University of Dusseldorf, Dept. of Medical Sociology, D-40225 Dusseldorf, Germany. E-mail: siegrist{at}uni-dusseldorf.de
In his Logic of Scientific Discovery Karl R Popper argued that a theory is based on risky assumptions that deviate from what is already known.1 Therefore, a (preliminary) confirmation of a theory has the potential of creating new knowledge while its refutation by empirical data produces some pressure towards reconsidering the theoretical formulation. It is one of the advantages of the job strain model proposed by Karasek2 that risky assumptions can be derived from it. Up to now, there is impressive empirical support in favour of this model, especially so with respect to cardiovascular risk and disease.3 It is true that the majority of investigations were concerned with men. Therefore, a test of the model in a large prospective study of women, as the one reported in this issue of the International Journal of Epidemiology by Lee et al.,4 is particularly instructive. Its negative results give rise to several considerations.
First, one may ask whether the study design, the measures used or the sample were adequate to test the model. Even when taking the authors own arguments about limitations of their study into account it is difficult to question the strengths of this report. The Nurses Health Study covers a large sample of a homogeneous professional group. Despite considerable sample loss its size provides adequate statistical power to test the hypothesis and to adjust for multiple confounders. Moreover, confounding of findings by socioeconomic status is excluded given the rather homogeneous professional background. Restriction of health outcomes to medically diagnosed incident cases of coronary heart disease must be considered an additional strength of the study.
A second consideration concerns the specific work stress profile of nursing. One may argue that for many nurses a low level of task control in combination with high job demands and low workplace social support is perhaps less stressful than coping daily with patients needs and expectations or exposure to their suffering and ill health.5 Moreover, lack of esteem by physicians, inadequate salaries and restricted promotion prospects may contribute to work stress among nurses.6 These aspects have not been measured in the present study. Similarly, as briefly discussed by the authors, the work-non work interface, and particularly stressful conditions at home, have not been adequately assessed in the current investigation, thus giving rise to potential underestimation of the contribution of psychosocial stressors to cardiovascular risk and disease in nurses (and women in general).
However, in keeping with Poppers argument the negative findings of this study might challenge the job strain model in a productive way as they urge us to consider gender (or gender role) differences more carefully. It may well be that a substantial part of stressful experience at work is contingent on the perceived threats associated with ones occupational position, and that men, as a result of socialized gender roles, are generally more vulnerable to these threats than women. For instance, in terms of the social cognitive theory of gender differentiation,7 women may be better suited to combine different roles or to change roles with more flexibility and thus to profit from multiple sources of self-efficacy and self-esteem. Men, on the contrary, often stick more exclusively to their occupational role as it provides a major source of their self-reliance.
Broadly speaking, sociocultural factors influence the appraisal of demands and threats in salient social roles in adult life, and these influences are embedded in gender-based practices of coping. Models of work stress might be designed in a way that takes these considerations into account. Interestingly, two recent studies of occupational stress and cardiovascular risk comparing men and women in Sweden, found some evidence along these lines.8,9 In men, the threats to occupational status in terms of low control and low reward were more strongly associated with cardiovascular risk than in women. Conversely in women, inadequate or excessive ways of personal coping with the demands at work predicted disease risk more strongly than in men.
It is certainly premature to evaluate the relevance of gender or gender roles in explaining links between psychosocial stress at work and health. Yet, the negative findings reported by Lee et al.,4 rather than being discouraging, stimulate new thoughts in an area of research that continues to be seminal and important.
References
1 Popper KR. The Logic of Scientific Discovery. London: Routledge, 1959.
2 Karasek R, Theorell T. Healthy Work. Stress, Productivity, and the Reconstruction of Working Life. New York: Basic Books, 1990.
3 Schnall PL, Belkic K, Landsbergis P, Baker D (eds). The workplace and cardiovascular disease. Occupational Medicine: State of the Art Reviews 2000;15:1334.[ISI]
4 Lee S, Colditz G, Berkman L, Kawachi I. A prospective study of job strain and coronary heart disease in US women. Int J Epidemiol 2002;31:114753.
5 Dewe PJ. Identifying the causes of nurses stress: a survey of New Zealand nurses. Work & Stress 1987;1:1524.[ISI]
6 Bakker AB, Killmer CH, Siegrist J, Schaufeli WB. Effort-reward imbalance and burnout among nurses. J Adv Nurs 2000;31:88491.[CrossRef][ISI][Medline]
7 Bussey K, Bandura A. Social cognitive theory of gender development and differentiation. Psychol Rev 1999;106:676713.[CrossRef][ISI][Medline]
8 Peter R, Alfredson L, Hammer 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. J Epidemiol Community Health 1998;52:54047.[Abstract]
9 Peter R, Siegrist J, Hallqvist J, Reuterwall C, Theorell T & the SHEEP study group. Psychosocial work environment and myocardial infarction: improving risk estimation by combining two alternative job stress models in the SHEEP study. J Epidemiol Community Health 2002;56:294300.