The systemic use of steroids and habitual alcohol
intake are two major causative factors in the development of idiopathic
osteonecrosis of the femoral head (ONFH). To examine any interaction
between oral corticosteroid use and alcohol intake on the risk of
ONFH, we conducted a hospital-based case-control study of 71 cases
with ONFH (mean age 45 years (20 to 79)) and 227 matched controls
(mean age 47 years (18 to 79)). Alcohol intake was positively associated
with ONFH among all subjects: the adjusted odds ratio (OR) of subjects
with ≥ 3032 drink-years was 3.93 (95% confidence interval (CI) 1.18
to 13.1) compared with never-drinkers. When stratified by steroid use,
the OR of such drinkers was 11.1 (95% CI 1.30 to 95.5) among those
who had never used steroids, but 1.10 (95% CI 0.21 to 4.79) among
those who had. When we assessed any interaction based on a two-by-two
table of alcohol and steroid use, the OR of those non-drinkers who
did use steroids was markedly elevated (OR 31.5) compared with users
of neither. However, no further increase in OR was noted for the
effect of using both (OR 31.6). We detected neither a multiplicative
nor an additive interaction (p for multiplicative interaction 0.19;
synergy index 0.95), suggesting that the added effect of alcohol
may be trivial compared with the overwhelming effect of steroids
in the development of ONFH. Cite this article:
