Because experimental and epidemiologic evidence indicates that the col
on is particularly sensitive to stress, and because work conditions co
ntribute to an individual's stress experience, we examined the relatio
n of both job stress (defined in terms of perceived job demand and job
control) and job social support to the risk of colon cancer in a larg
e population-based case-control study (744 pairs) in Los Angeles. Cont
rols were individually matched to cases on age, sex, and neighborhood.
For jobs held 5 years before, participants in the lowest tertile of j
ob control had a slightly increased risk when compared with those in t
he highest tertile (multivariate adjusted odds ratio = 1.3; 95% confid
ence interval = 1.0-1.6), but there was no evidence of a trend. Lower
levels of job social support were associated with a decreased risk of
colon cancer (odds ratio = 0.6 for lowest us highest tertile; 95% conf
idence interval = 0.4-1.0). We saw no effect for job demand. The effec
t of job control appeared to be independent of the level of job demand
. We found no consistent pattern of effects associated with jobs held
30 years before. These findings indicate that if job stress, as reflec
ted by perceived job demand or control, is a determinant of colon canc
er, it is probably not a strong one.