haaga and Stewart, 1992; Bear and Lichtenstein, 1988; Diclemente, 1981). For self-quitters the predictive power of self-efficacy may be even stronger. Gritz et al., 1992 found that self-efficacy predicted long-term abstinence (12-18 months) from smoking as a result of self-initiated quit attempts in agroup of female smokers (Mudde et al., 1995).
A recent prospective study by Mudde et al., 1995 provides further insight by comparing different measures of self-efficacy for smoking cessation in regards to which is the strongest predictor for smoking cessation. Each measure represented different combinations of the three dimensions of self-efficacy. The perceived difficulty scale used (PDS- Strecher et al.,1985) represented magnitude and generality. The perceived ability scale (PAS- Coletti et al., 1985) and a 1-item perceived ability measure (PAM- Mudde et al., 1995) both incorporated generality and strength in different degrees. Their results confirmed that perceived self-efficacy for smoking cessation predicts short and long-term smoking cessation. Results also suggested that the PDS may predict short-term cessation, while the PAS may be a predictor of long-term abstinence. By means of factor analyses, various subscales were found in both of these scales. The negative/affective subscale of the PDS and the negative/moodstates subscale of the PAS appeared to be the most important elements of these two scales. However, it was the PAM measure that showed the greatest predictive power for post-follow-up abstinence from smoking. Since all three measures included the dimension generality, a conclusion might be that magnitude is the dimension of greatest importance for the prediction of short-term cessation success, while strength may be the dimension that best determines long-term abstinence. Please see Mudde et al., 1995 for further details or for a discussion of the external validity of the study.
“In summary, the predictive ability and consistency of self-efficacy evaluations for smoking behavior have been impressive. Few constructs in the social sciences can boast such a record. In almost every case, efficacy evaluations, particularly abstinence efficacy evaluations, have been the most significant, or among the only significant, predictors of smoking cessation treatment outcome that emerged from studies that included a wide range of other predictors (DiClemente, 1986; DiClemente et al., 1995).
Self-efficacy and Weight Loss
Being overweight is a significant risk factor for the development of hypertension and coronary heart disease (see the link for a review of obesity and cardiovascular health by clicking on “overweight” among the listed risk factors our web page: www.workhealth.org). Weight control self-efficacy to perform behaviors that lead to weight loss has been examined in a number of ways. In fact, the assessment of self-efficacy varies greatly from study to study and is more diverse than in smoking cessation self-efficacy research. Due to the great variety in assessment of the construct of weight control self-efficacy, it should be noted that it is difficult to summarize the findings and make generalizable conclusions. The closest thing to a standard assessment in the field of weight control self-efficacy is Glynn and Ruderman’s (1986) Eating Self-Efficacy scale (ESES) (DiClemente et al., 1995).
There is a large amount of evidence suggesting that weight control self-efficacy plays an important role in weight loss. Chambliss and Murray (1979) have found cross-sectional evidence that a self-efficacy enhancing treatment group had greater weight loss than a comparison group. However, this effect was only apparent for those with an internal locus of control. Much stronger evidence comes from numerous prospective studies investigating the predictive powers of weight control self-efficacy on weight control. “Efficacy to resist the urge to overeat increases during the course of treatment (Glynn & Ruderman, 1986; Forster & Jeffrey, 1986). Expectations that seem more like outcome expectancies than efficacy expectancies (i.e., subjects’ confidence in reaching their goal weight, confidence in losing a certain amount of weight, or confidence in their ability to lose weight and maintain that loss) have been able to predict dropout from a weight control program (Mitchell & Stuart, 1984), as well as weight loss (Weinberger et al., 1984), and the maintenance of that weight loss (Blair et al., 1989). Most studies that use efficacy to resist the urge to eat or refrain from overeating have found these efficacy evaluations to be predictive of weight loss during the active phase of treatment (Glynn & Ruderman, 1986; Forster & Jeffrey, 1986). In addition, posttreatment efficacy evaluations have been related positively to maintenance of weight loss (Patsis & Hart, 1991; Rodin et al., 1988; DiClemente et al., 1995).
“Despite all the difficulties and differences in the assessment of self-efficacy related to eating behavior in weight control, the role that self-efficacy appears to play is quite similar to that in smoking cessation, where the assessments have been a bit more uniform. Efficacy evaluations appear to be useful and unique predictors of weight loss. Few constructs predict weight loss and maintenance of that loss in as consistent a fashion as self-efficacy focused on overeating behaviors (DiClemente et al., 1995).
Self-Efficacy and Low-Fat Diet
“There is substantial evidence that reducing saturated fat in the diet decreases the risk of coronary heart disease (CHD) in populations (Kromhout and Lezenne-Coulander, 1984; Kushi et al., 1985; Keys et al., 1986). Efforts to alter dietary habits through various programs have met with limited effectiveness (Advisory Board – IHHC, 1992). One reason for the limited success is the failure to fully understand the cognitive mediators of dietary change (Plotnikoff and Higginbotham, 1995) .One recent study by Plotnikoff and Higginbotham (1995) found significant positive association between self-efficacy for following a low-fat diet (dietary self-efficacy) and outcome measures related to low-fat diet. Future research is needed focusing specifically on dietary self-efficacy to determine whether substantial saturated fat reductions can be obtained by treatment programs developed to increase dietary self-efficacy.
Self-efficacy and Recovery from Heart Attack
Ewart and colleagues have studied the relationship between physical activity self-efficacy and recovery from heart attack. According to Ewart, 1992, “large numbers of heart attack survivors experience unnecessary distress and put themselves at significant medical risk due to excessive fear of physical activity.” Self-efficacy theory has improved their ability to identify and alleviate these inappropriate fears. Research reviewed by Ewart(1992) suggests that self-efficacy appraisals influence patient involvement in exercise regimens and mediate beneficial effects of exercise participation.” Furthermore, evidence is reviewed that self-efficacy predicts physical over-exertion and has called for the development of scales of self-efficacy to identify individuals who may be at risk of dangerous overexertion due to unrealistically optimistic appraisals of their physical capabilities (Ewart, 1992).
Self-efficacy and Physical Activity
Of course, physical activity is not only important to recovery from heart attack. Physical activity appears to decrease the risk of coronary artery disease. The United States Center for Disease Control reviewed existing observational studies and found a significant and graded relationship between physical activity and the risk of coronary artery disease (Powell et al., 1987; Littman, 1993). Exercise is widely recommended for health promotion and primary-risk reduction in people who have not developed symptoms of cardiovascular illness, and who are not inhibited by the anxieties patients experience after a heart attack (Ewart, 1995). However, relatively few people engage in regular exercise for a period of time to secure the benefits of moderate exercise to physical health (Dubbert, 1992). “The attrition rate for both clinical and community-based exercise programs can be as high as 50% within the first 3 to 6 months of participation (Brawley & Rogers, 1993; Dishman, 1988). Social-cognitive variables, including self-efficacy, seem to play a major role in this attrition.” Research suggests that self-efficacy expectancy, outcome expectancy, and outcome value are important in the initiation and maintenance of a variety of exercise programs (Brawley & Rogers, 1993; Brawley & Horne, 1988; Desharnais et al., 1986; Dzewaltowski et al., 1990; Garcia & King, 1991; McAuley, 1991, 1994; McAuley & Courneya, 1993; McAuley & Jacobson, 1991; Poag-Ducharme & Brawley, 1991a, 1991b; Rogers & Brawley, 1991a, 1991b). The relative influence of self-efficacy and outcome expectancy on exercise behavior has been shown to differ at different stages of exercise experience (Marcus et al., 1992; McAuley, 1991; McAuley & Jacobson, 1991; McAuley & Rowney, 1990; Poag-DuCharme & Brawley, 1993). Although initial experience with exercise may base their decision to try it largely upon their beliefs of the value of the benefits of exercise, this initial experience with exercise strongly influences self-efficacy (Ewart et al., 1983) which become the primary determinant of persistence (Maddux et al., 1995). Recently, studies have attempted to learn more about what types of self-efficacy best predict exercise behavior over time. Self-efficacy for the exercise components, self-efficacy for scheduling, and self-efficacy for overcoming barriers have been studied by Poag-DuCharme and Brawley (1993). The types of self-efficacy that predicted exercise intentions varied at different points in their 12-week community-based exercise program. This study and others (Poag-Charme, 1993; McAuley, 1992, 1993) suggest the need for further study concerning changes in the relationship between self-efficacy and exercise over time (Maddux et al., 1995).
Self-efficacy as a Component of Active Coping to Stress, and Resulting Enhanced Cardiovascular Reactivity
Stress, specifically model of job stress called “job strain,” has been shown in over a dozen epidemiological studies over the last decade to be a significant risk factor for CHD (Schnall et al., 1994; See our review of job strain). The concept of active coping is one model in which we think about the stress-illness relationship (Gerin et al., 1996). Active coping to stress enhances cardiovascular reactivity, elevating blood pressure and heart rate (Gerin et al., 1996). However, it is unclear which aspect of the process actually produces the elevations. A recent study by Gerin et al., 1996 has “concluded that self-efficacy for a particular task may be an integral part of the active coping process, indirectly affecting the blood pressure response by acting on the effort involved in the coping response.” Subjects engaged in a video game shape-matching task, who were preevaluated to have a high self-efficacy for this task, had greater blood pressure cardiovascular reactivity than subjects engaged in the same task who were preevaluated to have a low efficacy for this task.
Social determinants of Self-efficacy
There has been remarkably little research on the determinants of self-efficacy, particularly social class and job stress. Such factors might shape personality development in childhood. For example, certain parental behavior patterns (i.e., overly strict, critical and demanding of conformity) are more common in low SES households, and may be viewed as a reflection of the parents’ occupational and other life experiences, which are characterized by low control and insecurity (see Sennett &Cobb, 1973, The Hidden Injuries of Class; Rubin, 1976, Worlds of Pain: Life in the Working Class Family). Similarly, an adult’s experience, which might include stressful, low control jobs, may shape their personality development (Kohn and Schooler, 1982). The active motivated persistent personality style described in the beginning of this review as resulting from “self-efficacy” is quite similar to the hypothesized effects of active (high demand-high latitude) jobs on personality and coping in Karasek’s model (Karasek, 1979; 1981; Landsbergis et al., 1992). In the Cornell worksite blood pressure study, a three year increase in job decision latitude among men was associated with quitting smoking (Landsbergis et al., 1998). It is not known whether increased self-efficacy was a mediator in this causal pathway. Thus, research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
Smoking cessation, a CVD risk factor, has been the most common cardiovascular health-related behavior that self-efficacy has been linked to. “The predictive ability and consistency of self-efficacy evaluations for smoking behavior have been impressive. Few constructs in the social sciences can boast such a record. In almost every case, efficacy evaluations, particularly abstinence efficacy evaluations, have been the most significant, or among the only significant, predictors of smoking cessation treatment outcome that emerged from studies that included a wide range of other predictors.” (DiClemente, 1986; DiClemente et al., 1995).
Numerous studies have also investigated the relationship self-efficacy has on weight control, since obesity is a well-established CVD risk factor. Despite all the difficulties and differences in the assessment of self-efficacy related to eating behavior in weight control, the role that self-efficacy appears to play is quite similar to that in smoking cessation, where the assessments have been a bit more uniform. Efficacy evaluations appear to be useful and unique predictors of weight loss. Few constructs predict weight loss and maintenance of that loss in as consistent a fashion as self-efficacy focused on overeating behaviors.” (DiClemente et al., 1995).
Self-efficacy has been shown to play a role in influencing other CVD risk factors, such as high-fat diet, physical inactivity, and high blood pressure (via active coping to stress). In regards to high-fat diet, future research is needed focusing specifically on dietary self-efficacy to determine whether substantial saturated fat reductions can be obtained by treatment programs developed to increase dietary self-efficacy.
Concerning physical inactivity, research suggests that self-efficacy expectancy, outcome expectancy, and outcome value are important in the initiation and maintenance of a variety of exercise programs. Self-efficacy for physical activity may become particularly important for those recovering from a heart attack. However, the links between self-efficacy and these CVD health-related behaviors have not been as thoroughly studied as for smoking cessation and weight control. Finally, concerning high blood pressure, it appears that self-efficacy for confronting mentally and physically vigorous psychosocial environmental stressors may be a component of active coping. Active coping to stress enhances cardiovascular reactivity, elevating blood pressure and heart rate (Gerin et al., 1996). High blood pressure is of course a major CVD risk factor. As for high-fat diet and physical activity, the link between self-efficacy and high blood pressure through active coping to stress has not been studied thoroughly enough. More research is needed investigating the role that self-efficacy plays in active coping, work-stress models (i.e., job strain, effort-reward and John Henryism), and as a possible risk factor for hypertension.
Further research is definitely needed to better understand how much self-efficacy influences each of these CVD health-related behaviors, and the mechanism of that change by further narrowing in on what tasks self-efficacy is particularly important for in affecting behavioral change. Lastly, very little is known about the determinants of self-efficacy. Research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
Advisory Board of the International Heart Health Conference (1992, May 28). Bridging the Gap: Science and Policy in Action – Declaration, Victoria, Canada.
Bandura, A.(1976). Social foundations of thought and action. Englewood Cliff, NJ: Prentice-Hall.
Bandura, A.(1977). Self-efficacy: Toward a unifying theory of behavior change. Psychol. Rev., 84, 191-215.
Bear, J.S., & Lichenstein, E.(1988). Classification and prediction of smoking relapse episodes: An explanation of individual differences. Journal of Consulting and Clinical Psychology, 56, 104-110.
Blair, A., Booth, D., Lewis, V., & Wainwright, C.(1989). The relative success of official and informal weight reduction techniques: Retrospective correlational evidence. Psychology and Health, 3(3), 195-206.
Brawley, L.R., & Horne, T.E.(1988, December). Refining attitude-behavior models to predict adherence in normal and socially supportive conditions: Part I & II. Report: Project No. 8706-4042-2099, Canadian Fitness and Lifestyle Institute, Ottawa.
Brawley, L.R., & Rogers, W.M.(1993). Social psychological aspects of fitness promotion. In P. Seraganian (Ed.), Exercise psychology: The influence of physical exercise on psychological processes (pp. 254-298). New York, NY: Wiley.
Carey, M., Snel, D., Carey, K., & Richards, C.(1989). Self-initiated smoking cessation: A review of the empirical literature from a stress and coping perspective. Cognitive Therapy and Research, 13, 323-341.Chambliss, C.A., & Murray, E.J.(1979). Efficacy attribution , locus of control of weight loss. Cognitive Therapy and Research, 3, 349-353.
Coelho, R.J.(1984). Self-efficacy and cessation of smoking. Psychological Reports, 54, 309-310.
Coletti, G., Supnick, J., & Payne, T.(1985). The smoking self-efficacy questionnaire(SSEQ): Preliminary scale development and validation. Behavioral Assessment, 7, 249-260.
Desharnais, R., Bouillon, J., & Godin, G.(1986). Self-efficacy and outcome expectations as determinants of exercise adherence. Psychological Reports, 59, 1157-1159.
DiClemente, C.C.(1981). Self-efficacy and smoking cessation maintenance: A preliminary report. Cognitive Therapy and Research, 5, 175-187.
DiClemente, C.C.(1986). Self-efficacy and the addictive behaviors. [Special Issue: Self-efficacy theory in contemporary psychology]. Journal of Social and Clinical Psychology, 4(3), 302-315.
DiClemente, C.C., Fairhurst, S., & Piotrowski, N.(1995). In J. Maddux (Ed.), Self-efficacy, adaptation, and adjustment: theory, research, and application (pp. 109-141). New York, NY: Plenum Press.
Dishman, R.K.(1988). Exercise adherence: Its impact on public health. Champaign, IL: Human Kinetics.
Duppert, P.M.(1992). Exercise in behavioral medicine. Journal of Consulting and Clinical Psychology, 60, 613-618.
Dzewaltowski, D., Noble, J., & Shaw, J.(1990). Physical activity participation: Social cognitive theory versus the theories of reasoned action and planned behavior. Journal of Sport and Exercise Psychology, 12, 388-405.
Ewart, C.K., Taylor, B., Reese, L., & Debusk, R.(1983). Effects of early postmyocardial infarction exercise on self-perception and subsequent physical activity. The American Journal of Cardiology, 51, 1076-1080.
Ewart, C.K.(1992). Role of physical self-efficacy in recovery from heart attack. In R. Schwarzer (Ed.), Self-efficacy: thought control of action (pp. 287-304). Washington D.C.: Hemisphere Publishing Corporation.
Ewart, C.K.(1995). Self-efficacy and recovery from heart attack: Implications for a social cognitive analysis of exercise and emotion. In R. Schwarzer (Ed.), Self-efficacy, adaptation, and adjustment: theory, research, and application(pp.203-226). New York, NY: Plenum Press.
Forster, J.L., & Jeffery, R.W.(1986). Gender differences related to weight history, eating patterns, efficacy expectations, self-esteem, and weight loss among participants in a weight reduction program. Addictive Behaviors, 11(2), 141-147.
Garcia, A.W., & King, A.C.(1991). Predicting long-term adherence to aerobic exercise: A comparison of two models. Journal of Sport and Exercise Psychology, 13, 394-410.
Gerin, W., Litt, M., Deich, J., & Pickering, T.(1995). Self-efficacy as a moderator of perceived control effects on cardiovascular reactivity: Is enhanced control always beneficial? Psychosomatic Medicine, 57, 390-397.
Gerin, W., Litt, M., Deich, J., & Pickering, T.(1996). Self-efficacy as a component of active coping: Effects on cardiovascular reactivity. Journal of Psychosomatic Research, 5, 485-493.
Glynn, S.M., & Ruderman, A.J.(1986). The development and validation of an eating self-efficacy scale. Cognitive Therapy and Research, 10(4), 403-420.
Gritz, E., Berman, B., Bastani, R., & Wu, M.(1992). A randomized trail of a self-help smoking cessation intervention in a nonvolunteer female population: Testing the limits of the public health model. Health Psychology, 11, 280-289.
Haaga, D.A.(1989). Articulated thoughts and endorsement procedures for cognitive assessment in the prediction of smoking relapse. Psychological Assessment, 1, 112-117.
Haaga, D.A., & Stewart, B.L.(1992). Self-efficacy for recovery from a lapse after smoking cessation. Journal of Consulting and Clinical Psychology, 60, 24-28.
James, S.A.(1986). John Henryism: The JH Active Aoping Scale (JHAC12) (unpublished scoring instructions).
James, S.A., Harnett, S., & Kalsbeek, W.(1983). John Henryism and blood pressure differences among black men. J. Behav. Med., 6, 259-278.
James, S.A., Strogatz, D., Wing, S., & Ramsey, D.(1987). Socioeconomic status, John Henryism, and hypertension in blacks and whites. Am. J. Epidemiol., 126, 664-673.
Keys, A., Menotti, A., Karvonen, M., Aravanis, C., Blackburn, H., Buzina, R., Djordjevic, B., Dontas, A., Fidanza, F., Keys, M., et al. (1986). The diet and 15-year death rate in the seven countries study. American Journal of Epidemiology, 124, 903-915.
Kohn, M.L., & Schooler, C.(1982). Job conditions and personality: A longitudinal assessment of their reciprocal effects. American J. Sociology, 87(6), 1257-1286.
Kromhout, D., & de Lezenne-Colander, C.(1984). Diet, prevalence and 10 year mortality from coronary heart disease in 871 middle-aged men. The Zutphen Study. American Journal of Epidemiology, 119, 733-741.
Kushi, L., Lew, R., Stare, F., Ellison, C., el Lozy, M., Bourke, G., Daly, L., Graham, I., Hickey, N., Mulcahy, R., et al.(1985). Diet and 20-year mortality from coronary heart disease. The Ireland-Boston Diet-Heart Study. New Zealand Journal of Medicine, 312, 811-818.
Littman, A.B.(1993). Review of Psychosomatic Aspects of Cardiovascular Disease. Psychother. Psychosom., 60, 148-167.
Maddux, J., Brawley, L., & Boykin, A.(1995). Self-efficacy and healthy behavior: Prevention, promotion, and detection. In R. Schwarzer(Ed.), Self-efficacy, adaptation, and adjustment: theory, research, and application(pp.173-202). New York, NY: Plenum Press.
Marcus, B., Selby, V., Niaura, R., & Rossi, J.(1992). Self-efficacy and the stages of exercise behavior change. Research Quarterly for Exercise and Sport, 63, 60-66.
McAuley, E.(1991). Efficacy, attributional, and affective responses to exercise participation. Journal of Sport and Exercise Psychology, 13, 382-393.
McAuley, E.(1992). The role of exercise cognitions in the prediction of exercise behavior of middle-aged adults. Journal of Behaviorla Medicine, 15, 65-88.
McAuley, E.(1993). Self-efficacy and the maintenance of exercise participation in older adults. Journal of Behavioral Medicine, 16, 103-113.
McAuley, E.(1994). Physical activity and psychosocial outcomes. In C. Bouchard, R.J. Shephard, & T. Stephens (Eds.), Physical activity, fitness, and health: International proceedingsand consensus statement(pp. 551-568). Champaign, IL: Human Kinetics.
McAuley, E., & Courneya, K.S.(1993). Adherence to exercise and physical activity as health-promoting behaviors: Attitudinal and self-efficacy influences. Applied and Preventative Psychology, 2, 65-77.
McAuley, E., & Jacobson, L.(1991). Self-efficacy and exercise particpation in sedentary adult females. American Journal of Health Promotion, 5, 185-191.
McAuley, E., & Rowney, T.(1990). Exercise behavior and intentions: The mediating role of self-efficacy cognitions. In L. VanderVelden & J.H. Humphrey (Eds.), Psychology and sociology of sport(Vol. 2, pp. 3-15). New York, NY: AMS Press.
McIntyre, K., Lichtenstein, E., & Mermelstein, R.(1983). Self-efficacy and relapse in smoking cessation: A replication and extension. Journal of Consulting and Clinical Psychology, 51, 632-633.
Mitchell, C., & Stuart, R.B.(1984). Effect of self-efficacy on dropout from obesity treatment. Journal of Consulting and Clinical Psychology, 52, 1100-1101.
Mudde, A., Kok, G., & Strecher, V.(1995). Self-efficacy as a predictor for the cessation of smoking: Methodological issues and implications for smoking cessation programs. Psychology and Health, 10, 353-367.
Patsis, P., & Hart, K.E.(1991). Coping and self-efficacy in weight-loss maintenance. Paper presented at the convention of the Society of Behavioral Medicine, Washington, D.C.
Pederson, L., Strickland, C., DesLauriers, A.(1991). Self-efficacy related to smoking cessation in general practice patients. International Journal of the Addictions, 26, 467-485.
Plotnikoff, R.C., & Higginbotham, N.(1995). Predicting Low-fat diet intentions and behaviors for the prevention of coronary heart disease: an application of Protection Motivation Theory among an Australian population. Psychology and Health, 10, 397-408.
Poag-DuCharme, K.A.(1993). Goal-related perceptions of social-cognitive predictors of exercise behavior. Unpublished dissertation. University of Waterloo, Waterloo, Ontario.
Poag-DuCharme, K.A., & Brawley, L.R.(1991a, October). The goal dynamics of fitness classes: A preliminary analysis. Paper presented at the annual meeting of the Canadian Society for Psychomotor Learning and Sport Psychology, London, Ontario.
Poag-DuCharme, K.A., & Brawley, L.R.(1991b, October). The relationship of self-efficacy and social support to exercise intentions in the aged. Paper presented at the annual meeting of the Canadian Society for Psychomotor Learning and Sport Psychology, London, Ontario.
Poag-DuCharme, K.A., & Brawley, L.R.(1993). Self-efficacy theory: Use in the prediction of exercise behavior in the community setting. Journal of Applied Sport Psychology, 5, 178-194.
Powell, K., Thompson, P., Casperson, C., & Kendrick, J.(1987). Physical activity and the incidence of coronary artery disease. Annu. Rev. Publ. Health, 8, 253-287.
Rodin, J., Elias, M., Silberstein, L., & Wagner, A.(1988). Combined behavioral and pharmacologic treatment for obesity: Predictors of successful weight maintenance. Journal of Consulting and Clinical Psychology, 56(3), 399-404.
