It is well documented that alcohol abuse is associated with an increas
ed risk of fracture in men, but the relationship of moderate alcohol i
ntake to bone mineral density (BMD) is not as clear. We examined the r
elationship between self-reported alcohol intake and BMD measured at t
he hip and spine using dual X-ray absorptiometry in a population-based
cohort of men. The men were categorized as (1) non-drinkers and drink
ers and (2) as having no, low, medium, and high alcohol intake (units/
week). 458 men aged 69.1 +/- (SD) 2.8 (range 64-76) years were studied
. Drinkers were significantly younger (p = 0.01), had lower recorded c
affeine intake (p = 0.02), and had fewer pack years of smoking (p = 0.
04) than non-drinkers. The mean BMD at the hip was significantly highe
r in drinkers as compared with non-drinkers (p = 0.02) before and afte
r adjusting for age and weight. However, after further adjusting for p
ack years of smoking, caffeine intake, and physical activity, the diff
erence remained significant at the trochanter only. Men categorized in
the highest tertiles of alcohol intake (> 11 units/week) were younger
(p = 0.3), had lower caffeine intake (P = 0.008), and had significant
ly higher gamma-glutamyltransferase levels (p = 0.002) than the other
categories. The crude mean BMD at femoral neck, trochanter, and Ward's
triangle was significantly higher in men who drank 1-4 units of alcoh
ol per week. However, most of this association disappeared when BMD wa
s adjusted for age and weight except at the trochanter. The associatio
n at the trochanter did not remain after adjusting for all covariates.
In men moderate alcohol intake (1-2 drinks per day) does not appear t
o have a detrimental effect on BMD, and even high intakes of alcohol,
in this older cohort, did not appear to decrease BMD. Non-drinkers hav
e consistently low BMD measurements, which confirms that a small amoun
t of alcohol may benefit BMD.