2006 OPEN FORUM Abstracts
Smoking and Bone Mineral Density in Non-Hispanic White Women
Michael E. Anders, PhD,
Lori W. Turner, PhD, RD, Lorraine S. Wallace, PhD, FACSM, Horace J. Spencer III, MS,* and Donald D. Simpson, MPH*
*University of Arkansas for Medical Sciences, Little Rock, Arkansas
University of Arkansas, Fayetteville, Arkansas
University of Tennessee Graduate School of Medicine, Knoxville, Tennessee
Background. Osteoporosis is a devastating chronic disease that causes skeletal fragility. The occurrence of disability following hip fracture often leads to a profound forfeiture of independence. Sequelae of hip fractures, such as pneumonia and pulmonary embolism, are frequently lethal. The purpose of this study was to determine the correlation of smoking with a bone mineral density (BMD) ≤ the National Osteoporosis Foundation (NOF) treatment threshold in non-Hispanic white women in accordance with the World Health Organization reference criteria.
Methods. Data from a nationally representative sample from the Third National Health and Nutrition Examination Survey (NHANES III) were extracted for non-Hispanic white women aged 50 through 90 years who completed surveys, laboratory testing, and measurement of BMD via dual energy x-ray absorptiometry. A NHANES III quality control program reviewed each BMD test and monitored instrument stability and technologist performance and education. Random blind replications of BMD tests assessed reliability. Multivariable logistic regressions tested whether independent variables increased the odds of a BMD ≤ the NOF treatment threshold versus the odds a BMD > the NOF treatment threshold. First, univariable regressions screened the effect of each independent variable, including age, weight, weight-loss, smoking status, weight-bearing physical activity, calcium intake, alcohol intake, chronic disease status, and serum vitamins A, C, and E, on a BMD ≤ the NOF treatment. Second, those variables with p values ≤ 0.25 in the univariable analyses were tested in multivariable regression full models. Third, independent variables with a p value < 0.01 in the full model were tested at the 0.01 level in multivariable regression reduced models. Fourth, each variable excluded from the full model were then re-tested at the 0.01 level in the reduced model to identify variables that might have affect BMD in combination with other variables.
Results. The study sample included 1,719 non-Hispanic white women; 703 (37.4%) had a BMD ≤ the NOF treatment threshold, and 500 (26.6%) were current smokers.
Reduced Model Logistic Regression
|n = 1,719||ß||S.E.||Wald||p||Odds Ratio (95% CI)|
|Age (per 10 years)||.644||.063||104.93||< .001||1.90 (1.68, 2.15)|
|Weight (per 10 lbs.)||-.390||.026||219.03||< .001||.68 (.64, .71)|
|Lack of estrogen therapy||.710||.180||15.59||< .001||2.03 (1.43, 2.89)|
|Current smoker||.382||.135||7.98||.005||1.47 (1.12, 1.91)|
study results provided sufficient evidence to suggest that for a BMD ≤
the NOF treatment threshold: (a) increased
age, lack of estrogen therapy, and current smoking were risk factors and (b) increased
weight was a protective factor.