Rogers, W.M., & Brawley, L.R.(1991a). The role of outcome expectancies in participation motivation. Journal of Sport and Exercise Psychology, 13, 411-427.
Rogers, W.M., & Brawley, L.R.(1991b, June). Evaluating fitness messages promoting involvement: Effects of attitudes and behavioral intentions. Paper presented at the annual meeting of North American Society for the Psychology of Sport and Physical Activity, Kent, Ohio.
Schnall, P., Landsbergis, P., & Baker, D.(1994). Job strain and cardiovascular disease. Annu. Rev. Publich Health, 15, 381-411.
Schwarzer, R.(1992). Preface. In R. Schwarzer(Ed.), Self-efficacy: thought control of action(pp. ix-xiv). Washington D.C.: Hemisphere Publishing Corporation.
Strecher, V., Becker, M., Kirscht, J., Eraker, S., & Graham-Tomasi, R.(1985). Evaluation of a minimal-contact smoking cessation program in a health care setting. Patient Education and Counseling, 7, 395-407.
Strecher, V., DeVellis, B., Becker, M., & Rosenstock, I.(1986). The role of self-efficacy in achieving health behavior change. Health Education Quarterly, 13, 73-91.
Vlieststra, R., Kronmal, R., Oberman, A., Frye, R., & Killip, T.(1986). Effect of cigarette smoking on survival of patients with angiographically documented coronary artery disease. JAMA, 255, 1023-1027.
Weinberger, R., Hughes, H., Critelli, J., England, R., & Jackson, A.(1984). Effects of preexisting and manipulated self-efficacy on weight loss in a self-control program. Journal of Research in Personality, 18, 352-258.
Introduction- What is Locus of Control?
Originally developed within the framework of Rotter’s (1954) social learning theory, the locus of control construct refers to the degree to which an individual believes the occurrence of reinforcements is contingent on his or her own behavior. The factors involved with reinforcement expectancy are labeled “external” and “internal” control. In short, internal locus of control refers to the perception of positive or negative events as being a consequence of one’s own actions and thereby under one’s own personal control. In contrast, external locus of control refers to the perception of positive or negative events as being unrelated to one’s own behavior in certain situations and thereby beyond personal control. As a general principle, the locus of control variable may be thought of as affecting behavior as a function of expectancy and reinforcement within a specific situation (Carlise-Frank, 1991).
The following sections will review the evidence supporting links between locus of control and specific cardiovascular health-related behaviors. In addition, methods of assessment of locus of control will be discussed, as well as the current limitations of studies investigating the relationship between locus of control and cardiovascular health-related behavior.
How is Locus of Control Assessed in Health-Related Research?
Rotter’s Locus of Control Scale (Rotter, 1966), a generalized measure of internal vs. external locus of control, continues to be widely used to assess perceived control in health-related research (see reviews by Strickland, 1978; Wallston and Wallston, 1978). However, many health researchers have chosen to use more situation-specific, health-related locus of control measures in their investigations. The most widely used instrument of this sort is the Multidimensional Health Locus of Control (MHLC) Scale (Wallston et al., 1978). The MHLC Scale consists of two alternative forms (A and B) each of which contains 18 items. Each form, in turn, contains three six-item Likert scales which, in “normal healthy” populations are uncorrelated, or only slightly correlated, with one another (Wallston and Wallston, 1981). The Internal Health Locus of Control (IHLC) dimension assesses the degree to which one believes one’s health status is influenced by one’s own behavior. People who score high on the IHLC are said to have a sense of responsibility for their own health (Wallston and Wallston, 1982). PHLC measures the belief that powerful other people (one’s family, friends or health-care providers) control one’s health. Lastly, CHLC assesses perceived non-control of health, or the belief that fate, luck, or chance determines one’s health status. (Wallston, 1989)
Another approach to measuring locus of control are “Sphere-specific measures of perceived control” developed by Delroy Paulhus (1983). Paulhus’ questionnaire has a total of 30 items, broken down into 3 subscales called “personal efficacy”, “interpersonal control”, and “socio-political control”. The internal consistency reliabilities for the scales are typically .75-.80, higher than those obtained for the Potter scale on the same samples. Pauljus argues convincingly that the construct of locus of control is multidimensional, and that therefore a scale (such as his) that measures perceived control in different domains of life is a better measure than Rotter’s single scale.
Locus of Control and Cardiovascular Health-Related Behaviors
The literature concerning research on the relationship between locus of control and health-facilitating behavior as a whole points toward internal locus of control as a mediating factor of actions taken to prevent health problems (See Lefcourt and Davidson-Katz, 1991; Carlise-Frank, 1991). Physical activity is one of these health-facilitating behaviors whose relationship to locus of control has been studied. Physical activity appears to decrease the risk of coronary artery disease. The United States Center for Disease Control reviewed existing observational studies and found a significant and graded relationship between physical activity and the risk of coronary artery disease
(Powell et al., 1987; Littman, 1993). Exercise is widely recommended for health promotion and primary-risk reduction in people who have not developed symptoms of cardiovascular illness (Ewart, 1995). However, relatively few people engage in regular exercise for a period of time to secure the benefits of moderate exercise to physical health (Duppert, 1992). The attrition rate for both clinical and community-based exercise programs can be as high as 50% within the first 3 to 6 months of participation (Brawley & Rogers, 1993; Dishman, 1988). Sonstroem and Walker (1973) studied locus of control and attitudes toward physical fitness and found that internals had more favorable attitudes towards physical activity, obtained significantly better fitness scores, and engaged in greater amounts of voluntary physical exercise than did externals (Carlise-Frank, 1991). It must be recognized, however, that these authors used a theoretical and methodological approach that supported a generalized expectancy for control beliefs. Validity of studies investigating the relationship between locus of control and health-related behaviors have been questioned due to the fact that an individual may have a tendency towards internality in many life areas, but have an external belief with regard to the particular health-related behavior in question. Please see the end of this section for further discussion of this methodological problem.
Like the research on health-facilitating behaviors, research on individuals who attempt to overcome health-damaging behaviors has also shown internals are often better off than externals (Coan, 1973; James et al., 1965; Mlott and Mlott, 1975; Naditch, 1975; Pryer and Distefano, 1977; Williams, 1967). Unlike research on health-facilitating behaviors, however, results of research on overcoming health-damaging behaviors has been far less consistent in favoring internality. In this area, results have shown that externals and internals are equally successful in overcoming their deleterious health behaviors when placed in treatment programs consistent with their personal control beliefs (Carlise-Frank, 1991).
Being overweight is a significant risk factor for the development of hypertension and coronary heart disease. In studies concerning weight reduction as a function of locus of control, Balch and Ross (1975) found internal beliefs to be predictive of success and completion of an overweight treatment program. (Carlise-Frank, 1991) (Wallston et al.1976), using the Health Locus of Control scale, found internals in weight reduction programs were more satisfied with the results of their treatment than externals, but these results failed to reach significance. “Some researchers have been unable to relate the internal/external (I-E) variable to weight reduction (Bellack et al., 1974; Manno and Marston, 1972; Tobias and MacDonalad, 1977).” (Carlise-Frank, 1991).
Cigarette smoking has also been identified as a risk factor for cardiovascular disease (Schnall et al., 1994). Furthermore, studies indicate that patients with coronary artery disease who stop smoking have a lower mortality from all causes of death and a less frequent occurrence of myocardial infarction in particular than those who continue to smoke (Vlieststra et al., 1986). Results from research on stopping smoking have shown that internals are far more likely than externals to be affected by the Surgeon General’s report and are more likely to stop smoking (James et al., 1965). In another study, internals were found to profit more from a stop smoking program using a saturation technique than externals (Best and Steffy, 1975). Though some researchers have reported that internals are more likely to reduce their smoking rate than externals (Best and Steffy, 1971; Steffy et al., 1970), others have failed to demonstrate a relationship between locus of control and control of smoking behavior. In these studies, I-E scores have failed to predict treatment success, suggesting that locus of control data have little predictive value in smoking cessation (Berstein, 1970; Best and Steffy, 1971; Danaher, 1977; Keutzer, 1968; Carlise-Frank, 1991).
Perhaps one of the most important reasons for the discrepancy of studies investigating the relationship between locus of control and cardiovascular health-related behaviors is the approach used in determining whether individuals should be placed in internal or external treatment groups (Carlise-Frank, 1991). “The most commonly used measure of subjects’ control beliefs has been a generalized unidimensional scale of overall expectancy. If an individual’s belief system pointed towards internality, then he or she would be placed in a treatment regime consistent with that belief. However, an individual may have a tendency towards internality in many life areas but have an external belief with regard to health or the particular health-related behavior in question. It is unlikely that a generalized unidimensional scale would be able to detect such distinctions. If such individuals were placed in treatment groups congruent with internal beliefs, success might be marginal at best. From this perspective, internally oriented individuals who have learned externality or passive acceptance towards their health-damaging behavior would not be expected to have a high success rate in such programs.”(Carlise-Frank, 1991). This would explain the failure by many researchers to demonstrate a relationship between locus of control as a predictor for overcoming the cardiovascular health-damaging behaviors previously discussed (Carlise-Frank, 1991).
Diagnosis of Myocardial Infarction, Recovery from Myocardial Infarction, and Locus of Control
Another area of investigation concerning locus of control and cardiovascular health involves how locus of control influences health-related behavior once myocardial infarction has already begun. Early detection of acute myocardial infarction(AMI) reduces myocardial infarction morbidity (Genton and Sobel, 1987), yet up to one third of AMIs are not recognized (Roseman, 1954; Lindberg et al., 1960; Roseman et al., 1967; Kannel and Abbott, 1984; Kannel et al., 1990). Unrecognized AMI includes completely asymptomatic events and those with a typical symptoms, so that neither the patient nor physician entertains the diagnosis of AMI (Bertolet and Hill, 1989). In a recent study by Theisen et al.(1995), patients with unrecognized AMI scored higher on the CHLC (“chance” locus of control) than patients with diagnosed AMI. Unrecognized AMI is known to result from a lack of symptom experience and/or avoidance of medical care when symptoms are experienced. In analysis of their results, Theisen et al.(1995) hypothesize that “chance” locus of control, the belief in chance or fate as determining health, may inhibit treatment-seeking for AMI. More research is needed to rigorously test this hypothesis (Theisen et al., 1995).
The role of locus of control in recovery from AMI has also been explored in a study by Cromwell et al. (1977). “Patients who were classified as internal from Rotter’s Internal-External Locus of Control Scale (Rotter, 1966) were rated by the professional staff as being more cooperative and less depressed than were externals throughout their stay in the intensive care unit. On three highly intercorrelated physiological measures (sedimentation rates, serum glutamic oxaloacetic transaminase levels, and lactate dehydrogenase levels), externals were found to have worse prognoses than internals. Additionally, externals had higher peak temperatures during intensive care and remained longer in the unit, and in the hospital, than did internals. One tempting hypothesis is that internals simply behave in a manner that does not aggravate their fragile conditions. Rather than becoming distressed, with all of the personal and physiological consequences of distress, internals showed greater cooperation and less depression, possibly reflecting their more active participation and greater hope in the struggle for survival. It is plausible that responses to life-endangering threats such as myocardial infarction may be at least partially determined by personality characteristics such as locus of control.”(Lefcourt and Davidson-Katz, 1991) However when a measure of control beliefs made up of items relating specifically to perceived control over recovery from heart disease was used, no support was found that perceived control is important for psychological adjustment among cardiac patients (Flowers, 1994). Further research is needed before an important evaluation of the hypothesis that an internal locus of control leads to increased likelihood of behavior favoring recovery from AMI.
Locus of Control and Job Strain
Karasek (1979) has defined “job strain” as work in jobs with high psychological demands and low control. In more than a dozen epidemiological studies over the last decade, occupational stress researchers have implicated job strain as a risk factor for heart disease (Schnall et al., 1994). Although the mechanism by which the stress of job strain influences development of CHD is unknown, previous findings suggest that job strain may be related to elevations of blood pressure at work (Schnall et al., 1990, Van Egeren, 1992). High blood pressure has long been known to be a marker for individuals at high risk for developing CHD, and therefore is one possible mechanism by which job strain might exert its deterious influences on the heart.
One basic question yet to be answered by the job strain model “relates to the issue of objective versus subjective control. Clearly, the job strain model considers control as an objective characteristic of the work situation. However, cognitive and affective responses of the workers to these characteristics vary considerably according to their individual patterns of appraisal and coping (Lazarus and Folkman, 1984). Generalized control beliefs have been found to moderate the effects of objective job conditions on well-being (Spector, 1987). Furthermore, through regression analysis, Hendrix (1989) found locus of control to be a statistically significant predictor of job stress (Beta=.39; p<.001). These findings (as well as findings where individual coping characteristics such as Type A behavior, hostility, or lack of hardiness were associated with increased ischemic heart disease) call for a conceptual clarification of the relationship between control-limiting job conditions and those personal characteristics (particularly locus of control) which influence the perception of control (Siegrist et al., 1990). One possible relationship is interaction between job conditions and personality characteristics. For example, Parkes (1991) found a significant three-way interaction between job demands, job decision latitude, and Paulhus’ locus of control scale, in predicting affective distress and anxiety. For externals, demands and latitude combined intractively to predict outcome (consistent with Karasek’s model), whereas for internals, additive findings (main effects for demands and latitude) were obtained.
Social determinants of Locus of Control
There has been remarkably little research on the determinants of locus of control, particularly social class and job control. Such factors might shape personality development in childhood. For example, certain parental behavior patterns (i.e., overly strict, critical and demanding of conformity) are more common in low SES households, and may be viewed as a reflection of the parents’ occupational and other life experiences, which are characterized by low control and insecurity (Sennett and Cobb, 1973; Rubin, 1976). Similarly, an adult’s experience, which might include stressful, low control jobs, may shape their personality development (Kohn and Schooler, 1982). For example, Lefcourt (1982, p. 31) pointed out that locus of control is “positively associated with access to opportunity”
In a study of 2174 Dutch men and women aged 25-74 (Bosma et al., 1998), subjects with higher chidlhood socioeconomic status (SES) had much higher levels of “perceived control” (locus of control) in adulthood. In addition, “perceived control” appeared to be an important mediatorof the association of SES with later mortality. The association between SES and mortality (RR=2.6) was reduced substantially (RR=1.8_ agter controlling for levels of “perceived control”. Thus, research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
Validity of studies investigating the relationship between locus of control and health-related behaviors have been questioned due to the fact that an individual may have a tendency to be internal in many life areas, but have an external belief with regard to the particular health-related behavior in question. Valid locus of control measures more specific to the particular behavior being studied need to be developed. However, results using Rotter’s Locus of Control Scale and the more health behavior-related MHLC, have in general found that internals tend to behave in a more healthy way in regardseveral important cardiovascular health-related behaviors. Specifically,one study has shown internals to have better attitudes toward physical fitness, better physical fitness scores, and engaged in greater amounts of voluntary physical exercise (Sonstroem and Walker, 1973). Furthermore, there have also been findings suggesting that an internal locus of control may be beneficial in weight loss among the overweight, and successful smoking cessation. It should be noted however that there have been contradictory findings regarding the link between locus of control and these cardiovascular health-damaging behaviors. Another area of investigation concerning locus of control and cardiovascular health involves how locus of control influences health-related behavior once myocardial infarction has already begun. Patients with unrecognized AMI in one study were shown to score higher on the CHLC than patients with diagnosed AMI (Thiesen et al., 1995). This had led to the hypothesis that “chance” locus of control may inhibit treatment-seeking for AMI. Secondly, work by Cromwell et al. (1977) led to their hypothesis that an internal locus of control may lead to increased likelihood of behavior favoring recovery from AMI. There has been little further work investigating either of the aforementioned hypotheses. More research is definitely needed in these areas.
Locus of control may also be important in studies of job strain, implicated as a risk factor for heart disease by over a dozen epidemiological studies over the last decade. The current job strain model considers control as an objective characteristic of the work situation. Locus of control and other measures of generalized control beliefs have been shown to moderate the effects of objective job conditions on well being and as predictors of job stress. These findings call for future studies investigating the relationship between control-limiting job conditions and personal characteristics, such as locus of control, which influence the perception of control. Such work may lead to a better understanding of how the perception of control relates to job stress-related cardiovascular disease via the cardiovascular risk factor of job strain.
Lastly, there has been very little research on the determinants of locus of control. Research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
Balch, P. and Ross, A.W.(1975). Predicting success in weight reduction as a function of locus of control: A unidimensional and multidimensional approach. Journal of Consulting and Clinical Psychology, 43, 119.
Bellack, A., Rosensky, R., and Schwartz, J.(1974). Comparison of two forms of self-monitoring in a behavioral weight reduction program. Behavioral Therapy, 5, 523-530.
Berstein, D.A.(1970). The modification of smoking behavior: A search for effectively variables. Behavior Research and Therapy, 8(2), 133-136.
Bertolet, B., and Hill, J.(1989). Unrecognized myocardial infarction. In C.Pepine (Ed.), Acute Myocardial Infarction (pp. 173-182). Philadelphia, MA: FA Davis.
Best, J.A., and Steffy, R.A.(1971). Smoking modification procedures tailored to subject characteristics. Behavior Therapy, 2, 177-191.
Best, J.A., and Steffy, R.A.(1975). Smoking modication procedures for internal and external locus of control clients. Canadian Journal of Behavioral Science, 7(2), 155-165.
Bosma H, Van de Mheen, Mackenbach JP. Childhood socioeconomic conditions and adult health: The contribution of psychological attributes. Presented at the International Congress of Behavioral Medicine, Copenhagen, Denmark, August 21, 1998.
Brawley, L.R., & Rogers, W.M.(1993). Social psychological aspects of fitness promotion. In P. Seraganian(Ed.), Exercise psychology: The influence of physical exercise on psychological processes (pp. 254-298). New York, NY: Wiley.
Carlisle-Frank, P.(1991). Examining personal control beliefs as a mediating variable in the health-damaging behavior of substance use: an alternative approach. The Journal of Psychology, 125(4), 381-397.
Coan, R.(1973). Personality variables associated with cigarette smoking. Journal of Personality and Social Psychology, 26, 86-104.
Cromwell, R., Butterfield, E., Brayfield, F., and Curry, J.(1977). Acute Myocardial Infarction. St. Louis, MO: C.V. Mosby.
Danaher, B.(1977). Rapid smoking and self-control in the modification of smoking behavior. Journal of Consulting and Clinical Psychology, 45, 1068-1075.
Dishman, R.K.(1988). Exercise adherence: Its impact on public health. Champaign, IL: Human Kinetics.
Duppert, P.M.(1992). Exercise in behavioral medicine. Journal of Consulting and Clinical Psychology, 60, 613-618.
Ewart, C.K.(1995). Self-efficacy and recovery from heart attack: implications for a social cognitive analysis of exercise and emotion. In R. Schwarzer (Ed.), Self-efficacy, adaptation, and adjustment: theory, research, and application(pp.203-226). New York, NY: Plenum Press.
Flowers, B.J.(1994). Perceived control, illness status, stress, and adjustment to cardiac illness. The Journal of Psychology, 128(5), 567-576.
Genton, R., and Sobel, B.(1987). Early intervention for interruption of acute myocadial infarction. Mod. Concepts Cardiovasc. Dis., 56, 35-41.
Hendrix, W.H.(1989). Job and personal factors related to job stress and risk of developing coronary artery disease. Psychological Reports, 65, 1136-1138.
James, W., Woodruff, A., and Werner, W.(1965). Effect of internal and external control upon changes in smoking behavior. Journal of Consulting Psychology, 29(2), 184-186.
Kannel, W., and Abbott, R.(1984). Incidence and prognosis of unrecognized myocardial infarction. N. Engl. J. Med., 311, 1144-1147.
Kannel, W., Cupples, L., and Gagnon, D.(1990). Incidence, precursors, and prognosis of unrecognized myocardial infarction. In J.J. Kellermann and E. Braunwald (Eds.), Silent Myocardial Ischemia: A Critical Appraisal. Adv. Cardiol., 37, 202-214.
Karasek, R.A.(1979). Job demands, job decision latitude, and mental strain: Implications for job redesign. Adm. Sci. Q., 24, 285-308.
Keutzer, C.(1968). A measure of cognitive dissonance as a predictor of smoking treatment outcome. Psychological Reports, 22, 655-658.
Kohn, M.L., & Schooler, C.(1982). Job conditions and personality: A longitudinal assessment of their reciprocal effects. American J. Sociology, 87(6), 1257-1286.
Lazarus, R.S., and Folkman, S.(1984). Stress, Appraisal and Coping. New York: Springer.
Lefcourt HM. Locus of control. Hillsdale, NJ: Lawrence Erlbaum, 1982.
Lefcourt, H.M., and Davidson-Katz, K.(1991). Locus of control and health. In C.R. Snyder and D.R. Forsyth(Eds.), Handbook of Social and Clinical Psychology: the health perspective (pp. 246-266). New York: Pergamon Press.
Lindberg, H., Berkson, D., Stamler, J., and Poindexter, A.(1960). Totally asymptomatic myocardial infarction: an estimate of its incidence in the living population. Arch. Intern. Med., 5, 628-633.
Littman, A.B.(1993). Review of Psychosomatic Aspects of Cardiovascular Disease. Psychother. Psychosom., 60, 148-167.
Manno, B., and Marston, A.(1972). Weight reduction as a function of negative covert reinforcement versus positive covert reinforcement. Behavior Research and Therapy, 10, 201-207.
Mlott, R., and Mlott, Y.(1975). Dogmatism and locus of control in individuals who smoke, stopped smoking, and never smoked. Journal of Community Psychology, 3, 53-57.
Naditch, M.(1975). Locus of control and drinking behavior in a sample of men in army basic training. Journal of Consulting and Clinical Psychology, 43, 96.
Parkes KR. Locus of control as moderator: An explanation for additive versus interactive findings in the demand-discretion model of work stress? British Journal of Psychology 1991;82:291-312.
Paulhus D. Sphere-specific measures of perceived control. Journal of Personality and Social Psychology 1983;44:1253-1265.
Powell, K., Thompson, P., Casperson, C., & Kendrick, J.(1987). Physical activity and the incidence of coronary artery disease. Annu. Rev. Publ. Health, 8, 253-287.
Pryer, M., and Distefano, M.(1977). Correlates of locus of control among male alcoholics. Journal of Clinical Psychology, 33(1), 300-303.
Roseman, M.(1954). Painless myocardial infarction: review of literature and analysis of 220 cases. Ann. Intern. Med., 41, 1-8.
Rosenman, R., Friedman, M., Jenkins, C., Straus, R., Wurm, M., and Kositchek, R.(1967). Clinically unrecognized myocardial infarction in the Western Collaborative Group Study. Am. J. Cardiol., 19, 776-782.
Rotter, J.(1954). Social learning and clinical psychology. Englewood Cliffs, NJ: Prentice-Hall.
Rotter, J.B.(1966). Generalized expectancies for internal versus external control of reinforcement. Psychological Monographs, 33(1), 300-303.
Rubin LB. Worlds of pain. New York: Basic Books, 1976.
Schnall, P., Landsbergis, P., & Baker, D.(1994). Job strain and cardiovascular disease. Annu. Rev. Publich Health, 15, 381-411.
Schnall, P.L., Pieper, C., Schwartz, J.E., Karasek, R.A., Schlussel, Y, et al.(1990). The relationship between “job strain,” workplace diastolic blood pressure, and left ventricular mass index: Results of a case-control study. J. Am. Med. Assoc, 267, 1209.
Sennett R, Cobb J. The hidden injuries of class. New York: Vintage Books, 1973.
Siegrist, J., Peter, R., Junge, A., Cremer, P., and Seidel, D.(1990). Low status control, high effort at work and ischemic heart disease: prospective evidence from blue-collar men. Soc. Sci. Med., 31(10), 1127-1134.
Sonstroem, R., and Walker, M.(1973). Relationship of attitudes and locus of control to exercise and physical fitness. Perceptual and Motor Skills, 36, 1031-1034.
Spector, P.E.(1987). Interaction effects of perceived control and job stressors on affective reactions and health outcomes for clerical workers. Work & Stress, 1, 155.
