INTRODUCTION

Sleep duration has been associated with overweight individuals in many epidemiological studies; however, few studies have assessed sleep using objective methods. Our study was designed to evaluate the association between body mass index (BMI) and sleep duration measured by actigraphy (Acti), polysomnography (PSG) and the Pittsburgh sleep quality index questionnaire (PSQIO). Furthermore, we evaluated other biochemical and polysomnographic parameters.

METHODS

A representative sample of 1042 individuals from Sao Paulo, Brazil, including both genders (20-80 yrs), participated in our protocol. Weight and other anthropometric parameters were measured at the onset of the study. Sleep duration was calculated by Acti, PSG, and the PSQIQ. The population was sorted by sleep duration, body, slow wave sleep (SWS) and REM sleep (REMS) duration subsets. In addition, other biochemical and polysomnographic parameters were analyzed. Differences between population subsets were analyzed by one-way analysis of variance (ANOVA). Linear regression analysis was performed between sleep and anthropometric parameters.

RESULTS

Shorter sleep duration was associated with higher BMI and waist and neck circumference when measured by Acti and PSG (p<0.05). Lower leptin levels were associated with short sleep in normal-weight (BMI>18 and ⩽25) individuals (p<0.01). The association between short sleep duration Acti and higher BMI was present when apnea-hypopnea index (AHI) was less than 15 (p=0.049). Shorter REMS and SWS also were associated with higher BMI (p<0.01). Normal-weight individuals tended to sleep longer, have higher sleep efficiency and longer SWS and REMS than obese individuals (Acti, PSG; p=0.05). Sleep duration was negatively correlated with BMI (Acti, PSG; p<0.05). Short SWS and REMS were associated with higher cardiovascular risk factors (p<0.05).

CONCLUSION

Shorter sleep, SWS, and REMS duration were associated with higher BMI, central adiposity measurements, and cardiovascular risk factors when measured by objective methods.