Image: Hartmut Schwarzbach/Still Pictures
fatty deposit in the
intima (inner lining)
of an artery.
40 CHAPTER 2 • HEALTH INEQUALITIES
rates are about 25 percent higher among tenants than owner-occupiers
(Filakti and Fox 1995). Tenants also report higher rates of long-term illness
than owner-occupiers (Lewis, Bebbington, Brugha et al. 1998). Woodward,
Oliphant, Lowe et al. (2003), for example, found that after adjusting for age,
male renters were 1.48 times more at risk of developing CHD than male
owner-occupiers; women renters were 2.6 times more likely to develop CHD
than their owner-occupier counterparts. There are a number of explanations
for these differentials:
n Renters may experience more damp, poor ventilation, overcrowding and
n Rented occupation may be further away from amenities, making access to
leisure facilities or good-quality shops more difficult.
n Renters earn less than people who own their house.
n The psychological consequences of living in differing types of accommodation
may directly impact on health.
Although the fourth pathway has received little attention, Macintyre and
Ellaway (1998) found that a range of mental and physical health measures
were significantly associated with housing tenure, even after controlling for
the quality of housing and the age, sex, income and self-esteem of their
occupiers. They interpreted these data to suggest that the type of tenure itself
is directly associated with health. They suggested, for example, that the
degree of control we have over our living environment may influence mood,
levels of stress and perceived control over a wider set of health behaviours –
all of which may contribute to ill-health.
n Stress, strain and depression
The implication of the previous section is that poor housing leads to stress,
which in turn leads to ill-health. This argument can be widened to suggest
that differences in more general stress experienced across the social groups
leads to differences in health across the social groups. This seems a reasonable
hypothesis. There is consistent evidence that people in the lower socio-economic
groups not only experience more stress than those in the higher socio-economic
groups (Marmot, Ryff, Bumpass et al. 1997; McLeod and Kessler 1990),
they frequently have fewer resources to help them to cope. As a result, they
may be more vulnerable to stress than those in the higher economic groups.
They may also have fewer uplifts that reduce stress than people in higher
socio-economic groups. There is also significant evidence to suggest that stress
can adversely impact on health. We consider the evidence for this, and the
mechanisms that explain how stress can lead to disease, in Chapter 11.
Here we consider the first part of this linkage: the contention that people
in lower socio-economic groups experience more stress than those in higher
groups. Some of the stresses (and health-compromising behaviour and restrictions
in life opportunities) that may be experienced more by people in lower
socio-economic groups than the more economically better-off are summarised
below (see Carroll, Davey-Smith and Bennett 1996):
n childhood: family instability, overcrowding, poor diet, restricted educational
THE IMPACT OF POVERTY ON HEALTH 41
n adolescence: family strife, exposure to smoking and own smoking, leaving
school with poor qualifications, experiencing unemployment or low-paid
and insecure jobs;
n adulthood: working in hazardous conditions, financial insecurity, periods
of unemployment, low levels of control over work or home life, negative
n older age: no or small occupational pension, inadequate heating, food, etc.
These types of data have been formalised in Hobfoll and Lilly’s (1993) conservation
of resources model, which proposed that mental and physical
health are determined by the amount of resources available to the individual.
These may be:
n economic (e.g. job, income)
n social (e.g. family support)
n structural (e.g. housing)
n psychological (e.g. coping skills, perceived control).
A high level of resources is health-protective. Low levels of resources place an
individual at risk of health problems. Although the model has received general
support (e.g. Hobfoll 2001), its relation to socio-economic factors has
yet to be fully examined. The model is contrary to the individual appraisal
models of stress discussed in Chapter 11 and suggests that some factors
impact so markedly on our mental and physical health that they are not mediated
by individual appraisals but affect us all in similar ways. As such, the
model indicates the type of factors that will impact on the health of people
across society, albeit without explaining how these factors actually do so.
A more specific emotion that is increasingly associated with disease, and in
particular CHD, is depression. Just as in the case of stress, depression rates
are not equally distributed through society. They are relatively high among
the poor, ethnic minorities, and those with poor social or marital support.
Accordingly, differences in the prevalence of depression across the social
groups may also contribute to the differences in health. One final factor
that may interact with SES to influence risk for disease is the social support
available to the individual. A large number of positive social relationships
and few conflictual ones may buffer individuals against the adverse effects
of the stress associated with low economic resources. Conversely, low SES
combined with a poor social support system may significantly increase risk
for disease (Taylor and Seeman 1999). Sadly, the potentially protective effect
of good social support may be less widely available than previously. In contrast
to research conducted in the 1950s, those in the higher social groups
now appear to have more social support than those in the lower social groups
(e.g. Marmot, Davey-Smith and Stansfield 1991; Ruberman, Weinblatt,
Goldberg et al. 1984).
n Access to healthcare
Access to healthcare is likely to differ according to both personal characteristics
and the healthcare system with which the individual is attempting to
interact. Perhaps the most studies of this phenomenon have been conducted
in the USA, where different healthcare systems operate for those with and