Strickland, B.R.(1978). Internal-external expectancies and health-related behaviors. Journal of Consulting and Clinical Psychology, 46, 1192-1211.
Thiesen, M., MacNeill, S., Lumley, M., Ketterer, M., Goldberg, D., and Borzak, S.(1995). Psychosocial factors related to unrecognized acute myocardial infarction. The American Journal of Cardiology, 75, 1211-1213.
Tobias, L., and MacDonald, M.(1977). Internal locus of control and weight loss: An insufficient condition. Journal of Consulting and Clinical Psychology, 45, 647-653.
Van Egeren, L.F.(1992). The relationship between job strain and blood pressure at work, at home, and during sleep. Psychosomatic Medicine, 54, 337-343.
Vlieststra, R., Kronmal, R., Oberman, A., Frye, R., & Killip, T.(1986). Effect of cigarette smoking on survival of patients with angiographically documented coronary artery disease. JAMA, 255, 1023-1027.
Wallston, K.A.(1989). Assessment of control in health-care settings. In A. Steptoe and A. Appels (Eds.), Stress, Personal Control and Health (pp.85-105). New York: John Wiley & Sons.
Wallston, B.S., and Wallston, K.A.(1978). Locus of control and health: a review of the literature. Health Education Monographs, 6, 107-117.
Wallston, B.S., and Wallston, K.A.(1981). Health locus of control. In H. Lefcourt (Ed.), Research with the Locus of Control Construct(Vol. 1). New York: Academic Press.
Wallston, B.S., and Wallston, K.A.(1982). Who is responsible for your health? The construct of health locus of control. In G. Sanders and J. Suls (Eds.), Social Psychology of Health and Illness(pp. 65-95). Hillsdale, NJ: Erlbaum.
Wallston, B.S., Wallston, K.A., Kaplan, G., and Maides, S.(1976). Development and validation of the Health Locus of Control(HLC) Scale. Journal of Consulting and Clinical Psychology, 44, 580-585.
Wallston, K.A. Wallston, B.S., and DeVellis, R.(1978). Development of the multidimensional health locus of control (MHLC) scale. Health Education Monographs, 6, 5-25.
Williams, A.F.(1967). Self-concepts of college problem drinkers. Quarterly Journal of Studies in Alcoholism, 28, 267-276.
Defensiveness has been defined as a coping style characterized by an orientation away from threatening information and a denial or minimization of distress and negative emotions (Jamner et al., 1991). The standard measure of defensiveness is the Marlow-Crowne scale, or the MCSD (Crowne and Marlowe, 1960). Persons scoring high on this scale appear to underreport, deny, or suppress negative emotions such as anxiety and anger (Shapiro et al., 1995).
Previous research has found high scores on the MCSD to represent a defensive or self-deceptive response style which involves avoidance of anxiety-arousing cognitions (Weinberger, 1990; Walsh, 1990; Nicholson and Hogan, 1990; Paulhus, 1984; McCrae and Costa, 1983). High scores on the MCSD have also been shown to be reflective of the individual’s attention to the possibility that the revelation of anxiety-arousing cognitions might threaten social approval (Paulhus, 1984).
Hypertension has long been known to be an important risk factor for cardiovascular diseases (CVDs). Although hypertension is a health problem that affects all ethnic groups, hypertension has been shown to be particularly prevalent in African-Americans. Blacks in the U.S. are 2-4 times more likely than whites in the U.S. to develop hypertension by age 50 (Roberts and Rowland,1981). The reasons for the excess risk in African-Americans are not known. Numerous genetic and environmental factors have been hypothesized to contribute to the excess risk, but their relative contributions are still a matter of debate (Saunders, 1991). However, one thing is clear and universally accepted: socioeconomic status (whether measures are by education, occupation, or income) and hypertension tend to be inversely associated, for both Blacks and Whites (Tyroler, 1986). This has led to the suggestion that unrelieved psychosocial stress, generated by environments in which African-Americans live and work, is primarily responsible for their heightened susceptibility to hypertension.
In the early/mid 1970’s, numerous studies demonstrated that “high effort” coping (i.e., sustained cognitive and emotional engagement) with difficult psychological stressors produce substantial increases in heart rate and systolic blood pressure. Increases were shown to persist only as long as individuals actively worked at trying to eliminate the stressor. The effects were seen in a variety of different environments. Some of these studies were controlled laboratory experiments (Obrist et al., 1978), while others were field-based studies of “real life” stressors (Kasl and Cobb, 1970; Cobb and Rose, 1973; Harburg et al., 1973). This body of research led to a commentary by Syme (1979). Syme observed that persons of lower socioeconomic status (especially Blacks in these positions) by definition face more difficult psychosocial environmental stressors than more economically privileged individuals. He proposed that prolonged, high effort coping with difficult psychosocial stressors could be the explanation of both the inverse association between socioeconomic status and hypertension typically observed in U.S. communities and the increased risk for this disorder in Black Americans. This was the beginning of what later became known as the “John Henryism Hypothesis.”
What is the John Henryism Hypothesis and how is it assessed?
The term “John Henryism” was coined by Sherman James et al., 1983 as a synonym for prolonged, high-effort coping with difficult psychological stressors. The “John Henryism hypothesis” is the belief that John Henryism (JH) among lower socioeconomic groups that may not have the resources to successively cope with difficult psychological stressors are primarily responsible for the increased prevalence of hypertension among lower socioeconomic groups. James et al., 1983 also provided a 12 item scale called “The John Henryism Scale for Active Coping,” or the JHAC12, for measurement of John Henryism. Here are a few examples of the items on the JHAC12:
“I don’t let my personal feelings get in the way of doing a job.””Once I make up my mind to do something, I stay with it until the job is completely done.”
“Sometimes I feel that if anything is going to be done right, I have to do it myself.”
Subjects taking the JHAC12 respond to these items by selecting from the following responses:
1)completely false 2)somewhat false 3)somewhat true 4)completely true 5)don’t know
The JHAC12 is still currently the standard measurement of JH.
Research investigating the John Henryism Hypothesis
James et al., 1983, was also the first formal study of the John Henryism hypothesis. 132 randomly-selected, working-class Black men, ages 17-60, from a rural North Carolina community were given the original version of JHAC12 and had their blood pressure measured. This area of the country (Edgecombe county) has among the highest death rates in the country due to stroke and heart disease. Socioeconomic status(SES) was measured by years of formal education. High school graduates were classified as high SES, while high school dropouts were classified as low SES. Consistent with most other published studies, non-high school graduates in this study had higher adjusted diastolic blood pressures than high school graduates(81.1 mm Hg versus 77.1 mm Hg). When divided into high and low John Henryism groups, the difference in mean blood pressure for high school graduates versus non-graduates in the low John Henryism group was very small (1.7 mmHg). Furthermore, as predicted by the John Henryism hypothesis, the difference in mean blood pressure for high school graduates versus non-graduates in the high John Henryism group was considerably larger (6.3 mmHg).
The 1983 by James was largely a pilot study to test the validity of the JHAC12. The positive findings led to a larger study in 1987, again by Sherman James, that consisted of a larger randomly-selected sample that included both Blacks and Whites, ages 21-50, from the same North Carolina rural community. The sample of Whites largely consisted of skilled, blue-collar and lower mid-level white collar workers, while the sample of Blacks consisted mostly of unskilled and semi-skilled workers. Although null findings were found in Whites, the results again showed strong support of the John Henryism hypothesis in Blacks. Among Blacks, the low SES group had a higher mean diastolic blood pressure than the high SES group. More importantly, among Blacks, the difference in mean blood pressure between the high SES group and the low SES group was greater in the high John Henryism group (3.8 mmHg) versus the low John Henryism group (1 mmHg).
Even more striking was the difference in hypertension prevalence between the high SES group and the low SES group when again first divided into high JH and low JH groups. In the low JH group, differences in SES were not associated with drastic differences in hypertension prevalence (25% vs. 23.4%). However, in the high JH group, hypertension prevalence was almost three times greater for Blacks in the low SES group (31.4%) versus those in the high SES group (11.4%). According to James et al.,1987, the 11.5% hypertension prevalence in the high JH/high SES group is unusually low for any group of adult Blacks, and suggests that high JH/high SES might be protective against hypertension for Black adults.
The next major study investigating the John Henryism hypothesis sought to replicate James’ previous findings in a totally different population group. Both studies by James et al. previously discussed were confined to the community of Edgecombe County in North Carolina, making generalizations regarding the John Henryism hypothesis outside of the rural South in the United States difficult. Duijkers et al., 1988, studied the relationship between JH, SES, and blood pressure in the Dutch town of Zutphen. Of the 100 men and 100 women randomly selected participants in the study, all were between the ages 20-59 and 96% were Caucasian. As usual, John Henryism was measured using the JHAC12, shown to have high levels of internal consistency by James et al. in their U.S.-based studies. Years of education was used as an approximation of SES. Unfortunately, results for the most part were not statistically significant. After adjustment for age, alcohol consumption, physical activity, and Quetelet Index (a dependent of SES), a statistically significant positive correlation (F(1,92) = 8.04, p<.01) remained only between John Henryism and systolic blood pressure in men. When years of education was also taken into account, the only statistically significant difference (p<.05) in systolic blood pressure observed was in the group with fewer years of education. In this group, those with low JH had a mean systolic blood pressure of 124.6, while those with high JH had a mean systolic blood pressure of 134.9 (after adjustments for the other hypertension risk factors). Comparatively, in the group with greater years of education, the difference between low JH and high JH mean systolic blood pressure was 6.1 mmHg. Similar differences in the sample of 100 women were also found, but due to sample size problems when dividing into subgroups, any generalizations of the John Henryism hypothesis to women became impossible. The results of this study are support that the potential contribution of John Henryism to explaining variance in blood pressure in men is not limited to blacks in the rural South of the United States.
Other studies have attempted to replicate James’ findings in study populations that differ not only geographically, but also in age range and/or educational level. Jackson and Campbell, 1994 examined the relationship between John Henryism and blood pressure in 162 male and 259 female black college students from the University of Pittsburgh (Pittsburgh, PA), University of Massachusetts (Amherst, MA), and Paine and Augusta Colleges (Augusta, GA). In this study, no association was found between John Henryism and blood pressure measures. This failure to reproduce the finding of earlier papers has been explained by Jackson and Campbell, 1994 to be likely due to the diversity between this study population and previous ones. John Henryism has almost exclusively been examined in rural, low-SES male populations (Jackson and Campbell, 1994). Their results suggest that John Henryism may not be a significant factor for elevated blood pressure in groups with access to certain economic and social resources, such as those that are college educated. In other words, John Henryism may be a moderator variable between hypertension and other unknown variables that were not present in the study population of Jackson and Campbell, 1994. Another possibility is that the effect of stress beyond the college experience is important in the John Henryism-blood pressure relationship. Longitudinal studies of a similar population of Blacks would be need to test this hypothesis. Whatever the reason for the null findings, it has little relevance to the John Henryism hypothesis, because the study population was not subdivided into groups that differed in their SES. The John Henryism hypothesis predicts that only in low SES groups will John Henryism be positively correlated with blood pressure. It is extremely unlikely that all the study participants were from low SES backgrounds, and so therefore this study does not undermine the John Henryism hypothesis. This study further suggests that John Henryism is a moderator variable and that other additional variables must be taken into account to understand the relationship between John Henryism and blood pressure levels. There is substantial evidence that SES is one of these additional variables, but others may need to be unmasked before relationships become clear.
Although Wilst and Jackson did not find an association between John Henryism and blood pressure in their Black college student study population, there is evidence suggesting that the John Henryism hypothesis is valid in youth, as well as among adults. Wright et al., 1996 have found that high JH scores were associated with higher blood pressure, higher total peripheral resistance (TPR), and lower cardiac output (CO) in their study of 173 normotensive 10- to 17-year-old Black and White children. Consistent with expectations from the John Henryism hypothesis, those children from lower SES backgrounds who were high on John Henryism had particularly high levels of resting cardiovascular reactivity.
However, there have been studies whose null findings have shed doubt on the John Henryism hypothesis. Wilst and Flack, 1992, found no association between an interaction of John Henryism and SES and the risk of elevated blood pressure or definite hypertension. Identical methods to that of James were used in classifying SES and measuring JH. Wilst and Flack, 1992 identified research design and sample characteristics that may have been responsible for their null findings regarding the John Henryism hypothesis. They have proposed that psychological strategies to cope with environmental stressors among southwest urban African-Americans, Wilst and Flack’s study population, may differ from those of the rural southeast African-Americans studied by James and colleagues. Additionally, James studied a “relatively poor” community with a low level of education and a high unemployment rate (James et al., 1983), while Wilst and Flack’s study population was better educated and had only one-half the unemployment rate. Wilst and Flack identified many other differences in their study population in comparison to James’ that may also have contributed to the lack of association, but conclude that more studies of the John Henryism hypothesis in geographically diverse areas are needed to resolve the conflict of their results with that of James’ findings.
James and colleagues most recent study also supports the idea that the John Henryism hypothesis does not apply to all African-American population groups. In James’ third study (James et al., 1992), 1,784 black adults aged 25-50 years were randomly selected from inhabitants of Pitt County, North Carolina. This study population differed from James’ previous studies in Edgecombe County in that Pitt County has experienced more rapid urbanization and economic diversification than Edgecombe County. As a result, many more professional, middle-class Blacks were included in this study, which allowed for the creation of low SES, medium SES, and high SES classification groups. Defying James’ predictions, only a very modest and nonstatistically significant inverse association between SES and hypertension prevalence was observed. Furthermore, division of the sample into high and low John Henryism produced data that showed no support for the John Henryism hypothesis. However, upon reanalysis, an additional factor was discovered that James believed was responsible for the lack of an inverse association between SES and hypertension prevalence. Self-reported psychological stress was significantly (p<.05) positively correlated with mean blood pressures for both men and women in the Pitt County study (James et al., 1992). This self-reported psychological stress was noticed to be quite high among managerial level, white collar workers, and presumably raised the prevalence of hypertension to a surprisingly high level for the high SES group as a whole. To test this hypothesis, James and colleagues conducted a pot hoc analysis of the John Henryism hypothesis. They excluded all high SES persons whose psychological stress scores were above the sample median. Additionally, they excluded all low SES persons whose stress scores were below the sample median. With these exclusions, a strong inverse association between SES and psychological stress was observed for the remaining 1,131 participants of the study. These exclusions also had a significant effect on the inverse association between SES and hypertension prevalence: 24.7%, 23.4%, and 17.4% for the low, medium, and high SES groups, respectively. Most importantly, when subdivided into high and low JH groups, it was found hypertension prevalence varied little by SES in the low JH group while a strong, inverse association between SES and hypertension prevalence existed among the high JH group. This has led to the argument that the John Henryism hypothesis might only be observed in the following case:
It is only when chronic psychological stress is higher among lower SES groups than among groups of higher SES (the usual case) that the inverse association between SES and blood pressure will be strong, allowing data supporting the John Henryism hypothesis to be observed. However, the John Henryism hypothesis has not been sufficiently tested under these specific conditions, and so support for the John Henryism hypothesis remains fairly weak. Another important point that also drastically weakens support of the John Henryism hypothesis is that all studies have been cross-sectional. To provide more convincing evidence for the validity of the John Henryism hypothesis, prospective studies must be done that show the combination of low SES and high JH scores at one point in time contributes to an accelerated increase in blood pressure by some well defined, second point in time.
John Henryism and Job Strain
If indeed the inverse relationship between SES and blood pressure strongly depends on chronic psychological stress being more prevalent in lower SES groups, this would further implicate psychological stress as a risk factor for hypertension. Interestingly, high “job strain,” a major source of chronic psychological stress, has already been shown to be more strongly associated with hypertension and cardiovascular disease among men with lower SES than with men of higher SES (Johnson and Hall, 1988; Johnson et al., 1989; Karasek, 1981; Theorell et al., 1988). Furthermore, the association between high job strain and high blood pressure is about twice as strong among men with only 14 yrs or less of education vs. those with greater education (Landsbergis et al., 1994). “Job strain” has been defined by Karasek (1979) as work in jobs with high psychological demands (work pace + conflicting demands) and low decision latitude (control + variety and skill use). Five out of nine studies which have studied the relationship between job strain and ambulatory blood pressure found significant positive correlations, while the remaining four yielded a mixture of insignificantly positive and null results (Schnall et al.,1994). Not included in the review by Schnall et al., 1994 was Landsbergis et al.’s 1994 study. Landsbergis et al., 1994 found employees experiencing job strain had a systolic BP that was 6.7 mm Hg higher and a diastolic blood pressure that was 2.7 mm Hg higher at work than other employees, and that the odds of hypertension were also increased (odds ratio = 2.9, 95% CI). According to Schnall et al., 1994, the results taken as a whole, suggest that job strain acts, in part, to cause cardiovascular disease through the mechanism of elevated blood pressure. The link between job strain and hypertension appears to be even stronger than the link between John Henryism and hypertension. Yet although occupational
stress studies have implicated both John Henryism and job strain as likely risk factors for the development of hypertension, there is a major weakness in the literature in regards to whether these two work-stress models are independent or mutually reinforcing risk factors for hypertension and subsequent cardiovascular health problems. Furthermore, it would be interesting to determine whether part of the inconsistencies observed in the literature concerning the relationship between John Henryism and blood pressure could be resolved by taking job strain into account (in a similar way that SES and psychological stress were taken into account in recent John Henryism/hypertension studies).
Social determinants of John Henryism
There has been remarkably little research on the determinants of John Henryism, particularly social class and job stress. Such factors might shape personality development in childhood. For example, certain parental behavior patterns (i.e., overly strict, critical and demanding of conformity) are more common in low SES households, and may be viewed as a reflection of the parents’ occupational and other life experiences, which are characterized by low control and insecurity. Similarly, an adult’s experience, which might include stressful, low control jobs, may shape their personality development (Kohn and Schooler, 1982).” Thus, research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
High JH scores have been associated with elevated blood pressure most reliably among low SES, adult African-Americans. Although less convincingly, this association has also been observed in a wide variety of other sample populations. Positive findings for the association between high JH scores and elevated blood pressure have been found in sample populations that are diverse in sex, age, and ethnicity. The only common thread between the results from these diverse sample populations has been that the high JH/elevated blood pressure association has been consistently strongest in the lowest SES groups.
Based on James’ 1994 review, the feature most likely responsible for the pronounced effects among these low SES groups is higher chronic psychological stress. Despite some confliction among the results of studies investigating the relationship between John Henryism, SES, and blood pressure, the evidence as a whole supports that the John Henryism hypothesis is valid at least among specific population groups. However, all studies have been cross-sectional. To provide more convincing evidence for the validity of the John Henryism hypothesis, prospective studies must be done that show the combination of low SES and high JH scores at one point in time contributes to an accelerated increase in blood pressure by some well defined, second point in time. Additionally, future studies are needed to try to integrate the John Henryism hypothesis with other work-stress models, such as job strain, that have been linked more strongly with elevated blood pressure and adverse cardiovascular health consequences.
Lastly, very little is known about the determinants of John Henryism. Research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
Cobb, S., & Rose, R.(1973). Hypertension, Peptic Ulcer and Diabetes in Air Traffic Controllers. JAMA, 224, 489-492.
Duijkers, T., Drijver, M., Kromhout, D., & James, S.(1988). “John Henryism” and Blood Pressure in a Dutch Population. Psychosomatic Medicine, 50 353-359.
Harburg, E., Erfurt, J., Hauenstein, L., Chape, C., Schull, W., & Schork, M.(1973). Socioecological Stressor Areas and Black-White Blood Pressure. Journal of Chronic Disease, 26, 595-611.
Jackson, L., & Campbell, L.(1994). John Henryism and Blood Pressure in Black College Students. J Behavioral Medicine, 17, 69-79.
James, S.(1994). John Henryism and The Health of African-Americans. Culture, Medicine, and Psychiatry, 18, 163-182.
James, S., Hartnett, S., & Kalsbeek, W.(1983). John Henryism and blood pressure differences among black men. J Behav Med, 6, 259-278.
James, S., Keenan, N., Strogatz, D., Browning, S., & Garrett, J.(1992). Socioeconomic Status, John Henryism, and Blood pressure in Black Adults. Am J Epidemiology, 135, 59-67.
James, S., Strogatz, D., Wing, S., et al.(1987). Socioeconomic status, John Henryism, and hypertension in blakcs and whites. Am J Epidemiology, 126, 664-673.
Johnson, J., & Hall, E.(1988). Job strain, work place social support, and cardiovascular disease: A cross-sectional study of a random sample of the Swedish working population. Am. J. Public Health, 78, 1336-1342.
Johnson, J., Hall, E., & Theorell, T.(1989). Combined effects of job strain and social isolation on cardiovascular morbidity and mortality in a random sample of the Swedish male working population. Scand. J. Work Environ. Health, 15, 271-279.
Karasek, R.(1979). Job demands, job decision latitude, and mental strain: Implications for job redesign. Adm. Sci. Q., 24, 285-308.
Karasek, R., Baker, D., Marxer, F., Ahlbom, A., & Theorell, T.(1981). Job decision latitude, job demands, and cardiovascular disease: A prospective study of Swedish men. Am. J. Public Health, 71, 694-705.
Kasl, S., & Cobb, S.(1970). Blood Pressure Changes in Men Undergoing Job Loss: A Preliminary Report. Psychosomatic Medicine, 32, 19-38.
Kohn, M.L., & Schooler, C.(1982). Job conditions and personality: A longitudinal assessment of their reciprocal effects. American J. Sociology, 87(6), 1257-1286.
Landsbergis, P., Schnall, P., Warren, K., Pickering, T., & Schwartz, J.(1994). Association between ambulatory blood pressure and alternative formulations of job strain. Scand J Work Environ Health, 20, 349-363.
Obrist, P., Gaebelein, C., Teller, E., Langer, A., Grignoto, A., Light, K., & McCubbin, J.(1978). The Relationship Among Heart Rates, Carotoid dp/dt, and Blood Pressure in Humans as a Function of the Type of Stress. Psychophysiology, 15, 102-115.
Roberts, J., & Rowland, M.(1981). Hypertension in Adults 25-74 years of age, United States, 1971-1975. Hyattsville, MD: National Center for Health Statistics(Vital and Health Statistics, Series II: Data from the National Health Survey, no. 221) DHHS publication no 81-1671.
Saunders, E.(1991). Cardiovascular Diseases in Blacks. Philadelphia, PA: F.A. Davis Company.
Schnall, P., Landsbergis, P., & Baker, D.(1994). Job strain and Cardiovascular Disease. Annu. Rev. Public Health, 15, 381-411.
Syme, S.L.(1979). Psychosocial Determinants of Hypertension. In E. Oresti and C. Klint(eds.), Hypertension Determinants, Complications and Imtermention (pp.95-98). New York, NY: Grune and Stratton.
Theorell, T., Perski, A., Akerstedt, T., Sigala, F., Ahlberg-Hulten, G., et al.(1988). Changes in job strain in relation to changes in physiological states. Scand. J. Work Environ. Health, 14, 189-196.
Tyroler, H.A.(1986). Hypertension. In J.M. Last(eds.), Public Health and Preventive Medicine (pp. 1195-1214). Norwalk, CT: Appleton-Century-Crofts.
Wilst, W., & Flack, J.(1992). A Test of the John Henryism Hypothesis: Cholesterol and Blood Pressure. J Behavioral Medicine, 15, 15-29.
Wright, L., Treiber, F., Davis, H., & Strong, W.(1996). Relationship of John Henryism to Cardiovascular Functioning at Rest and During Stress in Youth. The Society of Behavioral Medicine, 18, 146-150.
Denial is psychological coping strategy that allows people to engage in behavior with little conscious awareness of the consequences (Allan and Scheidt, 1996). A review by Sirous (1992) traced 21 empirical studies on denial in coronary heart disease (CHD) published in the past 25 years and concluded that denial has a long-term negative effect on cardiovascular outcome (Allan and Scheidt, 1996). However, cardiac denial is a complex phenomenon. It appears that certain types of denial might be beneficial, while others might be very detrimental to cardiovascular health. This review of the scientific literature will discuss what is currently known about denial as a possible psychosocial risk factor for CHD. In the process, methods of assessment of denial and resulting methodological problems will be discussed, as well as evidence which is helping coronary-prone behavior researchers to better distinguish between dangerous and harmless, or even beneficial, uses of the basic defense mechanism of denial.
