P040 Association between sleep duration and macronutrient intake in people with Type 2 Diabetes

Abstract

Introduction Diet and sleep duration are both associated with Type 2 Diabetes (T2D). However, the longitudinal associations between macronutrient intake and sleep duration in people with T2D are unknown. We aimed to explore associations over 12 months in the Early-ACTID trial of usual care vs. a diet or diet+ physical activity interventions.1 Method Diet was assessed using 4-day estimated food diaries and average sleep duration in minutes was computed from self-reported usual sleep and wake times at baseline, 6- and 12-months post-intervention. Associations between percent total energy intake (%TEI) from fat, protein, carbohydrate and sleep duration were assessed using isoenergetic multiple linear regression substitution models, adjusting for TEI, and potential confounders. Results Median (IQR) sleep duration was 8 hrs (7.5, 8.9) at baseline. The% TEI from each macronutrient (fat/carbohydrate/protein) at 0, 6 and 12 months was 35/42/18%, 34/43/18% and 33/44/19%. Substituting 1% TEI from protein or fat with carbohydrate was associated with 2.6 minutes shorter sleep cross-sectionally at 12-months but not at baseline or 6 months (figure 1). Higher carbohydrate intake (in place of protein or fat) was associated with a decrease in sleep duration by 1.5–2 minutes from 0–6 and 6–12 months. All longitudinal and cross-sectional macronutrient specific substitution models (but 12 months cross-sectional) suggested replacing protein with carbohydrate results in more sleep duration reduction (about one minute) than replacing fat with carbohydrate. Discussion Lower protein and higher carbohydrate intake are consistently associated with shorter sleep duration. Replacing 5% protein with 5% carbohydrate results in 10–15 minutes less sleep duration 6 months later. Reference Andrews RC, Cooper AR, Montgomery AA, Norcross AJ, Peters TJ, Sharp DJ, et al. Diet or diet plus physical activity versus usual care in patients with newly diagnosed type 2 diabetes: the Early ACTID randomised controlled trial. The Lancet 2011;378(9786):129–39.