106 COPING RESOURCES
scores on the index were about twice as likely to die as those with high scores, even
after controlling for self-reports of social class, smoking, obesity and health at the outset.
One limitation of this study was the use of self-reports of health in the initial
measurements but findings were later replicated using physical examinations. In a
review of a range of such studies, House, Landis and Umberson (1988) concluded that
evidence consistently supports the view that there is an increased risk associated with
having few social relationships even after adjusting for other risk factors. However,
there are gender differences. In particular, marriage has more health benefits for men
than for women and bereavement is more harmful for men. More recently poor social
networks, or social isolation (as it is now more frequently called), have been linked
to strokes (Rutledge et al., 2008), decline in cognitive functioning (Shankar et al., 2013)
and depression (Golden et al., 2009).
Quality of relationships
Being part of a large social network does not guarantee that people will receive greater
help when they need it. The correlation between the number of connections people
have and the actual support they receive tends to be quite low, perhaps because one
good relationship may provide better support than a large number of more superficial
contacts (Cohen and Wills, 1985). The importance of quality rather than quantity of
relationships is demonstrated by a well-known sociological study conducted by Brown
and Harris (1978), who studied the origins of depression in women. They interviewed
400 women about the life events they had experienced in the past year. They also
asked the participants to name the people they were able to confide in about their
worries. Women were classified into one of the four following categories: (a) those
who had a close relationship with someone in the same household; (b) those without
such a relationship who had a friend or relative they saw at least weekly; (c) those with
a close friend or relative they saw less than weekly; and (d) those with none of these
relationships. The study found that having a confiding relationship protected the
women from depression following major life events. Among the women who
experienced a stressful life event, only 1 in 10 of those in category a developed
depression, compared to 1 in 4 of those in category b and 1 in 2.5 of those in categories
c or d. This suggests that having confiding social support buffered the impact of life
stressors.
Functional social support
None of the above measures taps specific supportive behaviours. However, it is
important to consider what types of behaviour may be most helpful if we wish to
develop effective social support interventions (e.g. Dakof and Taylor, 1990). For
example, what type of support would be most helpful when an individual received
a diagnosis of cancer? Various types of specific social support have been assessed,
including emotional support (helping the person to feel accepted or valued), instru -
mental support (i.e. practical support) and informational support (e.g. Cohen and Wills,
1985). The match between support provided and an individual’s need may be crucial
to effectiveness. For example, a study of support from family and friends provided for
breast cancer sufferers (Reynolds and Perrin, 2004) compared a range of provided
support with the support desired by the women. The study found that two behaviours
that were intended to be supportive were unwelcomed by over 90 per cent of the