Background. Information on a possible relation between coffee, caffein
e, and other methylxanthine-containing beverages and hip fracture is r
elatively scanty and controversial. We present here the results of a c
ase-control study conducted in northern Italy. Methods. A total of 279
cases of hip fracture and 1,061 controls in hospital for acute, nonne
oplastic nontraumatic, non-hormone-related diseases were interviewed d
uring their hospital stay. Odds ratios (OR) and their 95% confidence i
ntervals (CI) according to consumption of coffee and other methylxanth
ine-containing beverages were derived from multiple logistic regressio
n equations including terms for age, education, body mass index, smoki
ng status, alcohol drinking, calcium intake, menopausal status, and es
trogen replacement therapy. Results. Compared with nondrinkers, the OR
was 1.2 (95% CI, 0.8 to 1.7) for coffee drinkers. No association emer
ged with number of cups/day (OR = 1.2 for 1 cup/ day, 1.0 for 2, 1.4 f
or 3 and 4, and 1.2 for 5 or more cups/day) or with duration of coffee
intake (OR = 1.0 for less than 30 years and OR = 1.1 for more than 30
years). Similarly, no statistically significant association was obser
ved with decaffeinated coffee (OR = 1.3), tea (OR = 1.3), or cola inta
ke (OR = 0.6). OR for coffee drinking were computed across strata of a
ge, menopausal status, education, smoking status, total alcohol drinki
ng, and calcium intake. Conclusions. The present study found no associ
ation between hip fractures among women and consumption of regular or
decaffeinated coffee, tea, and cola. (C) 1995 Academic Press, Inc.