Does Subjective Sleep Affect Bone Mineral Density in Older People with Minimal Health Disorders? The PROOF Cohort

Abstract

Study objectives: Clinical and epidemiological studies suggest a relation between bone mineral density (BMD) and self-assessment of sleep with an effect on bone formation and osteoporosis (OS) risk in short and long sleepers. This study explores this association in a large sample of older subjects.

Methods: We examined 500 participants without insomnia complaints aged 65.7 ± 0.8 y. Each participant had a full evaluation including anthropometric measurement, clinical examination and measurements of BMD at the lumbar spine and femoral sites by dual-energy X-ray absorptiometry. The daily energy expenditure (DEE) was measured by the Population Physical Activity Questionnaire. Sleep duration and quality were evaluated by the Pittsburgh Sleep Quality Index. The subjects were stratified into three groups according to sleep duration, i.e., short (< 6 h), normal (6-8 h), and long (≥ 8 h) sleepers.

Results: Osteopenia was found in 40% of the subjects at the femoral level and 43% at the vertebral level. The prevalence of OS was lower both at femoral (8%) and vertebral (12%) levels. Short, normal, and long sleepers accounted for 29%, 40%, and 31% of subjects, respectively. After adjustments for metabolic, anthropometric, and DEE, multinomial logistic regression analysis indicated that long sleepers were more likely to have femoral neck OS with a slight effect of DEE at vertebral spine.

Conclusions: In a sample of older subjects, self-reported long sleep was the best predictor of OS risk at the femoral level. This finding suggests an association between OS and self-reported sleep duration in older subjects.

Clinical trial registration: NCT 00759304 and NCT 00766584.

Trial registration: ClinicalTrials.gov NCT00759304 NCT00766584.

Keywords: bone mineral density; older people; osteoporosis; sleep duration; sleep quality.

© 2016 American Academy of Sleep Medicine

Figures

Figure 1. Percentage of subjects with normal BMD and subjects with osteopenia or osteoporosis at the femoral and vertebral levels.

**p