I. Foppa et Rh. Noack, THE RELATION BETWEEN THE NUMBER OF SYMPTOMS AND OTHER HEALTH INDICATORS IN WORKING MEN AND WOMEN, European journal of public health, 7(1), 1997, pp. 9-14
Simple symptom counts are widely used indicators of ill-health in heal
th survey research. However, there has only been little research aimed
at a better understanding of symptom counts. The purpose of this stud
y was to explore the number of symptoms (a summary score derived from
10 unspecific self-reported symptoms) in a 'healthy' population. We re
analysed data on a sample of 850 employed men and women (Berne Workpla
ce Health Project). Stepwise ordinal logistic regression analysis was
used to study the relation of the number of symptoms to contextual, pe
rson related, behavioural, and health-related variables. In neither me
n nor women were any of the sociodemographic indicators significantly
associated with the number of symptoms. Of the work-related variables,
low job discretion and adverse work environment were significantly as
sociated with the number of symptoms and there was a statistical trend
for high job demands. In women, dissatisfaction with salary was the o
nly work-related variable which was significant. The private context,
on the other hand, seemed to be more influential for women than for me
n and there was a statistical trend for strain at home. Among the pers
on-related factors studied, age and emotional problems were significan
tly associated to the number of symptoms in both men and women. Psycho
logical ill-health and negative future orientation were significant on
ly in men and immersion only in women. Physical inactivity in leisure
time was the only behavioural variable found to be significantly assoc
iated to the number of symptoms. Among the health-related factors, poo
r self-rated health was significant for both gender groups, while obes
ity was significant only in men. Only categories of variables being hy
pothesized to be unfavourable in terms of health were associated with
a higher number of symptoms. Most of the associations found seem to be
theoretically meaningful. Despite methodological provisos, our findin
gs support a conception of the number of symptoms as an indicator of g
eneral ill-health.