Is subjective social status a more important determinant of health than objective social status? Evidence from a prospective observational study.

Health and social position

Social disadvantage was associated with poorer health (as indexed by all cause mortality) in this population, whatever the measure of social position used. Having a father in a manual occupation was the strongest predictor of poorer health. Other strong predictors were height and car access. Workplace status, current occupation, educational status and area of residence were all strongly associated with health in age-adjusted analyses. Adjustment for health behaviours and physiological risk factors attenuated these associations. Adjustment for psychological stress had little influence. Workplace status, current occupation, educational status and area of residence showed little association with health after adjustment for other measures of social position.

Objective compared to subjective social position

Men in contradictory class locations, that is men who enjoyed higher workplace status but who were similar to their subordinates in terms of other social position measures, had, in general, a similar health experience to these subordinates. Their status advantage did not appear to confer any substantial health advantage. Foremen, compared to employees, experienced a reduced risk of most of the categories of morbidity and mortality. However, the evidence in support of these estimated differences did not provide any strong basis to discount the null hypothesis of no difference between the two groups of men. In terms of cardiovascular disease (the most important cause of health inequality within this population) there was actually a suggestion that foremen were at higher risk than employees.

The role of psychological stress

Within the psychosocial explanatory framework, the predicted health of individuals in contradictory class locations depends on the psychological impact that such a location is assumed to have on the individual occupying it. Muntaner and colleagues suggest that this impact will be negative. In their study a weak relation between “supervisor” status and increased prevalence of substance misuse disorders appeared to support this suggestion. Our data provide some further support for this conceptualisation. Foremen reported higher psychological stress than employees (though not managers) and were more likely than both employees and managers to be heavy smokers in cross-sectional analyses at recruitment (proportions of current smokers were almost identical between employees and foremen).

An alternative conceptualisation, as suggested by the work of Wilkinson, is that the superior relative status of foremen should have a positive psychological impact and that this should further translate into better health. This framework appears to be supported by the findings of Adler and colleagues who found a closer relation between higher subjective (as opposed to objective) social status and better self-reported physical and psychological health. More recently, Mustard and colleagues also found that occupational prestige was directly associated with perceived health status, particularly in men. Objective measures of health status were not available in this study and it is likely that reporting tendency influenced the associations found between prestige and perceived health. The association was attenuated following adjustment for psychosocial work characteristics but apparently not by adjustment for household income.

In the present study, foremen had a reduced risk of admission to hospital with a psychiatric disorder compared to employees. However it is difficult to attribute this to a positive psychological impact of their workplace status as foremen reported higher stress than employees.

These data cast further doubt on the importance of psychological stress as an important cause of social gradients in physical disease. Managers, the group reporting highest stress, had a substantially lower risk of admission to hospital for peptic ulcer—the condition that for much of the follow-up period of this study would have been considered the classic “stress disease”. Managers may have been more mobile than foremen or employees (75% of managers, compared to 69% of foremen and 42% of employees, regularly drove a car). Conceivably, this may have made them more likely to be admitted to hospital outside of Scotland. Such admissions would not have been recorded (death outside of Scotland is recorded) however it seems unlikely that the protective effect of managerial status is substantially attributable to bias of this nature. Managers also had a considerably lower experience of other health outcomes related to psychological distress such as deaths from accidents and violence, alcohol related morbidity and mortality and psychiatric hospitalisation. Adjustment for social position measures considerably attenuated all these relations. Adjustment for smoking and alcohol consumption (which in the cohort as a whole were strongly associated with stress) and for stress made little difference to estimates.

Implications

Our findings suggest several conclusions. They confirm the existence of social inequalities in health and demonstrate that such inequalities exist whether subjective or objective social classifications are used. They illustrate that different indicators of social position may capture different dimensions of disadvantage experienced across the life course and the health consequences of this. They also confirm that although unhealthy behaviours are commoner amongst socially disadvantaged people, such behaviours appear to explain only part of the social health gradient in cohorts of this age. They lend no support to the suggestion that psychosocial stress, associated with perceptions of relative social status is an important determinant of health. Furthermore they suggest that associations between perceived social status and objective health arise predominantly because most individuals appear objectively accurate in their assessment of their own relative social status.

Our findings give little support to the thesis that subjective social status is a more important determinant of health than objective social status. Rather they suggest that—even in a relatively economically advanced society—it is material inequality itself, rather than any psychosocial correlate of such inequality that is the key determinant of health inequalities.