How is denial assessed?
A number of scales have been created in attempts to accurately assess and quantify denial and have been used in multiple clinical trials (Levine et al., 1987; Ramanaiah et al., 1977; Yanagida et al., 1981; Jamner and Schwartz, 1986). There is no standard measurement of denial, but the best-known scale assessing cardiac denial is that of Hackett and Cassem (Hackett and Cassem, 1974). According to Fields, 1989, their scale originated from a 1964 study by Olin and Hackett which “noted that myocardial infarction patients would go to extremes to attribute chest pain to causes other than the heart. From this study, Hackett and Cassem developed an interview technique consisting of a few structured questions, such as “What did you feel caused your chest pain?” The interviewers rated patients on a 31-item scale that indicates behavior typically seen in patients who deny major illness. This denial scale significantly distinguished three groups that they labeled as major, partial, and minimal deniers. Interrater reliability and comparisons with clinical observers have been consistently significant (Hackett and Cassem, 1974).” However, according to Dracup et al., 1995, there is still no evidence related to the reliability and validity of this rating scale.
Since no standard well-validated measure of denial exists, when comparing studies investigating the relationship between denial and cardiovascular health, it is a rarity that two studies using an identical method of assessment of denial will be observed. Many of these assessments of denial have not been sufficiently tested in reliability and validity. Therefore caution must be taken in placing too much weight on data from any group of studies until consistent results have been obtained using a variety of measures of denial, each with acceptable levels of validity and reliability.
The Early Stages of Myocardial Infarction and Denial
As previously mentioned, it appears that only certain types of denial, or denial at certain critical points in time, appear to be detrimental to cardiovascular health. Among the greatest of dangers to cardiovascular health that denial may present is the denial of a possible cardiac event. Denial is one of the first adaptive behaviors or mechanisms that an individual uses during the stress-producing event of an acute episode of chest pain (Hackett and Cassem, 1982). The key question to be answered is whether denial of a cardiac event in order to allay fear, anxiety, or other unpleasant emotions directly leads to a greater time delay in seeking medical help. The individual who delays seeking medical treatment risks increased myocardial damage, morbidity, and mortality (Allan and Scheidt, 1996). Since the mid-1980’s, several large-scale studies have demonstrated that thrombolytic therapy can significantly reduce mortality from acute myocardial infarction (AMI)(Kennedy et al., 1985; Marder and Sherry, 1988; Simoons et al., 1985; White et al., 1987; Yusuf et al., 1985). The benefits of thrombolytic therapy are directly related to the interval between the onset of symptoms and the administration of the drug. The shorter the interval, the better the outcome (Dracup et al., 1995). In previous studies it has been shown that a large part of the total time for a patient to come under coronary care is taken up by the patientUs decision to ask for medical help, the so-called patient delay (Goldstein et al., 1972; Simon et al., 1972; Colling et al., 1976; Rawles and Haites, 1988; Wielgosz et al.,1988; Leitch et al., 1989). Patient delay is defined as the time between the start of complaints from an AMI and the moment the patient or significant other calls for medical help (Bleeker et al., 1995).
The significance of denial in relation to patient delay has not been assessed thoroughly. Wielgosz and Nolan(1991) pointed out that denial has not been definitively linked with patient delay. Two studies(Wielgosz and Nolan, 1988; Hackett and Cassem, 1969) found no significant relationship between denial and patient delay. A more recent study (Bleeker et al., 1995), however, found that denying patients waited longer before looking for medical help. Bleeker et al., 1995 suggest that the use of different operationalizations and instruments may be the cause of these conflicting findings. Additionally, only one of four dimensions of the Denial questionnaire (Trijsburg et al., 1987; Trijsburg et al., 1989), resentment, was robustly related to patient delay. Conflicting findings will only be resolved through future research, and as a result, denial has yet to definitively linked with patient delay.
Given the lengthy patient delay that is usually reported, it remains tempting to identify denial as an important psychological response that significantly increases delay time (Dracup et al., 1995). According to Dracup et al., 1995, “two methodological problems make it difficult to understand the role of denial in patient delay time. The first is that all data related to the patient’s decision process must, of necessity, be collected retrospectively. Denial is a transitory coping mechanism that is difficult to evaluate after the fact. Investigators can ask patients how serious they thought their symptoms were or if they labeled their symptoms as cardiac in origin, but in the end it is impossible to assess the degree to which the patient used denial to ward off anxiety.” Secondly, the lack of an appropriate psychometric instrument to measure denial presents a large methodological problem. Previous studies which have not found a relationship between denial and patient delay (Wielgosz et al., 1988; Schmidt and Borsch, 1990; Hackett et al., 1968; Hackett and Cassem, 1969) may relate more to the inherent difficulty in measuring denial than to the lack of such a relationship (Dracup et al., 1995).
Hospitalization and Recovery from AMI and Denial
Once under coronary care, denial appears to play a different role in AMI patient outcome. Several studies have found AMI patients with high denial levels had superior outcomes during the first three days of recovery (Prince et al., 1982; Levine et al., 1987; Lazarus, 1983). During the first three days of hospitalization, arrhythmias, tachycardia, strong anxiety, and elevated blood pressure levels all increase mortality. In the coronary care unit setting, patients with strong denial experience fewer of these problems, perhaps because of less adrenergic nervous system stimulation (Fields, 1989). Levine et al.(1987) followed 30 men for 1 year post-MI and reported that individuals with high scores on a denial scale spent fewer days in intensive care and had fewer signs of cardiac dysfunction while in the hospital in comparison with those who had low scores.
However, once post-MI patients have been discharged from the hospital setting, this same study (Levine et al., 1987) suggests that denial switches from an adaptive behavior to a maladaptive behavior. After 1 year, high deniers were more non-compliant with treatment recommendations and required more days of re-hospitalization. Further support that denial of illness leads to problems once removed from the hospital setting comes from a recent prospective study by Julkunen and Saarinen, 1994. Using return to work 1 year following MI as a MI recovery outcome variable, they found that denial of illness was the best single psychological predictor of return to work. Denial of illness, assessed using a subscale of the Coping with Illness Scale (Julkunen, 1989; Saarinen, 1992) which was developed specifically for CHD patients, was more strongly correlated to return to work than work stress, anxiety, Type A, anger-out, or depression measures. Specifically, denial of illness had a -0.32 correlation with return to work (p<0.001). Numerous other studies also support that prolonged denial leads to ignorance of necessary activity restrictions, refusal to appreciate the seriousness of the illness, or failure to take prescribed medications needed to recover (Gentry and Haney, 1975; Douglas and Druss, 1987; Fields, 1989; McKenday and Logan, 1982; Scalzi and Burke, 1982; Viswanathan and Vizner, 1984).
Denial of Emotions, CHD, and CHD risk factors
There is also reason to believe that denial is not only a danger to individuals who are either in the early stages of AMI or in the process of out-of-hospital recovery from AMI. Prolonged denial of negative emotions and problems may substantially increase the risk of developing CHD. Denial of problems has been significantly positively correlated (p<0.01) with CHD somatic risk factors, such as LDL cholesterol, body mass index, and triceps skinfold, in 15 year old girls (Keltikangas and Jokinen, 1989). Similarly, denial of problems has been significantly positively correlated (p<0.01) with LDL cholesterol and total cholesterol in 12 year old girls (Keltikangas and Jokinen, 1989). It should be noted that significant correlation between denial and these CHD risk factors were not observed in 12, 15, and 18 year old boys, or 18 year old girls (Keltikangas and Jokinen, 1989).
Grossarth-Maticek and colleagues provide much stronger evidence that prolonged repression and denial of emotions may greatly increase the risk for CHD (Grossarth et al., 1985). According to a review of psychosocial risk factors for CHD by Kabat-Zinn, 1992: They “investigated the relationship between a number of psychosocial and personality factors and the incidence of and mortality from heart disease. In a 10-year prospective study in Yugoslavia of approximately 1400 people, they found that an 11-item questionnaire assessing rationality-antiemotionality, or repression and denial of emotions, was the best single predictor of the subsequent development of CHD. The relative risk for CHD was ten times greater for those who scored high on this scale compared to those who scored lower.
Rationality-antiemotionality was a stronger predictor of heart disease than the traditional CHD risk factors in this study (smoking levels).”
Does Denial Play a Role in the Development of High Blood Pressure, Exaggerated Blood Pressure Reactivity, or High Stress Levels?
Several studies have investigated whether or not hypertensives tend to be less self-disclosing and more likely than normotensives to use denial to blot out negative affect, neurotic conflicts, or awareness of psychiatric disorder.
According a 1989 review of psychosocial variables and hypertension(Sommers-Flanagan and Greenberg, 1989), there were eight such studies from 1979 to 1986. Five of those studies (McLelland, 1979; Drummond, 1982; Cumes, 1983; Svensson and Theorell, 1983; Santonastaso et al., 1984) found elevated blood pressure tended to be associated with less personal disclosure or greater denial of neurotic conflicts, while three found no such relationship (Monk, 1980; Mann, 1984; Cottington et al., 1985). More recent studies have also supported a possible relationship between denial and hypertension. Asymptomatic hypertensives (Knox et al., 1988) , as well as normotensives with a family history of hypertension (Jorgensen and Houston, 1986; Theorell, 1990), seem to express fewer emotions, have a noncomplaining attitude, and lack an ability to differentiate feelings, similar to denial (Karasek, 1990). Although these results are far from conclusive, they do support the existence of a tendency for hypertensives to use denial more frequently than normotensives. However, due to a lack of prospective studies investigating whether denial may play a causative role in the development of hypertension, there is no strong evidence to suggest that denial poses a significant risk to cardiovascular health via increased risk for high blood pressure.
As previously mentioned, normotensives with a positive family history of hypertension (FH+ normotensives) seem to express fewer emotions, have a noncomplaining attitude, and lack an ability to differentiate feelings (Jorgensen and Houston, 1986). FH+ normotensives manifest greater sympathetically driven cardiovascular reactivity to stressors than normotensives with a negative family history(FH-)(Hastrup et al., 1982; Jorgensen and Houston, 1981; Manuck and Proietti, 1982). Furthermore, it appears that a subgroup of these FH+ normotensives characterized by denial and unwillingness to admit to neurotic feelings or aggressiveness may be responsible for the exhibited exaggerated blood pressure reactivity of FH+ normotensives (Jorgensen and Houston, 1986). This subgroup reported little negative affect in response to experimental stressors in which they exhibited exaggerated blood pressure reactivity, which suggests they deny or suppress their feelings. Denial of negative affect in response to experimental stressors and effects on exaggerated blood pressure reactivity have not been sufficiently studied.
In a recent study by Suter et al., 1997, an inverse association between self-perceived stress and systolic blood pressure was found. They suggest that “the inverse association between systolic blood pressure(SBP) and the self-perceived stress reflects a neuroendocrine and biochemical setting characterized by inadequate stress handling associated with a higher fat and alcohol intake and more abdominal fat tissue leading to a higher blood pressure.” They believe their “data suggest that stress denial in combination with abdominal obesity, alcohol consumption, and smoking may be proxy for a high stress level.” Further research is needed investigating whether denial indeed plays an important role in exaggerated blood pressure reactivity or high stress levels.
Social determinants of Denial
There has been remarkably little research on the determinants of denial, particularly social class and job stress. Such factors might shape personality development in childhood. For example, certain parental behavior patterns (i.e., overly strict, critical and demanding of conformity) are more common in low SES households, and may be viewed as a reflection of the parents’ occupational and other life experiences, which are characterized by low control and insecurity. Similarly, an adult’s experience, which might include stressful, low control jobs, may shape their personality development (Kohn and Schooler, 1982). Thus, research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
As Sirous (1992) concluded in his review on denial in CHD, denial likely has a long-term negative effect on cardiovascular health. The extent and importance of that negative effect on cardiovascular health is still quite unknown due to methodological problems concerning the assessment of denial. Prompt medical treatment is crucial to the survival of AMI, and denial of cardiac events may be a primary reason for patient delay. However, denial has yet to be definitively linked to patient delay. Further research is needed on the relationship between denial and patient delay, because if such a link exists, dramatic reductions of patient delay might be possible. As a result, thousands of lives per year could be saved by getting AMI patients to the hospital early enough to receive thrombolytic therapy. Although denial may be an adaptive behavior towards the first three days of recovery from AMI, there is strong evidence that prolonged denial of the significance of the illness negatively effects AMI recovery outcome once removed from the hospital setting.
There is also prospective evidence suggesting that prolonged repression and denial of emotions may greatly increase the risk for CHD. Unfortunately, there has been a lack of prospective studies on denial and repression of emotions as a CHD risk factor since the findings by Grossarth et al., 1985. More prospective studies on the cardiovascular effects of denial of emotions are needed to confirm and better understand the findings of Grossarth et al., 1985.
Other studies have found denial to be associated with elevated blood pressure, exaggerated blood pressure (BP) reactivity, and possibly even high stress levels. Once again, due to the lack of prospective evidence on denial as a risk factor for hypertension, BP reactivity, or high stress levels, there is no strong evidence yet for any these hypotheses.
The study of denial as a psychosocial risk factor for CHD has not been studied nearly as thoroughly as other more heavily focused-on psychosocial risk factors, such as hostility, anger, type A behavior pattern, anxiety, or depression. With improvements in the validity and reliability of assessments of denial, and increased attention from coronary-prone researchers, it will be possible to better understand how much of a role that denial plays in effecting cardiovascular health.
Lastly, very little is known about the determinants of denial. Research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
Allan, R., & Scheidt, S.(1996). The empirical basis for cardiac psychology. In R. Allen and S. Scheidt(Eds.), Heart & Mind: The Practice of Cardiac Psychology(pp. 63-123). Washington D.C.: American Psychological Association.
Bleeker, J., Lamers, L., Leenders, I., Kruyssen, D., Simoons, M., Trijsburg, R., & Erdman, R.(1995). Psychological and knowledge factors related to delay of help-seeking by patients with acute myocardial infarction. Psychother. Psychosom., 63, 151-158.
Colling, A., Dellipiani, A., Donaldson, R., & MacCormack, P.(1976). Teesside coronary survey: an epidemiological study of acute attacks of myocardial infarction. Br. Med. J., ii, 1169-1172.
Cottington, E., Brock, B., House, J., & Hawthorne, V.(1985). Psychosocial factors and blood pressure in Michigan statewide blood pressure survey. Am. J. Epidemiol., 121, 515-529.
Cumes, D.P.(1983). Hypertension, disclosure of personal concerns, and blood pressure response. J. Clin. Psychol., 39, 376-381.
Douglas, C., & Druss, R.(1987). Denial of illness: a reappraisal. Gen. Hosp. Psychiatry, 9, 53-57.
Dracup, K., Moser, D., Eisenberg, M., Meischke, H., Alonzo, A., & Braslow, A.(1995). Causes of delay in seeking treatment for heart attack symptoms. Soc. Sci. Med., 40(3), 379-392.
Drummond, P.D.(1982). Personality traits in young males at risk for hypertension. J. Psychosom. Res., 26, 585-589.
Fields, K.B.(1989). Myocardial Infarction and Denial. Journal of Family Practice, 28(2), 157-161.
Gentry, W., & Haney, T.(1975). Emotional and behavioral reaction to acute myocardial infarction. Heart Lung, 4, 738-745.
Goldstein, S., Moss, A., & Greeene, W.(1972). Sudden death in acute myocardial infarction. Arch. Intern. Med., 129, 720-724.
Grossarth-Maticek, R., Bastiaans, J., & Kanazir, D.(1985). Psychosocial factors as strong predictors of mortality from cancer, ischemic heart disease and stroke: the Yugoslav prospective study. J. Psychosom. Res., 29, 167-176.
Hackett, & Cassem, N.(1968). The coronary-care unit: an appraisal of its psychologic hazards. N. Engl. J. Med., 279, 1365.
Hackett, T., & Cassem, N.(1969). Factors contributing to delay in responding to the signs and symptoms of acute myocardial infarction. Am. J. Cardiol., 24, 651-658.
Hackett, T., & Cassem, N.(1974). Development of a quantitative rating scale to assess denial. J. Psychosom. Res., 18, 93-100.
Hackett, T., & Cassem, N.(1982). Coping with heart disease. Adv. Cardiol., 31, 212-217.
Hastrup, J., Light, K., & Obrsit, P.(1982). Parental hypertension and cardiovascular response to stress in healthy young adults. Psychophysiology, 19, 615-622..
Jamner, L., & Schwartz, G.(1986). Self-deception predicts self-report and endurance of pain. Psychosomatic Medicine, 48, 211-223.
Jorgensen, R.S., & Houston, B.K.(1981). Family history of hypertension, gender, and cardiovascular reactivity and stereotypy during stress. J. Behav. Med., 4, 175-189.
Jorgensen, R.S., & Houston, B.K.(1986). Family History of Hypertension, Personality Patterns, and Cardiovascular Reactivity to Stress. Psychosomatic Medicine, 48, 102-116.
Julkunen, J.(1989). Coping with illness(CILL): Development of a coping scale for coronary patients. Paper presented at the Thrid European Conference on Health Psychology, June 1989, Utrecht, Holland.
Julkunen, J., & Saarinen, T.(1994). Psychosocial predictors of recovery after a myocardial infarction: Development of a comprehensive assessment method. The Irish Journal of Psychology, 15(1), 67-83.
Kabat-Zinn, J.(1992). In I.S. Ockene, & J.K. Ockene(Eds.), Prevention of Coronary Heart Disease (pp.299-332). Boston, MA: Little, Brown and Company.
Karasek, R.A., & Theorell, T.(1990). Healthy Work. New York, NY: Basic.
Keltikangas-Jarvinen, L., & Jokinen, J.(1989). Type A behavior, coping mechanisms and emotions related to somatic risk factors of coronary heart disease in adolescents. J. Psychosom. Res., 33(1), 17-27.
Kennedy, J., Ritchie, J., Davis, K., Stadius, M., Maynard, C., Fritz, J.(1985). The Western Washington randomized trial of intracoronary streptokinase in acute myocardial infarction: a 12-month follow-up report. N. Engl. J. Med., 312, 1073.
Knox, S., Svensson, J., & Waller, D.(1988). Emotional coping and the psychophysiological substrates of elevated blood pressure. Behav. Med., 2, 52-58.
Lazarus, R.S.(1983). The costs and benefits of denial. In S. Breznitz(Ed.), The Denial of Stress. New York, NY: International Universities Press.
Leitch, J., Birbara, T., Freedman, B., Wilcox, I., & Harris, P.(1989). Factors influecing the time from onset of chest pain to arrival at hospital. Med. J. Aust., 150, 6-10.
Levine, J., Warrenburg, S., Kerns, R., et al.(1987). The role of denial in recovery from coronary heart disease. Psychosomatic Medicine, 49, 109-117.
Mann, A.H.(1980). Hypertension: Psychological aspects and diagnostic impact in a clinical trial. Psychol. Med., 5(suppl), 1-35.
Manuck, S.B., & Proietti, J.M.(1982). Parental hypertension and cardiovascular response to cognitive and isometric challenge. Psychophysiology, 19, 481-489.
Marder, V., & Sherry, S.(1988). Thrombolytic therapy: current status (second of two parts). N. Engl. J. Med., 318, 1585.
McClelland, D.C.(1979). Inhibited power motivation and high blood pressure in men. J. Abnorm. Psychol., 88, 182-190.
McKenday, M., & Logan, R.(1982). The recognition and management of denial in patients after myocardial infarction. Aust. N. Z. J. Med., 12, 607-611.
Monk, M.(1980). Psychologic status and hypertension. Am. J. Epidemiol., 112, 200-208.
Prince, R., Frasure-Smith, N., & Rolicz-Woloszyk, E.(1982). Life stress, denial and outcome in ischemic heart disease patients. J. Psychosom. Res., 26, 23-31.
Ramanaiah, N., Schill, T., & Leung, L.(1977). A test of the hypothesis about the two-dimensional nature of the Marlowe-Crowne social desirability scale. J. Res. Personality, 11, 251-259.
Rawles, J., & Haites, N.(1988). Patient and general practitioner delays in acute myocardial infarction. Br. Med. J., 296, 882-884.
Santonastaso, P., Canton, G., Ambrosio, G., et al.(1984). Hypertension and neuroticism. Psychother. Psychosom., 41, 7-11.
Saarinen, T.(1992). Coping with myocardial infarction and return to work. (In Finnish with English summary). Publications of Rehabilitation Foundation, Research reports 35, Helsinki, Finland.
Scalzi, C., & Burke, L.(1982). Myocardial infarction: behavioral responses of patient and spouse. In S. Underhill, S. Woods, E. Sivarajan, & C. Halpenny(Eds.), Cardiac Nursing. Philadelphia, PA: JB Lippincott.
Schmidt, S. & Borsch, M.(1990). The prehospital phase of acute myocardial infarction in the era of thrombolysis. Am. J. Cardiol., 665, 1411.
Simon, A., Feinleib, M., & Thompson, H.(1972). Components of delay in the pre-hospital phase of acute myocardial infarction. Am. J. Cardiol., 30, 476-482.
Simoons, M., Serruys, P., van den Brand, M., Bar, F., de Zwaan, C., Res, J., Verheugt, F., Krauss, X., Remme, W., Vermeer, F., et al.(1985). Improved survival after early thrombolysis in acute myocardial infarction. Lancet, 2(8455), 578.
Sirous, F.(1992). Le deni dans la maladie coronarienne[ Denial in coronary artery disease]. Canadian Medical Association Journal, 147, 315-321.
Sommers-Flanagan, J., & Greenberg, R.(1989). Psychosocial variables and hypertension: a new look at an old controversy. J. of Nervous and Mental Disease, 177(1), 15-24.
Suter, P., Maire, R., Holtz, D., & Vetter, W.(1997). Relationship between self-perceived stress and blood pressure. J. Human Hypertension, 11(3), 171-176.
Theorell, T.(1990). Family history of hypertension- an individual trait interacting with spontaneously occurring job stressors. Scand. J. Work Environ. Health, 16 (suppl. 1), 74-79.
Trijsburg, R., Bal, J., Parsowa, W., Erdman, R., & Duivenvoorden, H.(1989). Prediction of physical indisposition with the help of a questionnaire for measuring denial and overcompensation. Psychother. Psychosom., 51, 193-202.
Trijsburg, R., Erdman, R., Duivenvoorden, H., Thiel, J., & Verhage, F.(1987). Denial and overcompensation in male patients with myocardial infarction. Psychother. Psychosom., 47, 22-28.
Viswanathan, R., & Vizner, T.(1984). The experience of myocardial infarction as a threat to oneUs adequacy. Gen. Hosp. Psychiatry, 6, 83-89.
White, H., Norris, R., Brown, M., Takayama, M., Maslowski, A., Bass, N., Ormiston, J., & Whitlock, T.(1987). Effect of intravenous streptokinase on left ventricular function and early survival after acute myocardial infarction. N. Engl. J. Med., 317, 850.
Wielgosz, A., & Nolan, R.(1991). Understanding delay in response to symptoms of acute myocardial infarction. Circulation, 84, 2193-2195.
Wielgosz, A., Nolan, R., Earp, J., Biro, E., & Wielgosz, M.(1988). Reasons for patientsU delay in response to symptoms of acute myocardial infarction. Can. Med. Assoc. J., 139, 853-857.
Yanagida, E., Streltzer, J., & Siemsen, A.(1981). Denial in dialysis patients: Relationship to compliance and other variables. Psychosomatic Medicine, 43, 271-280.
Yusuf, S., Collins, R., Peto, R., Furberg, C., Stampfer, M., Goldhaber, S., & Hennekens, C.(1985). Intravenous and intracoronary fibrinolytic therapy in acute myocardial infarction: overview of results on mortality, reinfarction and side-effects from 33 randomized controlled trials. Eur. Heart J., 6, 556.
Large amounts of research has been aimed at substantiating hypothesized psychosocial risk factors for Coronary Heart Disease (CHD), such as Type A Behavior Pattern (TABP), hostility, and anxiety. Comparatively little work, however, has examined whether depression is a significant risk factor for CHD or how it influences other more widely accepted CHD risk factors. The following will contain a review of the literature on depression and cardiovascular disease risk, as well as how depression is assessed and how other known CHD risk factors may be related to depression.
