Abstract P344: Sleep Duration and Obesity Impact of Demographics, Socioeconomic Status, Health Behaviors, and Health Status

    loading  Checking for direct PDF access through Ovid


Introduction: Many studies have shown that sleep duration is associated with obesity. It is unclear, though, whether this relationship exists equally across groups and whether it depends on demographics, socioeconomics, or aspects of health.Methods: Nationally-representative data from the 2016 BRFSS was used. Obesity was calculated as BMI≥30. Sleep duration was categorized as very short (≤4), short (5-6), normal (7-8), and long (≥9). Covariates included demographics (age, sex, race/ethnicity, education, marital status), socioeconomics (education, income, employment, # children), health risk factors (smoking, heavy drinking, sedentary lifestyle, access to a doctor, foregoing medical care due to cost), and health status (physical health, mental health, health-related limitations). Weighted logistic regression examined 5 models (unadjusted, demographics, add socioeconomics, add health behaviors, add health status). Whether relationships depended on covariates were evaluated with interaction terms and followed up by stratified analyses.Results: See Table for associations between sleep duration and obesity across all 5 models. In all models, very short, short, and long sleep were all associated with obesity, with very short sleep carrying the greatest risk. Note that as the number of covariates increased, the analytic samples were smaller. Interaction terms for all variables were statistically significant (p<0.001). Very short and short sleep effects were strongest in the youngest adults. Relationships were stronger in women. Sedentary individuals, heavy drinkers, and smokers demonstrated a weaker relationship. Lack of care was associated with a stronger relationship.Conclusions: Both short and long sleep are associated with obesity, even after accounting for many covariates. However, this relationship depends on factors such as age, sex, race/ethnicity, socioeconomic status, and health. This will help towards understanding risk and targeting interventions.

    loading  Loading Related Articles