1996 Volume 6 Issue 1 Pages 37-43
A cooperative hospital-based case-control study of idiopathic avascular necrosis of the femoral head (IANF) was carried out to clarify the involvement of smoking, drinking, flushing pattern, and other factors in the development of IANF comparing 90 cases (64 males and 26 females) without history of systemic corticosteroid use with 180 matched controls (128 males and 52 females). The results of analyses were represented only for male subjects because of small number of female cases. There was no significant difference in smoking habits, daily and cumulative number of cigarettes smoked between case and control groups. Current drinkers had obviously higher risk (OR=11.47) of IANF compared to nondrinkers or exdrinkers. In addition, there was a consistent risk increase with increasing alcohol consumption and the highly significant dose-response relationship remained unchanged after adjustment for all other factors (X2=14.33, p<0.001 and X2=13.24, p<0.001 for daily and cumulative alcohol consumption, respectively). For flushing pattern, although nonflushers had a significantly elevated risk (OR=2.08) in the univariate analysis, the association disappeared (OR=0.73) after adjustment for alcohol and other factors. Since nonflushers tend to be heavy drinkers, perhaps, an apparent risk increase among nonflushers may be due to alcohol drinking. Body mass index (BMI) was inversely related to the development of IANF. The risk reduction was found among subjects with higher BMI and an adjusted linear trend of OR was significant (X2=6.65, p<0.05). However, further studies were required regarding the association between flushing pattern or BMI and IANF because of a few reports. History of liver diseases and occupational history were not significantly associated with the development of IANF after adjustment for other factors. J Epidemiol, 1996; 6:37-43.