What is depression and how is it assessed?
According to H. Friedman in his book “The Self-Healing Personality”, depression generally involves feelings of sadness, tiredness, indecisiveness, and worthlessness. Depressed people tend to be aware of their problems, and are likely to be prone to excessive stress which manifests itself in a disruption of hormonal systems that are used to maintain internal homeostasis.
The most frequently used measure of depression is probably the Depression scale of the Minnesota Multiphasic Personality Inventory(MMPI). Numerous other measures for depression exist including the Profile of Mood States, the Welsh Depression Scale, the Dempsey Depression Scale, and simple ratings of depression made by an interviewer (Booth-Kewley and Friedman, 1987).
Associations of anxiety with CVDs and known CVD risk factors
The most commonly cited evidence that depression is reliably related to CHD is Booth-Kewley and Friedman’s 1987 meta-analysis of psychological predictors of heart disease. Surprisingly, out of the 18 personality-variable categories that Booth-Kewley and Friedman performed separate meta-analyses on, only SI-assessed TABP was found to be more reliably related to CHD outcomes than depression (r=.205). Depression was found to be a better predictor for CHD outcome than hostility, anger, or anxiety. All 15 studies included in their meta-analysis on depression and CHD used some manifestation of CHD or atherosclerosis as a dependent variable. When their analysis was performed separately for only the six prospective studies, depression was again strongly related to CHD outcome (r=.168), the strongest association seen when only prospective studies were included in the meta- analyses (Goldstein and Niaura, 1992). Although they conclude that more attention should be focused on depression as a component of the coronary-prone personality, the findings remain in question even by Friedman and Booth-Kewley (1988) due to numerous questions over the validity of the studies included. Further doubt over whether depression is a risk factor for CHD comes from Matthews’ 1988 meta-analysis of prospective studies on psychological predictors of heart disease. He did not find depression to be a predictor for coronary artery disease (CAD).Matthews was more conservative in the studies he selected and weighted studies according to the number of participants.
Although patients who develop cardiovascular disease are at a greater risk for subsequent depression (Carney et al., 1988; Littman, 1993), evidence for the converse is not as convincing. In four studies of institutionalized patients diagnosed with depression, two report a positive association between depression and cardiovascular mortality (Baldwin,1980; Dreyfuss et al., 1969), while two fail to report an association (Martin et al., 1985; Tsuang et al., 1980). It should be noted that these studies have been questioned in their validity due to a number of design problems. None of these studies controlled for known cardiovascular disease risk factors. Secondly, these studies were retrospective in design, and therefore any possible cause and effect relationships between depression and cardiovascular disease could not be distinguished. In fact, subjects with symptoms of depression, such as fatigue, palpitations, and decreased energy, may have had undiagnosed cardiovascular problems before the start of the study, because these symptoms are also incipient manifestations of cardiovascular disease (Hayward, 1995). Another problem of these studies comes from using institutionalized patients. This particular bias overestimates the degree of the association, because a patient who has two illnesses, a physical illness and a psychiatric diagnosis, is more likely to be hospitalized than a patient with only one disorder (Berkson, 1946). One study by Murphy et al., 1987 has avoided these particular study design pitfalls. In this prospective study where subject selection was from a community sample, the risk for subsequent cardiovascular disease death was higher among individuals with an affective disorder than among individuals without an affective disorder. The positive association in this study was larger for males(relative risk = 2.5) than for females(relative risk = 1.5).
It has been suggested that if depression increases the risk for later cardiovascular disease, the relationship may be non-linear (Hayward, 1995). In other words, low to intermediate levels of depression, a quite common experience, may have no ascertainable effect of cardiovascular health, but high levels of depression, predominantly only seen in the clinical population, may then pose significant harm to cardiovascular health. Evidencethat low to intermediate levels of depression does not greatly increase cardiovascular disease risk comes from studies that use self-report measures of depression obtained from nonclinical samples. In four prospective studies, there has been no evidence that self-reported symptoms of depression predict cardiovascular disease (Ostfeld et al., 1964; Vogt. et al., 1994; Brozek et al., 1966; Goldberg et al., 1979).
However, many recent studies investigating the relationship between depressive symptoms and CHD outcomes in nonclinical samples contradict these null findings, and suggest that self-reported symptoms of depression do actually predict cardiovascular disease. In a study of 2,832 initially healthy U.S. adults, self-reported depressed affect and hopelessness were associated with increased relative risk of fatal and nonfatal ischemic heart disease over a mean follow-up period of 12.4 years (Anda et al., 1993). Depressive symptoms, assessed using the Depression scale of the MMPI, have also been associated with increased risk of myocardial infarction (MI) in a prospective study of 409 initially healthy males and 432 initially healthy females (Barefoot and Schroll, 1996). Lastly, a recent cross-sectional study, where depression was assessed through psychiatric interview of a nonclinical random sample in Finland, has found significant positive associations between depression and increased risk for CVDs (Aromaa et al., 1994). The age-adjusted relative risk of coronary heart disease of those ages 40-64 diagnosed with depression was 4.87 (CI, 2.91-8.16) compared to those ages 40-64 who were free of psychiatric diagnosis. The hypothesis that depression is a risk factor for CVDs definitely deserves further study.
Although the evidence is inclusive in regards to whether depression is a significant risk factor in the development of CHD, there is substantial evidence that depression is associated with known cardiovascular disease risk factors, particularly cigarette smoking. According to Hayward, 1995, 22 out of 25 studies reporting the frequency of cigarette smoking in depressed versus nondepressed people found increased rates of cigarette smoking in the depressed. Additionally, depressed subjects are less successful in attempts to quit smoking (Glassman et al., 1990). In one 9-year follow-up study, depressed smokers were 40 percent less likely than nondepressed smokers to have quit smoking (Anda et al., 1990).
There are far fewer studies investigating whether depression is associated with other known cardiovascular disease risk factors, but positive associations of depression with hypertension have been reported. Cross-sectional studies of depressed persons provide evidence of increased sympathetic activity and increased blood pressure reactivity (Jonas et al., 1997), and suggest that depression may have a pressor effect on the cardiovascular system that could lead to the development of hypertension (Julius, 1988). According to Hayward, 1995, three cross-sectional studies (Heines et al, 1969; Heines, 1970; Reus and Miner, 1985) have found higher rates of hypertension in depression patients, while two cross-sectional studies did not find such an association (Friedman and Bennet, 1977; Yates and Wallace,1987). In the Yates and Wallace study, although higher rates of hypertension were not found in patients with unipolar affective disorder (depression), higher rates of hypertension were found in patients with bipolar affective disorder (mania). According to Jonas et al., 1997, unfortunately only one prospective study (Goldberg et al., 1980) has examined the effects of depression on the subsequent development of hypertension. No association was observed, possibly because of a poor study design that consisted of few subjects (640) and only a 1-3 year follow-up period. Jonas et al., 1997 have further investigated whether symptoms of depression are risk factors for hypertension in their own prospective study that consisted of more subjects (2992) and a much longer follow-up period of 7-16 years. Symptoms of depression were assessed using the General Well-Being Schedule cheerful vs. depressed scale (Fazio, 1977). The risk of experiencing hypertension was increased among whites aged 45 to 64 years who had high depression symptoms scores compared with those who had low symptom scores(Relative risk=1.80, confidence interval, 1.16-2.78) but not among whites aged 25 to 44 years. The risk of experiencing hypertension was even greater among blacks aged 25 to 64 years who had high depression symptom scores (RR=2.99, CI, 1.41-6.33).
Hypertension and smoking are most likely not the only CHD risk factors that will be shown to be more prevelant in the depressed. Physical fitness is a CHD risk factor that has not been adequetely studied among the depressed.
Some studies report physical inactivity as a risk factor for later onset of depression or depressive symptoms (Farmer et al., 1988; Frederick et al., 1988). It is unclear whether physical activity is a risk factor and/or a symptom of depression. There are also strong theoretical ties between depressed mood and various coronary-prone behaviors (Allan and Scheidt, 1996). Type A Behavior Pattern (TABP) is thought to be a reaction against an underlying, unconscious depression. As long as the Type A individual is engaged in driven, ambition-related activities, he or she is protected from awareness of the depression. When such, activities cease, however, such as after acute MI, depression often becomes unmasked. Clearly, depression is embodied within the cynical and pessimistic orientation to life assessed by the Cook-Medley Ho scale, a standard self-report measure of hostility. In classic psychoanalytic theory, depression is sometimes considered “anger turned inward,” suggesting a strong theoretical link between hostility and depression. Lastly, depression has been found to contribute to self-destructive behavior and motivates people in the direction of “quick fix” forms of satisfaction, such as cigarettes, alcohol, and fast foods high in saturated fat (Allan and Scheidt, 1996).
It is important to keep in mind that depressed patients are a very heterogeneous group. Just as only those depressed patients with increased rates for known CVD risk factors may be at a greater risk for CVDs, it is possible that only certain types of depression lead to substantially increased risk for CVDs. Fava et al., 1996 hypothesized that patients with anxious or hostile depression may have a greater risk of mortality from coronary artery disease(CAD) than other depressed patients. They tested their hypothesis by assessing the relationship between CAD risk factors and anxiety and hostility in a sample of 138 depressed outpatients. They found that patients with anxious or hostile depression had a profile of greater CAD risk compared with patients with depression who had low anxiety and hostility scores. The most significant association seen in the anxious or hostile depression groups was with increased serum cholesterol levels. Fava et al., 1997 further suggest that the heterogeneity of major depression might account for null findings among unipolar depressed patients of increased risk for CVDs and known CVD risk factors. Limitations of the study include that their findings are derived from outpatients with mild-to-moderate major depression and may not be generalizable to more sever forms of depression. Also, this study did not control for other known CVD risk factors. Although more studies are surely needed before any reliable conclusions can be made, depression may only lead to significantly increased CVD risk when psychosocial CVD risk factors other than depression, such as hostility and anxiety, are present as well.
Depression, Job strain, and CHD
“Job strain” has been defined by Karasek (1979) as work in jobs with high psychological demands (work pace + conflicting demands) and low decision latitude (control + variety and skill use). In a half-dozen epidemiological studies over the last decade, occupational stress researchers have found job strain to be a significant risk factor for CHD. To our knowledge, there have been no published studies examining possible relationships between depression and job strain. Although it has been suggested, the potential influence of job characteristics in the development of psychological states, such as depression, has rarely been studied (Schnall et al., 1994). According to Schnall et al., 1994, it is unlikely though that personality variables such as depression account for the association between job strain and hypertension or CVD. Research is needed to develop a theory that specifies possible interactions between environmental stressors(e.g., job strain) and personality characteristics(e.g., depression), and to improve the methodology of such studies.
Depression and its determinants
There has been remarkably little research on the determinants of depression, particularly social class and job stress. Such factors might shape personality development in childhood. For example, certain parental behavior patterns (i.e., overly strict, critical and demanding of conformity) are more common in low SES households, and may be viewed as a reflection of the parents’ occupational and other life experiences, which are characterized by low control and insecurity. Similarly, an adult’s experience, which might include stressful, low control jobs, may shape their personality development (Kohn and Schooler, 1982). Thus, research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
Although it is widely accepted that depression is commonly observed in CHD patients, the question of whether depression can significantly contribute to the etiology of CHD remains in debate. Recent prospective studies that have found depressive symptoms to predict common CHD endpoints, such as myocardial infarction and ischemic heart disease, provide strong evidence for considering depression a CHD risk factor, but more studies of this sort are needed to settle this issue. There is overwhelming evidence, however, that depressed people are more likely to lead a self-destructive lifestyle that leads to increased risk for other known CHD risk factors, primarily smoking. Furthermore, depression has been theoretically tied to more proven coronary-prone behaviors, such as TABP and hostility. One possibility for the null findings regarding associations between depression and CVDs is that only certain types of depression may pose harm to cardiovascular health. As one study (Fava et al, 1996) suggests, only anxious or hostile depressed patients may be at significantly increased risk for future development of CVDs. More studies are needed which investigate how depression interacts with other heavily studied coronary-prone behaviors in relation to their combined effects on cardiovascular health. Similarly, further research is needed to develop theories as to how hypothesized CHD risk factor personality characteristics, such as depression, interact with environmental stressor models of CHD risk, such as job strain, and improve the methodology of such studies. Lastly, additional research is needed to elucidate the social determinants of depression, particularly social class and job stress.
Allan, R., & Scheidt, S.(1996). Empirical Basis for Cardiac Psychology. R. Allan & S. Scheidt(Eds.), Heart and Mind: The Practice of Cardiac Psychology(pp. 63-123). Washington, D.C.: American Psychological Association.
Anda, R., Williamson, D., Escobedo, L., Mast, E., Giovino, G., & Remington, P.(1990). Depression and the dynamics of smoking: anational perspective. JAMA, 264, 1541-1545.
Anda, R., Williamson, D., Jones, D., Macera, C., Eaker, E., Glassman, A., & Marks, J.(1993). Depressed Affect, Hopelessness, and the Risk of Ischemic Heart Disease in a Cohort of U.S. Adults. Epidemiology, 4, 285-293.
Aromaa, A., Raitasalo, R., Reunanen, A., Impivaara, O., Heliovaara, M., Knekt, O., Lehtinen, V., Joukamaa, M., & Maatela, J.(1994). Depression and cardiovascular diseases. Acta Psychiatr Scand, Suppl. 377, 77-82.
Baldwin, J.(1980). Schizophrenia and physical disease: a preliminary analysis of the data from the Oxford Record Linkage Study. In G. Hemmings (Eds.), The biochemistry of schizophrenia and addiction: in search of a common factor(pp. 297-318). Lancaster, England: MTP Press.
Barefoot, J., & Schroll, M.(1996). Symptoms of depression, acute myocardial infarction, and total mortality in a community sample. Circulation, 93, 1976-1980.
Berkson, J.(1946). Limitations of the application of fourfold table analysis to hospital data. Biometrics, 2, 47-53.
Booth-Kewley, S., & Friedman, H.(1987). Psychological Predictors of Heart Disease: A Quantitative Review. Psychological Bulletin, 101(3), 343-362.
Brozek, J. Keyes, A., & Blackburn, H.(1966). Personality differences between potential coronary and noncoronary subjects. Ann NY Acad Sci, 134, 1057-1064.
Carney, R., Rich, M., Freedland, K., et al.(1988). Major depressive disorder predicts cardiac events in patients with coronary artery disease. Psychosomatic Medicine, 50, 627-633.
Dreyfuss, F., Dasberg, H., & Assael, M.(1969). The relationship of myocardial infarction to depressive illness. Psychother Psychosom, 17, 73-81.
Farmer, M., Locke, B., Moscicki, E., et al.(1988). Physical activity and depressive symptoms: the NHANES I Epidemiologica Follow-up Study. Am J Epidemiology, 128, 1340-1351.
Fava, M., Abraham, M., Pava, J., Shuster, J., & Rosenbaum, J.(1996). Cardiovascular Risk Factors in Depression: The Role of Anxiety and Anger. Psychosomatics, 37, 31-37.
Fazio, A.(1977). A concurrent validational study of the NCHS General Well-Being Schedule. Vital Health Stat 2, No. 73.
Frederick, T., Fredrichs, R., & Clark, V.(1988). Personal health habits and symptoms of depression at the community level. Prev Med, 17, 173-182.
Friedman, H.(1991). In H. Friedman(Eds.), The Self-Healing Personality: why some people achieve health and others succumb to illness. New York: H. Holt.
Friedman, H. & Bennet, P.(1977). Depression and hypertension. Psychosom Med, 39, 134-142.
Friedman, H. & Booth-Kewley, S.(1988). Validity of the type A construct: a reprise. Psychological Bulletin, 104, 381-384.
Glassman, A., Helzer, J., Covey, L., et al.(1990). Smoking, smoking cessation, and major depression. JAMA, 264, 1546-1549.
Goldberg, E., Comstock, G., & Graves, C.(1980). Psychological factors and blood pressure. Psychol Med, 10, 243-255.
Goldberg, E., Comstock, G., & Hornstra, R.(1979). Depressed mood and subsequent physical illness. Am J Psychiatry, 136, 530-534.
Goldstein, M., & Niaura, R.(1992). Psychological factors affecting physical condition. Cardiovascular disease literature review. Psychosomatics, 33, 134-145.
Hayward, C.(1995). Psychiatric Illness and Cardiovascular Disease Risk. Epidemiologic Reviews, 17(1), 129-138.
Heine, B.(1970). Psychogenesis of hypertension. Proc R Soc Med, 63, 1267-1270.
Heine, B., Sainsbury, P., & Chynoweth, R.(1969). Hypertension and emotional disturbance. J Psychiatr Res, 7, 119-130.
Jonas, B., Franks, P., & Ingram, D.(1997). Are Symptoms of Anxiety and Depression Risk Factors for Hypertension? Longitudinal Evidence From the National Health and Nutrition Examination Survey I Epidemiologic Follow-up Study. Arch Fam Med, 6, 43-51.
Julius, S.(1988). Transition from high cardiac output to elevated vascular resistance in hypertension. Am Heart J, 116, 600-606.
Karasek, R.(1979). Job demands, job decision latitude, and mental strain: Implications for job redesign. Adm. Sci. Q., 24, 285-308.
Martin, R., Cloninger, C., Guze, S., et al.(1985). Mortality in a follow-up of 500 psychiatric outpatients. II. Cause-specific mortality. Arch Gen Psychiatry, 42, 58-66.
Murphy, J., Monson, R., Olivier, D., et al.(1987). Affective disorders and mortality. Arch Gen Psychiatry, 44, 473-480.
Ostfeld, A., Lebovits, B., Shekelle, R., et al.(1964). A prospective study of the relationship between personality and coronary heart disease. J Chronic Dis, 17, 265-276.
Reus, V., & Miner, C.(1985). Evidence for physiological effects of hypercortisolemia in psychiatric patients. Psychiatry Res, 14, 47-56.
Schnall, P., Landsbergis, P., & Baker, D.(1994). Job strain and cardiovascular disease. Annu. Rev. Public Health, 15, 381-411.
Tsuang, M., Woolson, R., & Fleming, J.(1980). Causes of death in schizophrenia and manic-depression. Br J Psychiatry, 136, 239-242.
Vogt, T., Pope, C., Mullooly, J., et al.(1994). Mental health status as a predictor of morbidity and mortality: a 15-year follow-up of members of a health maintenance organization. Am J Public Health, 84, 227-231.
Yates, W., & Wallace, R.(1987). Cardiovascular risk factors in affective disorder. J Affect Disord, 12, 129-134.
The idea that a link may exist between anxiety and the heart has been around for as long as the history of medicine has been documented. However despite a wide spread public perception that stress and anxiety are significant risk factors for coronary heart disease(CHD), numerous conceptual and methodological difficulties in studying whether a relationship between anxiety and CHD exists have scared away many researchers from even attempting such studies(Byrne and Rosenman, 1990). It is only very recently, with advances in methodology, that possible associations between certain types of anxiety and CHD have been uncovered.
What is anxiety and how is it assessed?
Anxiety is generally defined as a psychobiological emotional state or reaction that can be distinguished most clearly from other emotions such as anger or sadness by its experiential qualities. An anxiety state consists of unpleasant feelings of tension, apprehension, nervousness, and worry, and activation of the autonomic nervous system. The physiological manifestations in anxiety generally include increased blood pressure; rapid heart rate(palpitations or tachycardia); sweating; dryness of mouth; nausea; vertigo; irregularities in breathing; muscle tension; and muscular-skeletal disturbances such as restlessness, tremors, and feelings of weakness(Spielberger and Rickman, 1990).
Anxiety also refers to relatively stable individual differences in anxiety-proneness as a personality trait. People who have high trait anxiety are most likely to perceive stressful situations as being personally dangerous or threatening and to respond to such situations with elevations in state anxiety. The stronger the anxiety trait, the more often the individual has experienced state anxiety in the past, and the greater the probability that intense elevations in state anxiety will be experienced in threatening situations in the future (Spielberger and Rickman, 1990).
State and trait anxiety are measured using a variety of different approaches. Projective techniques such as the Rorschach inkblots and the Thematic Apperception Test are used extensively in the clinical evaluation of anxiety. However, these tests do not lend themselves to quantification and are therefore limited in the extent they can be used in research. Rating scales and psychometric self-report inventories and questionaires are by far the most popular procedures for assessing anxiety in research. An advantage of such instruments is that they are easily administered and scored and do not require a great deal of expensive professional time. The standard rating scale to measure anxiety, the Hamilton Anxiety Rating Scale(HARS- Hamilton, 1959), is composed of 100 symptoms of anxiety evaluated by the clinical examiner that are aggregated to define 13 scale variables. The most commonly used psychometric self-report inventory is probably Spielberger’s State-Trait Anxiety Inventory (STAI- Spielberger et al., 1970). Major revisions were made to STAI in 1979 to develop a “purer” measure of anxiety in order to provide a firmer basis for differentiating anxiety disorders from depressive reactions. Many other self-report psychometric inventories and questionnaires exist such as the Taylor Manifest Anxiety Scale(MAS-Taylor, 1953), Cattell’s Trait and State Anxiety Measures(Cattell and Scheier, 1963), the Affect Adjective Check List(AACL- Zuckerman and Lubin, 1965), the SCL-90 Symptom Check List(SCL-90 – Derogatis et al., 1973), the Profile of Mood States(POMS- McNair et al., 1971), and the Crown-Crisp experiental index(Crown and Crisp, 1966). (Spielberger and Rickman, 1990).
Associations of anxiety with CHD, cardiovascular disorders, and known cardiovascular disease risk factors
In the mid-1970’s, three studies(Jenkins, 1976; Lebovitis et al., 1975; Zyzanski et al., 1976) reported a positive association between anxiety and coronary heart disease. However, these studies and other studies in the 1960’s and 1970’s showed little evidence that anxiety plays a causal role in the pathogenesis of CHD. These positive assocations can be mostly explained by assuming that patients diagnosed or suffering from cardiovascular problems commonly suffer from anxiety over their cardiovascular health. Five studies during this time period examined the relationship between self-reported anxiety and the risk of subsequent cardiovascular disease. Three of them showed positive associations(Paffenberger et al., 1966; Medalie et al., 1973; Thiel et al., 1973), while the other two found negative associations(Thorne et al., 1968; Wardell and Bahnson, 1973). None of these studies controlled for known cardiovascular disease risk factors.
Many studies in the early 1980’s found that coronary atherosclerosis is nfrequent in subjects with anxiety states(Bass, 1984; Costa, 1981; DeMaria et al., 1980; Schocken et al., 1984; Sprafkin et al., 1984), even suggesting that anxiety may play a protective role against atherosclerosis(Rosenman, 1990). On the basis of finding positive associations between anxiety and angina pectoris(chest pain) and negative relationships between anxiety and coronary artery disease, Costa et al., 1985, suggested that anxiety was a risk factor for being referred for oronary angiography but not for coronary artery disease.
However, also in the 1980’s and continuing into the early 1990’s, six studies examined cardiovascular disease outcomes in patients with anxiety disorders. In two studies, there was a higher than expected death rate from diseases of the circulatory system in patients with panic disorder (Coryell et al., 1982; Coryell et al., 1986). In a third study, anxiety neurosis was associated with higher death from arterioschlerotic disease(Sims and Prior, 1982). Finally, a fourth study, utilizing the Epidemiologic Cachment Area database, also found an association between panic disorder and cardiovascular disease(Weissman et al., 1990). The other two studies failed to observe a relation between anxiety and cardiovascular disease. All of these studies were retrospective, they did not control for known cardiovascular disease risk factors, and, with the exception of one study(Weissman et al., 1990), they used registries for the purpose of case identification.
