Background. Few studies among young adults have examined clustering of
health behaviors affecting risk for lifestyle diseases. Methods. Smok
ing, alcohol consumption, physical activity, and diet were examined am
ong Australian 18-year-olds (301 males, 282 females) initially recruit
ed at the age of 9 years from 26 schools, Association analysis was use
d to recognize behavior clustering. Results. Fat intake was greater am
ong male smokers than nonsmokers (36% energy vs 34% energy). Women smo
kers ate less fiber (14.1 g/day) than did nonsmokers (17.8 g/day). Smo
king was significantly related, among males, to unsafe drinking (odds
ratio 2.38) and higher fat intake (odds ratio 1.06) and, among females
, to unsafe drinking (odds ratio 1.59), lower dietary fiber (odds rati
o 0.93), and less physical activity (odds ratio 0.36). Cluster analysi
s defined separate behavior clusters for men and women with smoking st
atus identifying further subgroups. Smoking, drinking alcohol to exces
s, and adverse dietary choices clustered among men and women, with phy
sical inactivity also clustering among women. Conclusion. Smoking amon
g adolescents is an important indicator of behaviors influencing risk
for later cardiovascular disease and other medical disorders. Multimod
al approaches allowing for gender differences in health-related behavi
ors are likely to be more successful than targeting a single behavior
in this age group. (C) 1997 Academic Press.