have higher resting blood pressure and their blood pressure returns more slowly to
resting levels following stressful encounters. High status animals also show changes
in hormonal and cardiovascular functioning indicative of stress when their position in
the hierarchy changes, for example, when a larger male is introduced into their group.
Similar correspondence between such physiological indices of stress response and social
position is observed in people, e.g. among civil servants working at different levels
within government (Marmot et al., 1991; Wilkinson, 1996).
A further puzzling finding is that, while SES is strongly linked to mortality in men
(when their SES is determined by their occupation) this is not the case for women.
In fact, women’s mortality is more strongly affected by socio-economic class when
they are classified according to their husband’s job and not their own (Bartley et al.,
2004). One explanation is that, traditionally, a husband has been the main breadwinner
and his participation in the workforce was often more enduring. Thus men’s
occupation has more influence on the overall living standards of the family and so
determined family members’ positions within the societal hierarchy.
Traditionally, it has been hypothesized that inequalities in health are due to
material/structural and cultural/behavioural differences between socio-economic
groups. However, researchers debated whether social inequality causes ill health or ill
health causes social inequality (Carroll, Bennett and Davey Smith, 1993). This has given
rise to two opposing explanations. The first is known as the social causation hypothesis,
which states that low SES causes ill health. In other words, factors associated with
occupying a low socio-economic position negatively impact on health. The alternative
explanation is known as the social drift hypothesis, which states that ill health causes
low SES, that is, when an individual becomes ill, they drift down the socio-economic
hierarchy because they may be unable to hold down a job. In general, more evidence
supports the former explanation. In longitudinal studies following large samples of
individuals over time, baseline measures of SES have been found to be good predictors
of subsequent health status, whereas health status has been found to be a weaker
predictor of SES. Moreover, if ill health caused SES decline one might expect to see
differences between fast-acting fatal illnesses (e.g. lung cancer) where there is usually
little time to change SES and chronic illnesses (e.g. chronic bronchitis). Yet the SES
gradient is seen equally strongly in both types of illness (Carroll et al., 1993).
Stress and social inequality
Cardiovascular disease has been found to be associated with SES and this relationship
is not eliminated after conventional risk factors such as smoking are taken into
account. This has led researchers to search for additional factors that explain these social
inequalities. Two hypotheses have been suggested relating to the role of stress in
contributing to social inequality (e.g. Adler et al., 1994). First, the differential exposure
hypothesis maintains that the higher prevalence of health problems in low socio-
economic groups may be associated with a greater exposure to psychological stressors
in these groups. Second, the differential vulnerability hypothesis suggests that indi -
viduals in lower socio-economic groups are less well equipped to cope with stressors
due to having fewer resources (e.g. having less money to buy healthy foods, choosing
less effective coping strategies and having limited social support networks) and as such
their impact is much greater in these groups. Evidence supporting both hypotheses
70 STRESS AND HEALTH