et al. 1981; Karasek and Theorell 1990). According to the model, there are two
aspects of job strain: job demands, which reflect conditions that effect performance, and
job autonomy, which reflects the control over the speed or the nature of decisions made
within the job. Karasek’s job demand and control model suggests that high job
demands and low job autonomy (control) predicts coronary heart disease. Karasek and
co-workers have since developed the job demand control hypothesis to include social
support. Within this context, social support is defined as either emotional support,
involving trust between colleagues and social cohesion, or instrumental social support
involving the provision of extra resources and assistance. It is argued that high social
support mediates and moderates the effects of low control and high job demand.
Karasek and Theorell (1990) report a study in which subjects were divided into low
social support and high social support groups, and their decisional control and the
demands of their job were measured. The results indicated that subjects in the high
social support group showed fewer symptoms of CHD than those subjects in the low
social support group. In addition, within those groups high job control and low job
demands predicted fewer CHD symptoms. A series of studies have tested and applied
Karasek’s model of job strain and associations have been reported between job strain
and risk factors for heart disease and heart disease itself (Pickering et al. 1996; Schnall
et al. 1996; Marmot 1998; Tsutsumi et al. 1998). For example, Kivimaki et al. (2002)
used a prospective design to explore the links between job strain and subsequent death
from cardiovascular disease. A total of 812 employees from a metal factory in Finland
that manufactures paper machines, tractors, firearms along with other equipment
completed a baseline assessment in 1973 including measures of their behavioural and
biological risks and their work stress. Those with cardiovascular disease at baseline
were excluded. Cardiovascular mortality was then recorded between 1973 and 2001
using the national mortality register. The results showed that 73 people had died from
cardiovascular disease since the study onset who were more likely to be older, male,
have low worker status, to smoke, have a sedentary lifestyle, high blood pressure, high
cholesterol and higher body mass index. Further, when age and sex were controlled for
death was predicted by high job strain and low job control. However, after occupational
group was also controlled for (i.e. a measure of class), high job strain remained the
best predictor.Relationship stress
There is much evidence indicating an association between relationship status, psycho-
logical distress and health status. For example, separated and divorced people have the
highest rates of both acute and chronic medical problems even when many demo-
graphic factor are controlled for (Verbrugge 1979). In addition, these people also have
higher rates of mortality from infectious diseases such as pneumonia (Lynch 1977).
They are also over represented in both inpatient and outpatient psychiatric populations
(Crago 1972; Bachrach 1975). However, it is not just the presence or absence or a
relationship that is important. The quality is also linked to health. For example, whereas
marital happiness is one of the best predictors of global happiness (Glenn and Weaver
1981) those in troubled marriages show more distress than those who are unmarried266 HEALTH PSYCHOLOGY