The more compelling evidence for a relationship between anxiety and CHD comes from studies examining self-reported anxiety and the risk of subsequent cardiovascular disease. As previously mentioned, these types of studies in the 1960’s and 1970’s produced conflicting results. However, in the 1980’s and 1990’s, there have been three such studies, all of which found strong positive associations between anxiety and cardiovascular disease(Haines et al., 1987; Kawachi, Colditz et al., 1994l Kawachi, Sparrow et al., 1994). Additionally, all three of these studies contained large samples, were prospective, and controlled for the effect of known CHD risk factors. Haines et al.(1987) and Kwachi, Colditz et al.(1994) both found that the risk of cardiovascular disease specifically related to the phobic anxiety scale of the Crown-Crisp index(Crown and Crisp, 1966). The study by Kwachi, Colditz et al.(1994) also found that the risk of sudden coronary death to be greater than the risk of nonsudden coronary death. Lastly, the most recent study (Kawachi and Sparrow et al., 1994) used a five-item anxiety symptom scale out of questions from the Cornell Medical Index (Coryell et al., 1982). The five questions selected were similiar to existing, validated psychological assessment scales such as STAI, HARS, and the Crown-Crisp index. Compared with men reporting no symptoms of anxiety, men reporting two or more symptoms had an elevated risk of fatal CHD, and particularly sudden coronary death(after adjusting for potential confounding variables, age-adjusted odds ratio(OR)= 4.46, 95% confidence interval: 0.92-21.6). Unfortunately, all of these studies involved only males. Anxiety disorders are 2-3 times more common in females, and therefore it will be important to replicate these findings in women(Hayward, 1995).
Anxiety has also been linked to known CHD risk factors. High cholesterol levels, a known CHD risk factor, have been reported in patients with anxiety disorders but not in patients with affective disorders(Hayward et al., 1989; Bajwa et al., 1992; Fava et al., 1994). According to a recent review (Hayward, 1995), only only study(Tancer et al., 1990) did not find elevated cholesterol levels in patients with anxiety disorders, suggesting that trait anxiety may be linked to high cholesterol levels. Anxiety has also been hypothesized to be a risk factor for hypertension. According to Hayward, 1995, several studies have found higher rates of hypertension in patients with anxiety disorders(Friedman and Bennet, 1977; Noyes et al., 1980; Katon, 1984; Davidson et al., 1991), while only two studies have failed to find this association (Dunner, 1985; Charney and Heninger, 1986). Most of the studies that have investigated the possible association between anxiety and the development of hypertension have been cross-sectional or have been limited because they had small sample sizes, used persons with borderline hypertension, or had been conducted on other highly selected populations (Jonas et al., 1997).
According to Jonas et al., 1997, a look at prospective studies of anxiety as predictors of hypertension shows that two recent studies did find a positive association for specific age and sex groups (Markovitz et al., 1991; Markovitz et al., 1993), while four earlier studies did not (Jenkins et al., 1983; Russek et al., 1990; Kahn et al., 1972; Sparrow et al., 1982). However, it should be noted that three out of four of these earlier studies did not use a standardized measure of anxiety. Both studies by Markovitz et al., which did find a positive association between anxiety and later development of hypertension, did use a standardized measure of anxiety. Jonas et al., 1997, has provided further evidence favoring a link between anxiety and increased risk for hypertension using longitudinal evidence from the National Health and Nutrition Examination Survey I Epidemiologic Follow-up Study. Relative risks for hypertension incidence were determined on the basis of age, race, and level of anxiety symptoms. Blacks between the ages of 25-64 and whites between the ages of 45-64, who had a high level of anxiety symptoms, were found to have a higher relative risk of hypertension compared to groups of the same age range and race that possessed intermediate or low levels of anxiety symptoms. Blacks(age 25-64) with high levels of anxiety symptoms showed the highest relative risk of hypertension during the follow-up, with a relative risk of 3.24 of this group later being subscribed antihypertensive medications. In comparison, blacks (age25-64) with low levels of anxiety symptoms had a baseline relative risk of 1.00. Since hypertension is more common in blacks than whites (Fries, 1973; James et al., 1976), it is not surprising that the possible effects of anxiety on hypertension development were most easily seen in blacks. This might be part of the reason why Somova et al.(1995) found that in young blacks(ages 18-23) but not in whites of the same age, anxiety was found as a partial predictor of hypertension. The evidence as a whole is not conclusive, but recent research particularly suggests a link between anxiety and increased risk for hypertension.
If anxiety is indeed a risk factor for hypertension and high serum cholesterol, it still does not necessrily mean that it is the emotional state of anxiety that directly leads to these physiological effects. Anxiety may influence behaviors and habits that are relevant to cardiovascular disorders, which then lead to development of physiological cardiovascular disease risk factors such as hypertension and high serum cholesterol levels (Rosenmen, 1990). An example of a behavioral CHD risk factor is physical inactivity or lack of fitness. There is evidence that psychiatric patients with anxiety disorders, specifically panic disorder, are more likely to be physically unfit(Taylor et al., 1987; Gaffner et al., 1988). Although this area has not been adequately studied to make any conclusions about how the relationship between anxiety and physical fitness effects cardiovascular health, it serves as an example as to how anxiety could influence behavior that is relevant to the development of CHD.
Finally, it must be noted that commonly used measures of trait anxiety have been found to have a distinct lack of validity and instead seem to measure a single general trait labeled dysphoria, neuroticism, or negative affectivity (Smith, 1985). Individuals high on this trait are hypothesized to report(italicized) more dissatisfaction and have low psychological well-being. Past studies on the effects of anxiety on CHD have usually not controlled for the possible confounding effects of “negative affectivity” or “neuroticism.”
Is trait anxiety linked to cardiovascular reactivity to stress only during anxiety provoking situations?
Investigators have proposed that excessive physiological cardiovascular reactivity (CVR) to psychological stress may constitute an important physiological link which mediates the relationships between psychological traits and CHD (Glass, 1977; Goldband et al., 1979; Kranz and Manuck, 1984). If this is so for anxiety, it is expected that anxiety levels should predict CVR to stressful situations. While studies addressing the issue of whether trait anxiety is linked to CVR to stress have not produced convincing evidence (Houston, 1986), partial relationships have been found for anxiety-specific trait measures and cardiovascular reactivity. Test anxiety, social anxiety, performance anxiety, and specific phobias have all been found in separate studies to be related to cardiovascular reactivity, but many null findings have also been found. One hypothesis accounting for null findings include the failure to match specific dimensions of anxiety with situations that are appropriately anxiety-provoking (Burns, 1995). Relationships observed in most studies between stimulus-specific anxiety(eg., test anxiety, social anxiety) and cardiovascular reactivity most likely depend on whether the laboratory task mimics natural anxiety-producing stimuli (Ward, 1990). Turner et al.(1986) found evidence supporting the importance of specific situations emphasizing social evaluative cues in distinguishing an association between social anxiety and CVR. Additionally, studies by Burns (1995) support the hypothesis that specific dimensions of anxiety may be related to CVR only under appropriately anxiety-provoking situations.
Anxiety, job strain, psychological distress and CHD
“Job strain” has been defined by Karasek(1979) as work in jobs with high psychological demands (work pace + conflicting demands) and low decision latitude (control + variety and skill use). In a half-dozen epidemiological studies over the last decade, occupational stress researchers have found job strain to be a significant risk factor for CHD. Furthermore, Stansfeld et al.(1995) have found positive associations between anxiety and conflicting psychological work demands, work pace, and low control. Few studies have examined whether work environments that are high in job strain are particularly anxiety-provoking, although Bourbonnais et al.(1996) provide evidence for a positive association between job strain and psychological distress. Psychological distress was assessed in this study using the Psychiatric Symptom Index, which measures the presence and intensity of anxiety, aggressivity, depressive symptoms, and cognitive trouble. In contrast, in the New York City blood pressure study, trait anxiety was not correlated with job strain (Landsbergis et al., 1992; Schnall et al., 1992). However, the measures of trait anxiety used in this study lacked validity in that they are more a measure of negative affectivity. The relationship between high job strain and high employee anxiety remains for the most part unknown. Furthermore, how job strain and anxiety interact in regards to cardiovascular health also remains largely unknown.
Anxiety and its determinants
There has also been remarkably little research on the determinants anxiety, particularly social class and job stress. Such factors might shape personality development in childhood. For example, certain parental behavior patterns (i.e., overly strict, critical and demanding of conformity) are more common in low SES households, and may be viewed as a reflection of the parents’ occupational and other life experiences, which are characterized by low control and insecurity. Similarly, an adult’s experience, which might include stressful, low control jobs, may shape their personality development (Kohn and Schooler, 1982). Thus, research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
Although there is a large amount of conflicting results in the literature regarding the hypothesized link between high levels of anxiety and CHD, recent research for the most part supports that the relationship exists to some extent. Recent positive findings tying anxiety to higher serum cholesterol levels and later development of hypertension showed anxiety may be related to a number of important cardiovascular disease risk factors. There is an obvious need for finding better ways to assess anxiety, and past lack of validity in anxiety assessment adds uncertainty to the positive findings, as well as to the null findings. Possibly, associations will be easier to distinguish if more consideration is taken into what situations are best used to elicit the dimensions of anxiety being studied. There is also a lack of studies that have examined the relationship between job strain and anxiety, and whether they are independent or mutually reinforcing risk factors for the development of coronary heart disease. Finally, more research is needed on the determinants of anxiety, such as social class and job stress.
Bajwa, W.K., Asnis, G.M., Sanderson, W.C., et al.(1992). High cholesterol levels in patients with panic disorder. American Journal of Psychiatry, 149, 376-378.
Bass, C.(1984). Mitral vavle prolapse and anxiety disorders: No longer an enigma. British Journal of Psychiatry, 145, 447-448.
Bourbonnais, R., Brisson, C., Moisan, J., & Vezina, M.(1996). Job strain and psychological distress in white-collar workers. Scandanavian Journal of Work, Environment, and Health, 22, 139-145.
Bryne, D.G., & Rosenman, R.H.(1990). In D.G. Bryne and R.H. Rosenman (Eds.), Anxiety and the Heart(pp. xv-xvii). New York: Hemisphere Publishing Corporation.
Burns, J.W.(1995). Interactive Effects of Traits, States, and Gender on Cardiovascular Reactivity During Different Situations. Journal of Behavioral Medicine, 18(3), 279-303.
Cattell, R.B., & Scheier, I.H.(1963). Handbook for the IPAT Anxiety Scale (2nd Ed.) Champaign, IL: Institute for Personality and Ability Testing.
Charney, D.S., & Heninger, G.R.(1986). Abnormal regulation of noradrenergic function in panic disorders: effects of clonidine in healthy subjects and patients with agoraphobia and panic disorder. Archives of General Psychiatry, 43, 1042-1054.
Coryell, W., Noyes, R. Jr., & Clancy, J.(1982). Excess mortality in panic disorder: a comparison with unipolar depression. Archives of General Psychiatry, 39, 701-3.
Coryell, W., Noyes, R. Jr., & House, J.D.(1986). Mortality among outpatients with anxiety disorder. American Journal of Psychiatry, 143, 508-510.
Costa, P.T., Zonderman, A., Engel, B., et al.(1985). The relation of chest pain symptoms to angiographic findings of coronary artery stenosis and neuroticism. Psychosomatic Medicine, 47, 285-293.
Costa, P.T.(1981). Neuroticism as a factor in the diagnosis of angina pectoris. Behavioral Medicine Update, 3, 18-20.
Crown, S., & Crisp, A.H.(1966). A short clinical diagnostic self-rating scale for psychoneurotic patients. The Middlesex Hospital questionnaire. British Journal of Psychiatry, 112, 917-923.
Davidson, J.R., Hughes, D., Blazer, D.G., et al.(1991). Post-traumatic stress disorder in the community: an epidemiological study. Psychological Medicine, 21, 713-721.
DeMaria, A.N., Lee, G., Amsterdam, E.A., Low, R., & Mason, D.T.(1980). The anginal syndrome with normal coronary arteries: Etiologic and prognostic considerations. Journal of the American Medical Association, 244, 826-828.
Derogatis, L.R., Lipman, R.S., & Covi, L.(1973). SCL-90: An outpatient psychiatric rating scale: Preliminary report. Psychopharmacology Bulletin, 9, 13-27.
Dunner, D.(1985). Anxiety and panic: relationship to depression and cardiac disorders. Psychosomatics, 26(11 suppl), 18-22.
Fava, M., Abraham, M., Pava, J., et al.(1996). Cardiovascular risk factors in depression: the role of anxiety and anger. Psychosomatics, 37(1), 31-37.
Friedman, M.J., & Bennet, P.L.(1977). Depression and hypertension. Psychosomatic Medicine, 39, 134-142.
Fries, E.D.(1973). Age, race, sex, and other indices of risk in hypertension. American Journal of Medicine, 55, 275-280.
Gaffney, F.A, Fenton, B.J., Lane, L.D., et al.(1988). Hemodynamic, ventilatory, and biochemical responses of panic patients and normal controls with sodium lactate infusion and spontaneous panic attacks. Archives of General Psychiatry, 45, 53-60.
Glass, D.C.(1977). Behavior Patterns, Stress, and Coronary Disease. Hillsdale, NJ: Erlbaum.
Goldband, S., Katkin, E.S., & Morell, M.A.(1979). Personality and cardiovascular disorder: Steps toward demystification. In I.G. Sarason and C.D. Spielberger(Eds.), Stress and Anxiety, Vol. 6. Washington, D.C.: Hemisphere.
Haines, A.P., Imeson, J.D., & Mease, T.W.(1987). Phobic anxiety and ischaemic heart disease. Br. Med. J., 295, 297-299.
Hamilton, M.(1959). The assessment of anxiety states by rating. British Journal of Medical Psychology, 32, 50-55.
Hayward, C., Taylor, C.B., Roth, W.T., et al.(1989). Plasma lipid levels in patients with panic disorder or agoraphobia. American Journal of Psychiatry, 146, 917-919.
Hayward, C.(1995). Psychiatric Illness and Cardiovascular Disease Risk. Epidemiologic Reviews, 17(1), 129-138.
Houston, B.K.(1986). Psychological variables and cardiovascular and neuroendocrine reactivity. In K.A. Matthews, S.M. Weiss, T. Detre, T.M. Dembroski, B. Falkner, S.B. Manuck, & R. Williams(Eds.), Handbook of Research Methods in Cardiovascular Behavioral Medicine. New York, NY: Plenum.
James, S.A., & Kleinbaum, D.G.(1976). Socioecologic stress and hypertension related mortality rates in North Carolina. American Journal of Public Health, 66, 354-358.
Jenkins, C.(1976). Recent evidence supporting psychologic and social risk factors for coronary disease. New England Journal of Medicine, 294, 987-994, 1033-1038.
Jenkins, C., Somervell, P.D., & Hames, C.G.(1983). Does blood pressure usually rise with age? Or with stress? Journal of Human Stress, 9, 4-12.
Jonas, B.S., Franks, P., & Ingram, D.D.(1997). Are Symptoms of Anxiety and Depression Risk Factors for Hypertension? Archives of Family Medicine, 6, 43-51.
Kahn, H.A., Medalie, J.H., Neufeld, N.H., Riss, E., & Goldhurt, U.(1972). The incidence of hypertension and associated factors: the Israel Ischemic Heart Disease Study. American Heart Journal, 84, 171-182.
Karasek, R.A.(1979). Job demands, job decision latitude, and mental strain: Implications for job redesign. Adm. Sci. Q., 24, 285-308.
Katon, W.(1984). Panic disorder and somatization: review of 55 cases. American Journal of Medicine, 77, 101-106.
Kawachi, I., Colditz, G.A., Ascherio, A., et al.(1994). Prospective study of phobic anxiety and risk of coronary heart disease in men. Circulation, 89, 1992-1997.
Kawachi, I., Sparrow, D., Vokonas, P.S., & Weiss, S.T.(1994). Symptoms of Anxiety and Risk of Coronary Heart Disease. Circulation, 90(5), 2225-2229.
Kranz, D.S., & Manuck, S.B.(1984). Acute psychophysiologic reactivity and risk of cardiovascular disease: a review and methodologic critique. Psychological Bulletin, 96, 435-464.
Lansbergis, P.A., Schnall, P.L., Schwartz, J.E., Warren, K., & Pickering, T.G.(1992). The patterning of psychological attributes and distress by job strain and social support in a sample of working men. Journal of Behavioral Medicine, 15, 379-405.
Lebovitis, B., Lichter, E., & Moses, V.(1975). Personality correlates of coronary heart disease. Soc. Sci. Med., 9, 207-219.
Markovitz, J.H., Matthews, K.A., Wing, R.R., Kuller, L.H., & Mellahn, E.N.(1991). Psychological, biological and health behavior predictors of blood pressure changes in middle-aged women. Journal of Hypertension, 9, 399-406.
Markovitz, J.H., Matthews, K.A., Kannel, W.B., Cobb, J.L., & DUAgostino, R.B.(1989). Psychological predictors of hypertension in the Framington Study: is there tension in hypertension? JAMA, 270, 2439-2443.
McNair, D.M., Lorr, M., & Droppleman, L.F.(1971). Profile of Mood States. San Diego: Educational and Industrial Testing Service.
Medalie, J.H., Snyder, M., Groen, J.J., et al.(1973). Angina pectoris among 10,000 men: 5 year incidence and univariate analysis. American Journal of Medicine, 55, 583-94.
Noyes, R. Jr., Clancy, J., Hoenk, P.R., et al.(1980). The prognosis of anxiety neurosis. Archives of General Psychiatry, 37, 173-178.
Paffenberger, R.S., Wolf, P.A., Notkin, J et al.(1966). Chronic disease in former college students. I. Early precursors of fatal coronary heart disease. American Journal of Epidemiology, 83, 314-28.
Russek, L.G., King, S.H., Russek, S.J., & Russek, H.I.(1990). The Harvard Mastery of Stress Study 35-year follow-up: prognostic significance of patterns of psychophysiological arousal and adaption. Psychosomatic Medicine, 52, 271-285.
Schnall, P.L., Landsbergis, P., & Baker, D.(1994). Job Strain and Cardiovascular Disease. Annu. Rev. Public Health, 15, 381-411.
Schnall, P.L., Lansbergis, P., Schwartz, J.E., Warren, K., & Pickering, T.G.(1992). The relationship between job strain, ambulatory blood pressure and hypertension. Presented at Int. Symp. Epidemiol. Occup. Health, 9th., Cincinatti, OH.
Schocken, D.C., Worden, T.J., Harrison, E.E., & Spielberger, C.D.(1984). Anxiety differences in patients with angina pectoris, nonanginal chest pain, and coronary artery disease(Abstract). Circulation, 70, II-387.
Somova, L.I., Connolly, C., & Diara, K.(1995). Psychosocial predictors of hypertension in Black and White Africans. Journal of Hypertension, 13, 193-199.
Sims, A., & Prior, P.(1982). Arteriosclerosis related deaths in severe neurosis. Compr. Psychiatry, 23, 181-185.
Smith, T., & Frohm, K.(1985). What’s so unhealthy about hostility? – Construct validity and psychosocial correlates of the Cook-Medley Ho scale. Health Psychology, 4, 503-520.
Sparrow, D., Garvey, A.J., Rosner, B., & Thomas, H.E. Jr.(1982). Circulation, 65, 789-794.
Spielberger, C.D., Gorsuch, R.L., & Lushene, R.D.(1970). Manual for the State-Trait Anxiety Inventory. Palo Alto, CA: Consulting Psychologists Press.
Spielberger, C.D., & Rickman, R.L.(1990). Assessment of State and Trait Anxiety in Cardiovascular Disorders. In D.G. Byrne and R.H. Rosenman(Eds.), Anxiety and the Heart(pp. 73-92). New York: Hemisphere Publishing Corporation.
Sprafkin, R.P., McCroskery, J.H., Lantinga, L.J., & Hills, N.(1984). Cardiovascular and psychological characteristics of patients with chest pain. Behavioral Medicine, 6, 11.
Stansfeld, S.A., North, F.M., White, I., & Marmot, M.G.(1995). Work characteristics and psychiatric disorder in civil servants in London. Journal of Epidemiology and Community Health, 49, 48-53.
Tancer, M.E., Stein, M.B., Moul, D.E., et al.(1990). Normal serum cholesterol in panic disorder. Biological Psychiatry, 19, 99-101.
Taylor, C.B., King, R., Ehlers, A., et al.(1987). Treadmill exercise test and ambulatory measures in panic attacks. American Journal of Cardiology, 60, 48J-52J.
Taylor, J.A.(1953). Apersonality scale of manifest anxiety. Journal of Abnormal and Social Psychology, 48, 285-290.
Thiel, H.G., Parker, D., & Bruce, T.A.(1973). Stress factors and risk of myocardial infarction. Journal of Psychosomatic Research, 17, 43-57.
Turner, S.M., Beidel, D.C., & Larkin, K.T.(1986). Situational determinants of social anxiety in clinic and nonclinic samples: Physiological and cognitive correlates. J. Consult. Clin. Psychol., 54, 523-527.
Ward, E.E.(1990). The relationship oof Anxiety and Cardiovascular Reactivity. In D.G. Bryne, & R.H. Rosenman (Eds.), Anxiety and the Heart(pp. 347-365). New York, NY: Hemisphere Publishing Corporation.
Wardell, W.I., & Bahnson, C.B.(1973). Behavioral variables and myocardial infarction in the southeastern Connecticut heart study. Journal of Chronic Disease, 26, 447-61.
Weissman, M.M., Markowitz, J.S., Ouellette, R., et al.(1990). Panic disorder and cardiovascular/cerebrovascular problems: results from a community survey. American Journal of Psychiatry, 147, 1504-1508.
Zuckerman, M., & Lubin, B.(1965). Manual for the Multiple Affect Adjective Check List. San Diego, CA: Educational and Industrial Testing Service.
Zyzanski, S., Jenkins, C., Ryan, T., et al.(1976). Psychological correlates of coronary angiographic findings. Archives of Internal Medicine, 136, 1234-1237.
As early as 1939, Franz Alexander suggested that the repression of anger is associated with chronic elevations in blood pressure and ultimately with essential hypertension (Alexander, 1939). However, because of inconsistent findings in efforts to support this hypothesis, interest in repressed and suppressed anger as a risk factor for coronary heart disease (CHD) declined for many decades (Diamond, 1982). It was not until Rosenman and Friedman’s work on Type A Behavior Pattern (TABP) that anger reentered the spotlight as potentially being involved in the development of CHD. Rosenman and Friedman clearly included expressed anger as part of the TABP construct (Rosenman, 1978). Yet anger once again became largely ignored as coronary-prone behavior researchers shifted their attention to another dimension of TABP, the “potential for hostility.” However, when the criteria for “potential for hostility was later carefully examined, it was found that these judgments of the “potential for hostility” involve both anger and hostile attitudes (Siegman, 1994). In fact, leaders in the field of hostility-CHD research such as Theodore Dembroski, later renamed what they were measuring from “potential for hostility” to “anger-hostility.” Only very recently has the hypothesis that anger is a significant risk factor for CHD been a question that coronary-prone behavior researchers have focused on.
What is anger and how is it assessed?
Psychologists distinguish between anger, hostility, and aggression, with the first referring to affect, the second to attitudes, and the last to destructive behavior. Anger is often described as the emotion evoked when a person is blocked in the attainment of a goal or in the fulfillment of a need. Hostility, on the other hand, is a general attitude of ill will and negative evaluation of people and events (Mendes de Leon and Meesters, 1991).
As previously mentioned, coronary-prone behavior researchers have concentrated on measuring hostility, not anger, and as a result popular techniques have not done well at separating the two. Assessment of anger and hostility can be divided into two general types: self-report and interview based. Each general type has its own strengths and weaknesses, and create a need for multiple types of self-report and interview-based methods. Self-report measures include the Cook-Medley Ho scale, Factor L, and the Buss-Durkee Hostility Inventory. Interview-based assessments have been developed from the Structured Interview (SI) method developed originally to assess TABP, consisting of a set of questions concerning competitiveness, time urgency, and anger expression.
