Smoking, heavy drinking, and weight changes are consequences of vulner
able health behaviour. Factors predicting adverse changes in health be
haviour are not well known. We monitored the workers of a wood-process
ing factory who were given notice simultaneously and a matched group o
f employed factory workers for 1 year. Health behaviour was considered
vulnerable if the responder reported increased smoking or alcohol con
sumption or if the responder's weight had changed by at least 2 kg dur
ing follow-up. At least two of the above three changes were required f
or inclusion in the vulnerable health behaviour group. Vulnerable heal
th behaviour change was observed in 26.6% of the unemployed and in 13.
7% of the employed (P = 0.009). Vulnerable health behaviour change was
not associated with sociodemographic background variables (age, sex,
education, financial status, social support) in either of the groups.
In univariate analysis high psychic stress levels reported at the begi
nning of follow-up by both employed and unemployed persons were predic
tive of vulnerable health behaviour. In the unemployed, abundant psych
osomatic symptoms also predicted vulnerable health behaviour change. I
n logistic regression analysis vulnerable health behaviour change was
predicted in the whole sample by the base-line psychic stress level an
d marital status other than marriage/cohabitation. To summarize, the m
ost interesting observation in our study was that psychic stress was a
risk factor for health behaviour harmful to physical health. Thus, at
tention should be paid to the functional connection between mind and b
ody to better maintain total well-being. Additionally, persons living
alone are a risk group for vulnerable health behaviour.