Article Text

  1. D. Buscemi,
  2. A. Kumar,
  3. K. Nugent
  1. Texas Tech University Health Science Center, Lubbock, TX, and R Nugent University of Washington, Seattle, WA


Background Obesity has become a major health problem in the United States. The identification of potentially modifiable risk factors should lead to better treatment of this problem. Epidemiological studies have shown an association between shorter sleep times and increased body mass index (BMI). We used a survey instrument in our clinic patients to evaluate the association between sleep time and BMI and to identify potential confounding factors.

Methods We obtained a convenience sample of consecutive patients in internal medicine resident clinics. Two hundred patients agreed to participate and four refused. Patients were given a structured questionnaire asking about sleeping, eating, and lifestyle habits. We used descriptive statistics, logistic regression, and multivariate linear regression for data analysis. Unstable sleep was defined as having greater than a 2 hr range in to bed times or greater than a 2 hr range in wake-up times, or naps of length greater than 1 hr.

Results The final sample had 199 patients (74 men and 125 women). The median age was 54 with a range of 18 to 89. The mean BMI was 28.8 in men and 29.1 in women (p < .05 by t-test). Forty-one percent of the patients had a BMI $ 30. The mean sleep time was 7.7 hr in men and 8.0 in women (p > .05 by t-test). BMI was higher in younger patients (18-49) than older patients (> 50) (p < .05). The average BMI was higher in working patients than nonworking patients, and these working patients had a shorter sleep time (p < .05). Patients with unstable sleep patterns had a higher BMI (p > .05). Logistic regression demonstrated that the probability of obesity (BMI $ 30) increased in women (odds ratio [OR]: 2.7),in young patients (18-49 yrs, OR: 4.9), with short sleep periods (first quartile: 3-6.9 hrs, OR: 2.82), with dieting (OR: 7.03), with drinking alcohol (OR: 4.11), in diabetics (OR: 2.39), in hypertensive patients (OR: 2.39), and in sleep apnea (OR: 8.77). Subgroup analysis with logistic regression demonstrated that short sleep periods increased the probability of obesity in males and in stable sleepers but not in women or in unstable sleepers. Sleep period had a U-shaped effect in young patients (18-49) but not older patients.

Conclusion Adult BMI depends on genetic background, environmental influences, and acquired diseases. In this clinic population, shorter sleep time (3-6.9 hrs), age (18-49), and female gender increased the probability of obesity. Our subgroup analysis suggests that weight loss programs should focus on young males with short sleep times (less than 7 hours). In addition, our results suggest that sleep research studies need to separate patients with stable sleep patterns from patients with unstable patterns.

Statistics from

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.