Assessing Modes of Anger Coping: Expression, Suppression, and Repression of Anger
Another interest of coronary-prone behavior researchers has been to determine which modes of coping with angry feelings are most detrimental to cardiovascular health. Anger-in and anger-out constructs have been created to help assess anger coping styles. Those that tend to hold anger inside and not display it outwardly are “anger-in,” while those that vent their anger readily are termed “anger-out.” (Spielberger et al., 1985) Assessing unexpressed anger (anger-in) has proved difficult without first having valid methods to determine whether anger is being experienced but not expressed, or whether anger is not being experienced. Assessments of whether anger is being experienced have been developed by Siegman and colleagues using factor analyses of the Buss-Durkee Hostility Inventory subscales (Siegman et al., 1987). Two factors were identified. The first was defined by subscales that measure the frequency with which feelings of anger and hostility are experienced, and the second factor was defined by subscales that measure the expression of anger that occurs in response to provocation (Siegman, 1994). If anger is being experienced, failure to express anger through overt aggression (anger-out) could reflect autonomic or voluntary inhibition of aggressive urges (i.e., repression or suppression of anger), or it could reflect alternative methods of coping through other social responses, such as assertiveness (Smith, 1994).
Cardiovascular Consequences of Repressing or Suppressing Anger vs. Expressing Anger
It has long been known that anger is associated with heightened levels of cardiovascular arousal, which has been identified as the mechanism that translates behavior into coronary heart disease. Anger is also associated with increased rates of testosterone production and platelet formation, which are also involved in the pathogenesis of CHD (Siegman, 1994). Other biological mechanisms by which anger may increase the risk of CHD include discharge of circulating catecholamines, increased myocardial oxygen demand, and vasospasm (Kawachi et al., 1996). Whatever the biological mechanisms are, if anger is indeed a crucial risk factor for CHD, it is important to understand which forms of coping with anger are most related to the development of CHD.
Heightened levels of systolic and diastolic blood pressure (BP) reactivity have been identified as a risk factor for future development of hypertension, CAD, and CHD (Manuck et al., 1992; Williams, 1989; Fredrickson & Matthews, 1990; Light et al., 1992). Experiments have been done to test the hypothesis that the expression of anger is associated with heightened levels of cardiovascular reactivity (CVR)(Siegman et al., 1990; Siegman & Boyle, 1992). The findings by Siegman and colleagues indicated that only the expression of anger, not its mere experience, was associated with appreciable increases in BP reactivity.
Correlational studies examining the relationship between the experience of anger, the expression of anger, and cardiovascular reactivity (CVR) have also been done. The distinction between the experience of anger versus the expression of anger is based upon the results of several factor analytic studies of the Buss-Durkee Hostility Inventory (BDHI). In three studies, significant positive correlation has been obtained only between BDHI-derived expression of anger and BP reactivity during an anger-arousing task (Siegman et al., 1988; Siegman et al., 1992; Suarez & Williams, 1990). However, findings showing insignificant correlation between BDHI-derived experience of anger and BP reactivity does not mean that suppression and repression of anger do not have any association with heightened BP reactivity or higher resting BP levels. As previously mentioned, people cope with anger in a variety of different ways, and therefore any attempts to correlate the experience of anger to BP reactivity says little about the effects of specific modes of coping with anger on BP reactivity. In fact, there is evidence suggesting that people who have a tendency to suppress their anger also have a tendency to have high blood pressure (Harburg et al., 1973; Harburg et al., 1979; Gentry et al., 1982). The suppression of anger has been considered an important component of the “hypertensive personality” by many investigators (Pernini et al., 1988; Schneider et al., 1986). Several of the previous prospective studies in this area have found suppressed anger to be predictive of increases in blood pressure (Kahn et al., 1972; Rose et al., 1978; McCelland, 1979; Pernini et al., 1991), and a number of cross-sectional studies have found increased anger suppression in hypertensive individuals (Schneider et al., 1986; Pernini et al., 1990; Goldstein et al., 1988). It should be mentioned that one of the most recent prospective studies (Markovitz et al., 1993) found no relationship between mode of anger expression and later hypertension, but a large number of studies still support the view that suppression of anger is an important risk factor for hypertension.
Numerous studies have also been completed analyzing the relationship between the expression of anger, the experience of anger, and CAD or CHD (Siegman et al., 1987; Helmig et al., 1991; Mendes de Leon, 1992; Dembroski et al., 1989). Results have been consistent with those where BP reactivity was the endpoint, with only the expression of anger, not its mere experience, being related to the development of CAD and CHD.
Evidence in regard to the possible relationship between the suppression or repression of anger, and CAD or CHD is fairly sparse. Positive findings include Haynes et al.’s 1980 study that found that suppression of anger in both men and women was related to future development of CAD. Additionally, two 1985 studies (Demroski et al., 1985; MacDougall et al., 1985) evaluated both anger suppression and hostility and demonstrated that those who were hostile and suppressed their anger had the most severe CAD.
Validity of these studies on anger suppression and CHD are still quite questionable, because satisfactory measures of suppression and repression have not yet been developed (Siegman, 1994). Many studies have used the “anger-in” construct to assess suppression and repression of anger, however none of the assessments of anger-in (Haynes et al., 1978; Spielberger et al., 1985; Dembroski et al., 1983, 1989) distinguish between suppressing anger and repressing anger. Also, there has been some confusion concerning the assessments of anger-in. Some anger-in measures appear to assess other characteristics such as shyness and unassertiveness, or cognitive factors such as brooding and resentment (Miller et al., 1996). Results from studies examining the relationship between anger-in and CVR mostly report null findings, as reviewed by Mills et al., 1989. One of the most recent, major prospective studies found that neither anger-in nor anger-out was associated with a higher incidence of CHD (Russek et al., 1990). Other than the problem of distinguishing what each anger-in measure really assesses, these negative findings may be partly due to the fact that in these studies some of the participants may not have been angered sufficiently, and therefore have little or no anger to suppress or repress (Siegman, 1994).
Work environments and anger-hostility
It is becoming increasingly clear that environmental factors have a large impact on how psychosocial factors, such as anger, effect CHD development. Research primarily by Burns and colleagues suggests that psychosocial traits of anger expression and hostility may be differentially related to CVR depending on the qualities of the situation and gender (Burns et al., 1993). Also according to Burns, one group of men at risk for CHD may be those who are chronically angry, who prefer to suppress their anger, and who are also faced with constant and excessive interpersonal harassment, demands, or criticism on the job or at home (Burns et al., 1995).
“Job strain” has been defined by Karasek (1979) as work in jobs with high psychological demands and low control. In a half-dozen epidemiological studies over the last decade, occupational stress researchers have implicated “job strain” as a risk factor for heart disease (Schnall et al., 1994). Although the mechanism by which the stress of “job strain” influences development of CHD is unknown, previous findings suggest that “job strain” may be related to elevations of blood pressure at work (Schnall et al., 1990; Van Egeren, 1992). In the Cornell prospective study of job strain and hypertension, anger was associated with lower job control (decision latitude) and lower workplace social support among men. Lower SES and non-White adults experience lower levels of control at work and in other aspects of their lives. Another recent study by Burns (Burns, Hutt, & Wiedner, 1993) examined whether anger-hostility could moderate the effects of demand and decision latitude on CVR. The results suggest that anger-hostility may operate in concert with dimensions of job stress to affect CVR, and presumably CHD. It remains to be seen how anger suppression and anger expression operate in relation to dimensions of job stress, and hopefully this weakness in the literature will soon be remedied.
Suppressed anger and its determinants
There has been remarkably little research on the determinants on suppressed anger, particularly social class and job stress. Such factors might shape personality development in childhood. For example, certain parental behavior patterns (i.e., overly strict, critical and demanding of conformity) are more common in low SES households, and may be viewed as a reflection of the parents’ occupational and other life experiences, which are characterized by low control and insecurity. Similarly, an adult’s experience, which might include stressful, low control jobs, may shape their personality development (Kohn and Schooler, 1982). Psychosocial work environments which enforce a non-complaining attitude and prevent development of active emotional coping with anger may result in suppressed anger that could have an adverse effect on long-term cardiovascular health of the employees (Theorell, 1990; Cottington et al., 1986). Thus, research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
There is an abundance of conflicting results in the literature regarding the cardiovascular consequences of different anger coping styles. Despite the controversy over which anger coping style is most detrimental to the heart, there is a mass of evidence suggesting that anger coping styles, not the mere experience of anger, are related to CVR and possible development of CHD. However, recent research (Burns et al., 1995) suggests that CVR depends not only on how anger is dealt with, but also on how much anger must be managed, at least among anger suppressors. To gain further insight into whether anger suppression versus anger expression is more important in the etiology of CHD, more valid measures of anger suppression must first be created. Additionally, environmental factors must be better accounted for, such that attempts to establish associations between anger suppression or anger expression and CHD are elicited during relevant stressful situations. Although job stress has already been identified as a risk factor for CHD, more research is needed to determine how specific anger coping styles and “job strain” are related to CHD, and whether they are independent relationships or mutually reinforcing. Lastly, more research is needed to better elucidate the determinants of depression, particularly social class and job stress.
Alexander, F.G.(1939). Emotional factors in essential hypertension. Psychosomatic Medicine, 1, 175-179.
Burns, J.W.(1995). Interactive Effects of Traits, States, and Gender on Cardiovascular Reactivity During Different Situations. Journal of Behavioral Medicine, 18(3), 279-303.
Burns, J.W., & Katkin, E.S.(1993). Psychological, Situational, and Gender Predictors of Cardiovascular Reactivity to Stress: A Multivariate Approach. Journal of Behavioral Medicine, 16(5), 445-465.
Burns, J.W., Hutt, J., & Weidner, G.(1993). Effects of Demand and Decisions Latitude on Cardiovascular Reactivity Among Coronary-Prone Women and Men. Behavioral Medicine, 19(3), 122-128.
Cottington, E., Matthews, K., Talbott, E., & Kuller, L.(1986). Occupational stress, suppresssed anger, and hypertension. Psychosomatic Medicine, 48, 249-259.
Dembroski, T.M., & MacDougall, J.M.(1983). Behavioral and psychophysiological perspectives on coronary-prone behavior. In T.M. Dembroski, T.H. Schmidt, & G. Blumchen(Eds.), Biobehavioral bases of coronary heart disease. New York: Karger.
Dembroski, T.M., MacDougall, J.M., Williams, R.B., Haney, T.L., & Blumenthal, J.A.(1985). Components of Type-A, hostility, and anger-in: Relationship to angiographic findings. Psychosomatic Medicine, 47, 219-233.
Dembroski, T.M., MacDougall, J.M., Costa, P.T., & Grandits, G.A.(1989). Components of hostility as predictors of sudden death and myocardial infarction in the Multiple Risk Factor Intervention Trial. Psychosomatic Medicine, 51, 514-522.
Diamond, E.L.(1982) The role of anger and hostility in essential hypertension and coronary heart disease. Psychological Bulletin, 92, 410-433.
Gentry, W.D., Chesney, A.P., Gary, H.E., et al.(1982). Habitual anger-coping styles. I. Effect on mean blood pressure and risk for essential hypertension. Psychosomatic Medicine, 44, 195-202.
Goldstein, H.S., Edelbery, R., Meier, C.F., & Davis, L.(1988). Relationship of resting blood pressure and heart rate to experienced anger and expressed anger. Psychosomatic Medicine, 50, 321-329.
Harburg, E., Erfurt, J.C., Hauenstein, L.S., et al.(1973). Socio-ecological stress, suppressed hostility, skin color, and black-white male blood pressure. Psychosomatic Medicine, 35, 276-96.
Harburg, E., Blakelock, E.H., & Roeper, P.J.(1979). Resentful and reflective coping with arbitrary authority and blood pressure. Psychosomatic Medicine, 41, 189-202.
Haynes, S.G., Feinleib, M., & Kannel, W.B.(1980). The relationship of psychosocial factors to coronary heart disease in the Framington study: III. Eight-year incidence of coronary heart disease. American Journal of Epidemiology, 111, 37-58.
Haynes, S.G., Levine, S., Scotch, N., Feinleib, M., & Kannel, W.B.(1978). The relationship of psychosocial factors to coronary heart disease in the Framington Study: I. Methods and risk factors. American Journal of Epidemiology, 107, 362-383.
Helmig, L., Houston, B.K., Vavak, C.R., & Mullin, J.(March 1991). Hostility rekated variables, self-schemata and CHD. Paper presented at the meetings of the Society for Behavioral Medicine, Washington, DC.
Kabat-Zinn, J.(1992). In I.S. Ockene, & J.K. Ockene(Eds.), Prevention of Coronary Heart Disease (pp.299-332). Boston, MA: Little, Brown and Company.
Kahn, H.A., Medaalie, J.H., Neufeld, H.N., Riss, E., Aldbourt, U.(1972). The incidence of hypertension and associated factors: the Israeli Ischemic Heart Disease Study. Am. Heart J., 84, 171-182.
Karasek, R.A.(1979). Job demands, job decision latitude, and mental stress: Implications for job redesign. Adm. Sci. Q., 24, 285-308.
Kawachi, I., Sparrow, D., Spiro, A., Vokonas, P., & Weiss, S.T.(1996). A Prospective Study of Anger and Coronary Heart Disease. Circulation, 94(9), 2090-2095.
MacDougall, J.M., Dembroski, T.M., Dimsdale, J.E., & Hackett, T.P.(1985). Components of Type-A, hostility, and anger-in: Further relationships to angiographic findings. Health Psychology, 4, 137-152.
Manuck, S.B., Olsen, G., Hjemdahl, P., & Renqvist, N.(1992). Does cardiovascular reactivity to mental stress have prognostic value in postinfarcation patients? A pilot study. Psychosomatic Medicine, 54, 102-108.
Markovitz, J.H., Matthews, K.A., Kannel, W.B., Cobb, J.L., & DUAgositino, R.B.(1993). Psychological Predictors of Hypertension in the Framingham Study. JAMA, 270(20), 2439-2443.
McCelland, D.C.(1979). Inhibited power motivation and high blood pressure in men. J. Abnorm. Psychol., 88, 182-190.
Mendes de Leon, C.F.(1992). Anger and impatience/irritability in patients of low socioeconomic status with acute coronary heart disease. Journal of Behavioral Medicine, 15, 273-284.
Mendes de Leon, C.F., & Meesters, M.G.(1991). In A. Appels, J. Groen, J. Koolhaas, J. van Dixhoorn, L. van Doornen, C. Mendes de Leon, & C. Meesters (Eds.), Behavioral Observations in Cardiovascular Research (pp. 107-128). Amsterdam: Swets & Zeitlinger B.V.
Miller, T.Q., Smith, T.W, Turner, C.W., Guijarro, M.L., & Hallet, A.J. (1996). A meta-analytic review of research on hostility and physical health. Psychological Bulletin, 119(2), 322-348.
Pernini, C., Mailer, F.B., Rauchfleisch, U, Battegay, R., Hobi, V., & Buhler, F.R.(1990). Psychosomatic factors in borderline hypertensive subjects and offspring of hypertensive parents. Hypertension, 16, 627-634.
Pernini, C., Muller, F.B., & Buhler, F.R.(1988). Suppressed aggression and hyperdynamic cardiovascular regulation in normotensive offspring of essential hypertensive patients. J. Cardiovasc. Pharmacol., 12 (suppl 3): S130-S133.
Pernini, C., Muller, F.B., & Buhler, F.R.(1991). Suppressed aggression accelerates early development of essential hypertension. J. Hypertension, 9, 499-503.
Rose, R.M., Jenkins, C.D., & Hurst, M.W.(1978). Air Traffic Controller Health Change Study: A Prospective Investigation of Physical, Psychological, and Work-related Changes. Boston, Mass: Boston University School of Medicine. A report to the Federal Aviation Administration on research performed under contract DCT-FA73WA-3211.
Rosenman, R.H.(1978). The interview method of assessment of the coronary-prone behavior pattern. In T.M. Dembroski, S. Weiss, J. Schillar, S.G. Haynes, & M. Feinlieb (Eds.), Coronary prone behavior. New York: Springer-Verlag.
Russek, L.G., King, S.H., Russek, S.J., & Russek, H.I.(1990). The Harvard Mastery of Stress Study 35-Year Follow-up: Prognostic Significance of Patterns of Psychophysiological Arousal and Adaptation. Psychosomatic Medicine, 52, 271-285.
Schnall, P.L., Landsbergis, P., & Baker, D.(1994). Job Strain and Cardiovascular Disease. Annu. Rev. Public Health, 15, 381-411.
Schnall, P.L., Pieper, C., Schwartz, J.E., Karasek, R.A., Schlussel, Y, et al.(1990). The relationship between Rjob strain,S workplace diastolic blood pressure, and left ventricular mass index: Results of a case-control study. J. Am. Med. Assoc, 267, 1209.
Schneider, R.H., Egan, B.M., Johnson, E.H., Drobny, H., & Julius, S.(1986). Anger and anxiety in borderline hypertension. Psychosomatic Medicine, 48, 242-248.
Siegman, A.W.(1994). In A.W. Siegman, & T.W. Smith (Eds.), Anger, Hostility, and the Heart (pp. 1-22, 173-198). Hillsdale, NJ: Lawrence Erlbaum Associates.
Siegman, A.W., Anderson, R.A., Herbst, J., Boyle, S., & Wilkinson, J.(1992). Dimensions of anger-hostility and cardiovascular reactivity in provoked and angered men. Journal of Behavioral Medicine, 15, 257-272.
Siegman, A.W., & Boyle, S.(1992, March 31-April 4). The expression of anger and cardiovascular reactivity in men and women: An experimental investigation. Paper presented at 50th anniversary international meeting of the American Psychosomatic Society, New York.
Siegman, A.W., Dembroski, T.M., & Ringel, N.(1987). Components of hostility and the severity of coronary artery disease. Psychosomatic Medicine, 49, 127-135.
Siegman, A.W., Dohm, F.A., & Gjesdal, J.(April 1988). The effects of repressive coping style on the relationship between hostility, neuroticism, and cardiovascular reactivity. Symposium paper presented at the annual meetings of the Society for Behavioral Medicine, Boston.
Smith, T.W.(1994). In A.W. Siegman, & T.W. Smith (Eds.), Anger, Hostility, and the Heart (pp.23-42). Hillsdale, NJ: Lawrence Erlbaum Associates.
Spielberger, C.D., Johnbson, E.H., Russel, S.F., Crane, R.J., Jacobs, G.A., & Worden, T.J.(1985). The experience and expression of anger: Construction and validation of an anger expression scale. In M. Cheney & R. Rosenman (Eds.), Anger and hostility in cardiovascular and behavioral disorders. New York: McGraw-Hill.
Suarez, E.C., & Williams, R.B.(1990). The relationships between dimensions of hostility and cardiovascular reactivity as a function of task characteristics. Psychosomatic Medicine, 52, 558-570.
Theorell, T.(1990). Family history of hypertension- an individual trait interacting with spontaneoulsy occurring job stressors. Scand. J. Work Environ. Health, 16 (suppl. 1), 74-79.
Van Egeren, L.F.(1992). The relationship between job strain and blood pressure at work, at home, and during sleep. Psychosomatic Medicine, 54, 337-343.
Williams, R.B.(1989). Biological mechanisms mediating the relationship between behavior and coronary heart disease. In A.W. Siegman & T.M. Dembroski (Eds.), In search of coronary-prone behavior: Beyond Type A. Hillsdale: NJ: Lawrence Erlbaum Associates.
There are few epidemiological studies directly investigating whether need for control is a risk factor for heart disease or CHD risk factors. Researchers in the 1970s and 1980s had suggested this association:
“The notion that a challenged need for control or loss of control is involved in the etiology of coronary heart disease has been present in the field of cardiovascular research for many years. However, no epidemiological study has directly addressed the question of whether or not those who are in high need for control as a personality characteristic, or those who lose control over their environment, are at increased risk for myocardial infarction. Generally speaking, need for control or loss of control are only mentioned in the discussion sections of empirical papers. Rosenman, for example, makes use of the control concept in an attempt to achieve greater insight into the origins of type A behavior.” (Appels, 1989). “The Type A behavior pattern (TABP) may be a characteristic style of response to environmental stressors that threaten an individual’s sense of control over his or her environment. Thus, Type A behavior appears to be an enhanced performance to assert and maintain control over the environment whenever this control is challenged or threatened. The high drive and pace of Type A persons reflect their need for mastery over their environment.” (Rosenman, 1986). David Glass should be credited as being the first author to draw attention to loss of control as a possible determinant of myocardial infarction and sudden cardiac death. His basic assertion is that Type A individuals exert greater efforts than their Type B counterparts to control stressful events that are perceived as threats to their sense of control. These active coping attempts eventually extinguish, for without reward, the relentless striving of the Type A individual leads to frustration and psychic exhaustion, which culminates in a reduction of efforts at control (Glass, 1977). This description illustrates the notion that type A individuals who suffer from coronary heart disease(CHD) have passed through a state of frustration and exhaustion prior to myocardial infarction. Glass described this state as a “prodromal depression.” He suggested that this state is characterized by a change from initial hyper-responsiveness into subsequent hypo-responsiveness.” (Appels, 1989). “The precise role of initial hyper-responsiveness and subsequent hypo-responsiveness (helplessness) in the development of cardiovascular pathology remains unclear.”(Glass, 1977).
Appels (1989) has “hypothesized that a successful type A individual is not at increased risk for myocardial infarction but that the type A person who has lost control over his or her environment is at elevated risk. An additional hypothesis was that type A individuals are at increased risk of losing control because of their deep commitment to their vocation (both in relation to occupation and family life) makes them rather vulnerable when negative events occur in their immediate environment.” (Appels, 1989)
However, this earlier work failed to adequately describe the nature of the environment which promotes or interacts with an unhealthy level of “need for control” in producing CHD risk. In other words, why does exertion of control sometimes lead success and rewards and sometimes lead to frustration, overcommitment and failure?
Recent research by Johannes Siegrist and colleagues has attempted to describe the possible harmful interactions between the personality characteristic “need for control” and environmental conditions which allow little control. The “effort-reward” model of work stress, developed by Siegrist and colleagues, defines threatening job conditions as a “mismatch between high workload (high demand) and low control over long-term rewards” (Siegrist, Peter et al., 1990, p. 1128). Effort is defined as either the demands of the job (“extrinsic effort”) or the personality characteristic of “immersion” in the job (“intrinsic effort”). Low reward includes concepts such as low “esteem reward” (e.g., low social support), low income, and low “status control” (e.g., poor promotion prospects, job insecurity). In this model, high effort and low reward can each have significant main effects on risk of cardiovascular disease (CVD), and they can also interact to further increase CVD risk.
Siegrist emphasizes personal control over long-term reward since “distressing experiences often result from basic threats to the continuity of a crucial social role”, among adults, often the occupational role. This is clearly the case “with job termination or job instability. However, related conditions of low reward and low security may also be identified, such as forced occupational change, downward mobility, lack of promotion prospects, jobs held with inconsistent educational background (status inconsistency)” (Siegrist, Peter et al., 1990, p. 1128). These aspects of work life may threaten a person’s “sense of mastery, efficacy and esteem by evoking strong recurrent negative emotions of fear, anger or irritation” (Siegrist, 1996, p. 30).
Factor analysis of items measuring the psychological component of the model (the coping pattern of “need for control”) identified two relevant coping variables — “vigor”, a state of active efforts with a high probability of positive feedback, and “immersion” a state of exhaustive coping reflecting frustrated, but continued efforts and associated negative feelings. “Immersion” is considered to increase an individual’s vulnerability to experiences of high extrinsic demand and low status control.
In a prospective study of 416 German factory workers aged 25-55, status inconsistency (odds ratio (OR)=4.4), job insecurity (OR=3.4), work pressure (OR=3.4), and immersion (OR=4.5) independently predict CHD incidence after adjusting for other behavioral and somatic risk factors (Siegrist, Peter et al., 1990). The effort and reward variables have also been associated with CHD or CHD outcomes in other studies. For example, in a cross-sectional study of Stockholm area residents, effort-reward ratio >1 was associated with hypertension (OR=1.62) and cholesterol/HDL ratio (OR=1.26) among men. Among women, the association with hypertension (OR=1.56) did not reach statistical significance (Peter, Alfreddson et al., 1997). Immersion was associated with LDL cholesterol (OR=1.39) but only among women.
