This epidemiological study will examine the impact of alcohol dependence and recovery on bone mineral density and bone turnover in premenopausal women. It will be completed in 2 phases. In phase l, data will be collected on 250 women, aged 20 to 40 years who are receiving treatment for alcohol dependence at a residential treatment facility in central Iowa. Bone mineral density of the femoral neck and lumbar spine will be measured using dual energy x-ray densitometry. Biological and self- reported factors which may impact bone density will also be evaluated. These include: l) reproductive characteristics 2) nutritional status ,measured by body composition, eating disorders, serum 25(OH)D and 1,25(OH)D and caffeine intake, 3) extent of alcohol use 4) liver function as measured by serological LFT's and 5) behavioral characteristics such as smoking, physical activity and other drug use. Data will be collected only on women who are alcohol dependent. Comparison data for Caucasian women will be obtained from the Iowa Bone Health Study, a population based longitudinal study of bone mass which includes approximately 300 women between the ages of 20 to 40 years. Comparison data for African-American women will be obtained from a Michigan based study of bone mass in African-American women. These studies contain comparable physical, biochemical and interview data. Multiple regression will be the major analytic strategy employed in this phase. Phase 2 is a pilot study which will describe potential changes in bone density and bone turnover following abstinence from alcohol. Approximately 50 women will be followed longitudinally for 12 months post-treatment. Additionally 100 women in recovery and abstinent for at least 3 years will be recruited from local support groups and evaluated cross-sectionally. For this phase all women will complete the interview and physical measurements as described in phase l. Women completing the longitudinal component will serve as their own controls. Changes in bone density and bone turnover will be examined using the paired t test. For the cross- sectional evaluation, those women selected as controls for phase l will also serve as controls for this component. Multiple regression will be the major analytic strategy employed.