In developing the effort-reward model, Siegrist and colleagues have made a major contribution to our understanding of the health effects of work stress. First, they have expanded the concept of control typically used in research on the job demands-control model to include job security, respect, reward and upward mobility (promotion prospects). They emphasize control over long-term rewards, primarily the continuity of a crucial adult social role (i.e., occupation). Second, they have shown associations between their exposure measures and a wide variety of outcome measures, including coronary heart disease, coronary risk factors, cardiovascular reactivity, musculoskeletal symptoms, gastrointestinal symptoms, fatigue and sickness absence. In addition, in some studies, they were able to employ objective measures of stressors, for example, job insecurity (employment in a factory undergoing layoffs), wage level, piecework, shiftwork, and noise (Siegrist, Peter et al., 1990). However, the model and its use raise several important theoretical and methodological questions, which need to be addressed.
Does an unhealthy level of “need for control” result from socialization or
“Need for control” (and its components “vigor” and “immersion”) is considered to be “a personal characteristic which is rather stable over time” (Siegrist, Peter et al., 1990, p. 1128) in Siegrist’s model. However, it remains to be determined to what extent they are influenced by levels of work control or other job characteristics over the course of a person’s work history. For example, in a group of Swedish men aged 45-64, “immersion” was more prevalent in men with job strain (43%) than those without job strain (30%) (Peter, 1997). In the Whitehall II study, personality measures related to “immersion” (hostility, psychiatric disorder, negative affectivity, angry coping and unassertive coping) were more common among men and women with lower job control compared to those with higher job control (Bosma, Stansfeld & Marmot, 1997).
In articles about the effort-reward model, there is little discussion of this possible socialization effect. In one, Peter and Siegrist (1997, p. 1113) do offer a hypotheses about occupational factors which may promote active as opposed to passive coping (withdrawal): “the possible costs produced by disengagement (e.g., the risk of being laid off or of facing downward mobility) may outweigh the costs of accepting inadequate benefits.” In another, Siegrist (1996, p. 31) points out that “blue-collar workers with reduced opportunities of changing jobs will not minimize their effort at work even if their gain is low” because of these “costs of disengagement”. However, the “need for control” variable is typically discussed as a “stable” person characteristic (Siegrist, Peter et al., 1990; Siegrist, 1996).
Implications for prevention and treatment
Therefore, the effort-reward model relies, in part, upon a clinical psychological (individual) explanation of a phenomenon which may have a strong sociological component. This issue is not only theoretical, but also very practical — it has important implications for intervention. What will our prescription be for “immersion”, “need for control” or, for other psychological states or traits that may be predictive of CHD, such as “hostility”? The psychological approach would prescribe counseling or psychotherapy, valuable components of any treatment for stress-related illness. A sociological approach would allow us to consider preventive public health measures such as workplace redesign or legislative reforms, in addition to individual approaches, to solve the problem of unhealthy “immersion”.
Occupational socialization — influences on “need for control”
Several studies provide evidence for the process of adult socialization. For example, in a U.S. study, the substantive complexity of work (analogous to decision latitude) predicted increased intellectual flexibility, non-authoritarianism, and intellectually demanding leisure time 10 years later (Kohn & Schooler, 1982). In Sweden, workers whose jobs became more “passive” (low demand-low latitude) over six years reported less participation in political and leisure activities. In contrast, workers in jobs, which became more “active”, participated more in these activities (Karasek & Theorell, 1990, p. 53). Statistical control for current measures of job characteristics may help to solve the question of the extent of influence of work history on personality, since job characteristics often change over time (Schnall et al., 1998; Johnson et al., 1996).
A broader sociological approach to this issue would also encompass influences on personality development that precede working life. For example, parental child-rearing practices, values, and self-esteem, which are shaped by their social class and work experiences, can play an important role in shaping their children’s psychological and personality development (Sennett & Cobb, 1973; Rubin, 1975). Thus, their children’s level of “need for control”, “vigor”, “immersion”, “need for approval”, “competitiveness”, “impatience” and “hostility”, which are created, to some extent, before their children’s entry in the workforce, are not random but the result of generations of class and work (as well as race and gender) experiences. A full sociological model of work stress would need to consider such factors. Thus, research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
Due to the lack of epidemiological studies directly assessing the issue of whether need for control is a risk factor for CHD, few conclusions can be drawn. Need for control has been suggested to be an important component of TABP (Please see our review of the evidence of TABP as a CHD risk factor). Type A individuals may have a higher need for control. Future research is needed to begin to determine whether TABP or unhealthy levels of “need for control” in combination with limited control over the environment leads to increased CHD risk. Lastly, more research is needed on the determinants of the need for control, such as social class and job stress.
Appels, A. (1989). Loss of control, vital exhaustion and coronary heart disease. In A. Steptoe and A. Appels (Eds.), Stress, Personal Control and Health (pp. 215-235). New York, NY: John Wiley & Sons.
Bosma H, Stansfeld SA, Marmot M. Personal characteristics and the association between work and heart disease. In: Socio-economic variations in cardiovascular disease in Europe: The impact of the work environment and lifestyle (The Heart at Work Network) (pp. 33-38). London: University College London, Department of Epidemiology and Public Health, 1997.
Glass, D.C.(1977). Behavior Patterns, Stress and Coronary Disease. Hillsdale: Lawrence Erlbaum
Johnson JV, Stewart W, Hall EM, Fredlund P, Theorell T. Long-term psychosocial work environment and cardiovascular mortality among Swedish men. Am J Public Health 1996;86:324-331.
Karasek RA, Theorell T. Healthy Work. New York: Basic Books, 1990.
Kohn ML, Schooler C. Job conditions and personality: A longitudinal assessment of their reciprocal effects. Am J Sociol 1982;87:1257-86.
Peter R, Siegrist J. Chronic work stress, sickness absence, and hypertension in middle managers: General or specific sociological explanation? Soc Sci Med 1997;45(7):1111-20.
Peter R. Comparative analysis of the effort-reward imbalance model and the job strain model: preliminary results from a Swedish case-control study. In: Socio-economic variations in cardiovascular disease in Europe: The impact of the work environment and lifestyle (The Heart at Work Network) (pp. 102-4). London: University College London, Department of Epidemiology and Public Health, 1997.
Peter R, Alfredsson L, Hammar N, Siegrist J, Theorell T, Westerholm P. High effort, low reward and cardiovascular risk factors in employed Swedish men and women – baseline results from the WOLF study. In: Socio-economic variations in cardiovascular disease in Europe: The impact of the work environment and lifestyle (The Heart at Work Network) (pp. 108-14). London: University College London, Department of Epidemiology and Public Health, 1997.
Rosenman, R.H.(1986). Current and past history of type A behavior pattern. In T. Schmidt, T. Dembroski, and G. Blumchen (Eds.), Biological and Psychological Factors in Cardiovascular Disease(p. 23). Berlin: Springer Verlag.
Rubin LB. Worlds of pain. New York: Basic Books, 1976.
Schnall PL, Schwartz JE, Landsbergis PA, Warren K, Pickering TG. The effect of job strain on the 3-year change in mean ambulatory blood pressure of male employees Psychosomatic Medicine (in press).
Sennett R, Cobb J. The hidden injuries of class. New York: Vintage Books, 1973.
Siegrist J, Peter R, Junge A, Cremer P, Seidel D. Low status control, high effort at work and ischemic heart disease: Prospective evidence from blue-collar men. Soc Sci Med 1990;31:1127-34.
Siegrist J. Adverse health effects of high-effort/low-reward conditions. Journal of Occupational Health Psychology 1996;1:27-41.Appels, A.(1989). Loss of control, vital exhaustion and coronary heart disease. In A. Steptoe and A. Appels(Eds.), Stress, Personal Control and Health (pp. 215-235). New York, NY: John Wiley & Sons.
Neuroticism, or negative affectivity, is one of a small set of global traits that reflect one’s general approach to life and summarize the tendencies of individuals (Denollet, 1993). The personality dimension of neuroticism reflects the tendency to experience emotional distress and the inability to cope effectively with stress. Highly neurotic people are extremely tense, anxious, insecure, suspecting, jealous, emotionally unstable, hostile and vulnerable (Maddi, 1980). “Although emotional distress is associated with invalid health complaints such as chest pain in the absence of coronary heart disease (CHD)(Costa and McCrae, 1987), evidence suggests that emotional distress is associated with actual CHD as well.” (Friedman, 1990)(Denollet, 1993).
How is neuroticism assessed?
There is no standard assessment of neuroticism. Previously used measures of neuroticism include the total Cornell Medical Index ( CMI) Psychiatric score (Brodman, et al., 1960), the Guilford Zimmerman Temperament Survey (GZTS) Emotional Stability (Guilford et al., 1976), GZTS Emotional Health (Guilford et al., 1976), the NEO Personality Inventory (NEO-PI; Costa & McCrae, 1985), the Minnesota Multiphasic Personality Inventory (MMPI) Neuroticism Component Scale (Costa et al., 1985), the Eysenck Personality Questionnaire (EPQ; Eysenck & Eysenck, 1975), and the Eysenck Personality Inventory (EPI; Eysenck & Eysenck, 1964). Of these, the EPI, EPQ, and the MMPI Neuroticism Component Scale have become the most commonly used measures in recent years for investigating the relationship between neuroticism and cardiovascular disease.
Is Neuroticism a Psychosocial Risk Factor for CHD?
There is little evidence suggesting that neuroticism is predictive of CHD. There have been studies that have shown neuroticism and its component traits to be positively associated with occurrence of CHD, however the interpretation of these results has been complicated by several problems (Jenkins, 1971; Jenkins, 1976; Friedman, 1987). “First, neuroticism is positively associated with occurrence of a condition clinically similar to angina pectoris but not due to atherosclerotic coronary artery disease(Costa, 1987). This produces errors of classification which, because they are correlated with neuroticism, bias results toward a positive association. Second, neuroticism is positively correlated with worry about health. Persons who score high on neuroticism are more likely than others to visit physicians and, therefore, to have asymptomatic diseases discovered. This could also bias results toward a positive association because a large proportion of nonfatal myocardial infarctions are asymptomatic.” (Almada et al., 1991)
Most importantly, studies finding positive associations between neuroticism and coronary heart disease endpoints have been cross-sectional. Neuroticism has not been shown to prospectively predict CHD outcomes such as myocardial infarction(MI) and CHD mortality (Costa, 1987). Four large-scale prospective studies that examined the association between neuroticism and subsequent myocardial infarction all yielded null findings(Ostfeld et al., 1964; Keehn et al., 1974; Goldbourt et al., 1975;Hallstrom et al., 1975). Because the temporal association of the variables in cross-sectional studies remain ambiguous, many prominent coronary-prone researchers (Stone and Costa, 1990; Matthews, 1988) believe that only prospective studies should be used to garner evidence about the causal role of personality in disease etiology (Costa, 1989). Although neuroticism has not been shown to predict CHD outcomes such as MI and CHD mortality, more research is needed before neuroticism should be ruled out as playing a causal role in CHD development. Validity of these studies are still quite questionable, because the best assessments currently available for neuroticism were not used in these studies, increasing the chance of interference due to confounders.
Although there is a lack of evidence suggesting that neuroticism plays a causal role in the development of heart disease, there is evidence that neuroticism is somehow associated to coronary heart disease. In a 1991 study by Cramer, neuroticism assessed using the EPI was found to be more positively correlated to self-reported coronary heart disease than Type A Behavior Pattern (TABP), a much more established psychosocial risk factor for CHD. Similar work by Lichtenstein et al., 1989 provides further support for the positive association of neuroticism with self-reported CHD. Neuroticism assessed using the EPI again was correlated with greater relative risk for CHD. Men under the age of 65 years had a relative risk for CHD of 2.56 (1.47-4.45, 95% confidence interval), while women under the age of 65 years had a relative risk of 2.73 (1.41-5.28, 95% confidence interval). Winstow et al., 1989 summarizes much of the research which has shown relationships between cardiovascular disease and neuroticism. According to Winstow et al., ten out of twelve studies found a positive relationship between neuroticism and various cardiovascular disease endpoints, such as MI and angina. This leads them to the conclusion that “in general there appears to be strong research support for a positive relationship between neuroticism and cardiovascular disease.” Winstow et al., 1989 also contained the results of their own study which found that neuroticism, assessed using the EPQ, was positively associated with stress and cardiovascular symptoms. Once again, due to the temporal ambiguity of these studies from their cross-sectional nature, it is impossible to determine which one, if any, of these factors play a causal role. However, these findings supporting that stress, neuroticism, and cardiovascular symptoms may be tied together in ways that are not yet understood do show that more research is needed concerning the interaction of these factors.
Despite these findings of a positive association between neuroticism and various CHD endpoints, null findings and questions of validity over many of the studies create doubt over whether any important relationship between neuroticism and CHD exists at all. Although neuroticism has been very consistently associated with chest pain complaints and frequently associated
with subsequent diagnoses of angina pectoris, neuroticism has consistently been found not to be positively associated with stricter, more objective CHD endpoints (Stone and Costa, 1990). For example, two angiographic studies found no relationship between neuroticism and the extent of coronary artery disease (CAD)(Blumenthal et al., 1979; Zyzanski et al., 1976). Other angiographic studies actually found an inverse relationship between neuroticism and extent of CAD (Elias et al., 1982; Bass & Wade, 1984). Furthermore, these results were confirmed in studies that use other objective criteria for CHD other than angiography, such as history of MI and electrocardiogram (ECG) evidence of coronary ischemia (Costa et al., 1982). The use of strict, objective endpoints is particularly important in studies examining the relationship between neuroticism and cardiovascular disease, because of the previously discussed biases toward a positive association of neuroticism with more subjective CHD endpoints, such as self-reported cardiovascular health problems and angina pectoris diagnoses. Until neuroticism is shown to be consistently positively associated with strict, objective CHD endpoints, the neuroticism-CHD association will remain highly controversial.
Associations between Neuroticism and High Blood Pressure
Studies have also examined whether neuroticism is associated with high blood pressure. As with the neuroticism-CHD association, results have been conflicting. Three studies observed higher neuroticism scores among hypertensives (persons in clinical care) in comparison to healthy normotensives (Robinson, 1962; Sainsbury, 1964; Kidson, 1973). However, in other studies in which a nonclinical study population was used, no significant differences in neuroticism were found between hypertensive and normotensive groups in seven studies (Cochrane, 1969; Cochrane, 1973; Kidson, 1973; Schnalling and Svensson, 1984; Santonastoso et al., 1984; Almada et al., 1991; Kohler et al., 1993). Only one nonclinical study found hypertensives to have higher neuroticism scores than normotensives (Cuelho et al., 1989), while another study even found an inverse relationship, lower neuroticism scores in hypertensives as opposed to normotensives (Davies, 1970). The evidence as a whole does not support an association between neuroticism and high blood pressure. Positive findings of such an association were primarily seen in the clinical population group studies (Robinson, 1962; Sainsbury, 1964; Kidson, 1973), but have been questioned in their validity. Hypertensives in early stages often goes unnoticed for many years. The highly neurotic hypertensives are more likely to seek medical attention. Therefore, by selecting hypertensives in clinical settings, high neuroticism scores may also have been selected for.
It should be noted that in the majority of these studies, heterogeneous study populations were used in which possible confounders such as gender, age, professional status, food intake, and physical exercise were not controlled for. However, the most recent study (Kohler et al., 1993) used a large sample of 624 subjects, as homogenous as possible in the aforementioned variables, and still found no association between blood pressure and neuroticism.
Determinants of Neuroticism
There has been remarkably little research on the determinants of neuroticism, particularly social class and job stress. Such factors might shape personality development in childhood. For example, certain parental behavior patterns (i.e., overly strict, critical and demanding of conformity) are more common in low SES households, and may be viewed as a reflection of the parents’ occupational and other life experiences, which are characterized by low control and insecurity. Similarly, an adult’s experience, which might include stressful, low control jobs, may shape their personality development (Kohn and Schooler, 1982). Thus, research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
Due to a lack of positive findings in prospective studies, there is no scientific evidence that neuroticism plays a causal role in coronary heart disease or hypertension. Very few large-scale, rigorous prospective studies have been done however, and those that have often have not used standard assessments of neuroticism. Future, more carefully designed prospective studies may uncover evidence of such a causal role for neuroticism, but there is little scientific evidence to suggest such future findings will be made. Cross-sectional positive associations observed between neuroticism and various CHD endpoints or hypertension have been inconsistent and controversial. Studies that have found positive associations of neuroticism with various CHD endpoints have been flawed, because biases toward a positive association had not been avoided through the use of strict, objective CHD endpoints. Additionally, many of the studies that found support for a neuroticism-CHD relationship cited by Winslow et al., 1989 did not use strict measures of neuroticism. When more reliable, objective CHD endpoints such as angiography have been used, no evidence for a positive association between neuroticism and cardiovascular disease has been observed. Until positive correlation between valid assessments of neuroticism with strict, objective CHD endpoints are observed, there will not be strong research support for such an association. Lastly, studies that have found positive correlation between neuroticism scores and high blood pressure have also been questioned, because this relationship has only been observed in clinical populations of hypertensives where biases toward a positive association exist. Numerous studies in nonclinical settings, more reliable due to the lack of selection bias, have produced consistent null findings. Therefore, there appears to be no strong evidence for an important association between neuroticism and high blood pressure, and significant evidence against such a relationship. Lastly, very little is known about the determinants of neuroticism. Research on the social determinants of personality measures believed to be associated with illness outcomes needs to be a major priority in future research.
Almada, S.J., Zonderman, A.B., Shekelle, R.B., Dyer, A.R., Daviglus, M.L., Costa, P.T., & Stamler, J.(1991). Neuroticism and Cynicism and Risk of Death in Middle-Aged Men: The Western Electric Study. Psychosomatic Medicine, 53, 165-175.
Bass, C., & Wade, C.(1984). Chest pain with normal coronary arteries: A comparative study of psychiatric and social morbidity. Psychological Medicine, 14, 51-61.
Brodman, K., Erdmann, A.J., & Wolff, H.G.(1960). The Cornell Medical Index-health questionnaire manual. New York, NY: Cornell UniversityPress.
Blumenthal, J.A., Thompson, L.W., Williams, R.B., & Kong, Y.(1979). Anxiety-proneness and coronary heart disease. Journal of Psychosomatic Research, 23, 17-21.
Cochrane, R.(1969). Neuroticism and the discovery of high blood pressure. J. Psychosom. Res., 13, 21-25.
Cochrane, R.(1973). Hostility and neuroticism among unselected essential hypertensives. J. Psychosom. Res., 17, 215-218.
Costa, P.T.(1987). Influence of the normal personality dimension of neuroticism on chest pain symptoms and coronary artery disease. Am. J. Cardiol., 60, 20J-26J.
Costa, P.T., Fleg, J.L., McCrae, R.R., & Lakatta, E.G.(1982). Neuroticism, coronary artery disease and chest pain complaints: Cross-sectional and longitudinal findings. Experimental Aging Research, 8, 37-44.
Costa, P.T., & McCrae, R.R.(1985). The NEO Personality Inventory manual. Odessa, FL: Psychological Assessment Resources.
Costa, P.T., & McCrae, R.R.(1987). Neuroticism, somatic complaints, and disease: Is the bark worse than the bite? J. Personal., 55, 299-316.
Costa, P.T., Zonderman, A.B., McCrae, R.R., & Williams, R.B. Jr.(1985). Content and comprehensiveness in the MMPI: An item factor analysis in a normal adult sample. J. Pers. Soc. Psychol., 48, 925-933.
Cramer, D.(1991). Type A behaviour pattern, extraversion, neuroticism and psychological distress. British Journal of Medical Psychology, 64, 73-83.
Cuelho, R., Hughes, A.M., Da Fonseca, F., & Bond, M.R.(1989). Essential hypertension: The relationship of psychological factors to the severity of hypertension. J. Psychosom. Res., 33, 187-196.
Davies, M.(1970). Blood pressure and personality. J. Psychosom. Res., 14, 89-104.
Denollet, J.(1993). Biobehavioral Research on Coronary Heart Disease: Where is the Person? J. Behavioral Med., 16(2), 115-141.
Elias, M.F., Robbins, M.A., Blow, F.C., Rice, A.P., & Edgecomb, J.L. 1982). Symptom reporting, anxiety and depression in arteriographically classified middle-aged chest pain patients. Experimental Aging Research, 8, 45-51.
Eysenck, H.J., & Eysenck, S.B.G.(1964). Manual of the Eysenck Personality Inventory. London: University Press.
Eysenck, H.J., & Eysenck, S.B.G.(1975). Manual of the Eysenck Questionnaire. San Diego, CA: EdITS.
Friedman, H.S.(1990). Where is the disease-prone personality? Conclusion and future directions. In H.S. Friedman(Ed.), Personality and Disease(pp. 283-292). New York, NY: Wiley.
Friedman, H.S., & Booth-Kewley, S.(1987). The Rdisease-prone personalityS: A meta-analytic view of the construct. Am. Psychol., 42, 539-555.
Goldbourt, U. Medalie, J.H., & Neufeld, H.N.(1975). Clinical Myocardial infarction over a five-year period- III. A multivariate analysis of incidence, The Israel Ischemic Heart Disease Study. Journal of Chronic Disease, 28, 217-237.
Guilford, J.S., Zimmerman, W.S., & Guilford, J.P.(1976). The Guilford-Zimmerman Temperament Survey Handbook: Twenty-five years of research and application. San Diego, CA: EdITS.
Hallstrom, T., Lapidus, L., Bengtsson, C., & Edstrom, K.(1986). Psychosocial factors and risks of ischemic heart disease and death in women: A twelve-year follow-up of participants in the population study of women in Gothenburg, Sweden. Journal of Psychosomatic Research, 30, 451-459.
Jenkins, C.D.(1971). Psychologic and social precursors of coronary disease. N. Engl. J. Med., 284, 244-255, 307-317.
Jenkins, C.D.(1976). Recent evidence supporting psychologic and social risk factors for coronary disease. N. Engl. J. Med, 294, 987-994.
Keehn, R.J., Goldberg, I.D., & Beebe, G.W.(1974). Twenty-four-year mortality follow-up of army veterans with disability separations for psychoneurosis in 1944. Psychosomatic Medicine, 36, 27-46.
Kidson, M.A.(1973). Personality and hypertension. J. Psychosom. Res., 17, 35-41.
Kohler, T., Scherbaum, N., Richter, R., & Bottcher, S.(1993). The relationship between neuroticism and blood pressure reexamined. Psychother. Psychosom., 60, 100-105.
Lichtenstein, P., Pedersen, N.L., Plomin, R., Faire, U.D., & McClearn, G.E.(1989). Type A behavior pattern, related personality traits and self-reported coronary heart disease. Person. individ. Diff., 10(4), 419-426.
Maddi, S.R.(1980). Personality theories: A comparative analysis (p.463). Homewood, IL: Dorsey Press.
Matthews, K.A.(1988). Coronary Heart Disease and Type A behaviors: Update on and alternative to the Booth-Kewley and Friedman(1987) quantitative review. Psychological Bulletin, 104, 373-380.
Ostfeld, A.M., Lebovits, B.Z., Shekelle, R.B., & Paul, O.(1964). A prospective study of the relationship between personality and coronary heart disease. Journal of Chronic Diseases, 17, 265-276.
Robinson, J.O.(1962). A study of neuroticism and causal arterial blood pressure. Br. J. Clin. Psychol., 2, 56-64.
Sainsbury, P.(1964). Neuroticism and hypertension in an out-patient population. J. Psychosom. Res., 8, 235-238.
Santonastoso, P., Canton, G., Ambrosini, G.B., & Zamboni, S.(1984). Hypertension and neuroticism. Psychother. Psychosom., 41, 7-11.
Schalling, D., & Svensson, J.(1984). Blood pressure and personality. Person. Individ. Diff., 5, 41-51.
Stone, S.V., & Costa, P.T.(1990). Disease-Prone Personality or Distress-Prone Personality? The Role of Neuroticism in Coronary Heart Disease. In H.S. Friedman(Ed.), Personality and Disease(pp. 178-200). New York, NY: John Wiley and Sons.
Winstow, D.J., Wakefield, J.A., & Goldsmith, W.M.(1990). The relationship between personality, health symptoms and disease. Person. individ. Diff., 11(3), 717-723.
Zyzanski, S.J., Jenkins, C.D., Ryan, T.J., Flessas, A., & Everist, M.(1976). Psychologic correlates of coronary angiographic findings. Archives of Internal Medicine, 136, 1234-1237.