Effects of ad libitum food intake, insufficient sleep and weekend recovery sleep on energy balance

Abstract

Study objectives: Insufficient sleep is believed to promote positive energy balance (EB) and weight gain. Increasing weekend sleep duration to "recover" from weekday sleep loss is common, yet little is known regarding how weekend recovery sleep influences EB. We conducted a randomized controlled trial to assess how: (1) 2 days and 8 days of insufficient sleep and (2) ad libitum weekend recovery sleep impact EB (energy intake [EI] - energy expenditure [EE]).

Methods: Following ten baseline days with 9 h per night sleep opportunities, participants completed one of three 10-day experimental protocols with ad libitum EI: control (9 h sleep opportunities; n = 8; 23 ± 5 years [mean ± SD]); sleep restriction (SR; 5 h sleep opportunities; n = 14; 25 ± 5 years); sleep restriction with weekend recovery sleep (SR + WR; 5 days insufficient sleep, 2 days ad libitum weekend recovery sleep, 3 days recurrent insufficient sleep; n = 14; 27 ± 4 years).

Results: Twenty-four hour EB increased (p < 0.001; main effect) by an average of 797.7 ± 96.7 (±SEM) kcal during the 10-day experimental protocol versus baseline with no significant differences between groups. Percent change from baseline in 24 h-EE was higher (p < 0.05) on day 2 of insufficient sleep (SR and SR + WR groups; 10 ± 1%) versus adequate sleep (control group; 4 ± 3%).

Conclusions: In this between-group study, the effects of adequate sleep and insufficient sleep, with or without or weekend recovery sleep, on 24 h-EB were similar. Examining EB and body weight changes using within-subject cross-over designs and "free-living" conditions outside the laboratory (e.g. sleep extension) are needed to advance our understanding of the links between insufficient sleep, weekend recovery sleep and weight-gain.

Keywords: catch-up sleep; circadian misalignment; interindividual variability; metabolic dysregulation; obesity; overeating; sleep deprivation; sleep loss; sleep restriction; timing of food intake.

Figure 1.

Experimental protocol. (A) Control group; (B) sleep restriction group; (C) sleep restriction with weekend recovery group. Underlines represent scheduled wakefulness and black boxes represent scheduled time in bed. Time of day is plotted as relative clock hour with scheduled wake time during baseline days arbitrarily assigned a value of 08:00 h and all other times referenced to this value (e.g. breakfast 2 h after scheduled waketime would be reported as occurring at a relative time of day of 10:00 h). Actual sleep timing was based on individual habitual sleep timing, so all participants were studied at their habitual circadian phase for baseline sleep assessments. Light exposure consisted of natural sunlight from a window in the room and room lighting at ~200 lux during scheduled wakefulness, except on energy expenditure assessment days. Energy expenditure was assessed on study days 3, 5, and 11 (labeled “C”) by whole room indirect calorimetry under dim light (<10 lux max) during scheduled wakefulness. Scheduled sleep occurred in darkness (0 lux). 24 h blood samples were collected on study days 3, 5, and 11 for assessment of leptin and ghrelin. For all groups, study days 1–3 served as the in-laboratory baseline segment, study days 4–8 served as the workweek-1 segment (WW-1), study days 9–10 served as the weekend segment (WE), and study day 11–13 served as the workweek-2 segment (WW-2). Meal timing for different study segments is presented on study days 1, 7, and 12 for the baseline, workweek-1, and workweek-2 study segments, respectively. B, breakfast; L, lunch; D, dinner; S, snack.

Figure 2.

24 h Energy intake, energy expenditure, and energy balance between groups. (A) Energy intake; (B) energy expenditure; (C) energy balance. CON, control group; SR, sleep restriction group; SR + WR, sleep restriction with weekend recovery sleep group; BL, baseline; WW-1, workweek-1; WE, weekend; WW-2, workweek-2. Solid blue and red line represents significant differences (p < 0.05) for workweek-1 (SR plus SR + WR groups combined) versus the CON group. Data are mean ± SEM. See panel “B” inset for raw 24 h energy expenditure data and Figure S1 for within group analyses of 24 h energy expenditure and energy balance.

Figure 3.

Hourly energy expenditure between groups. (A) Mean energy expenditure during the 4 h of extended wakefulness during insufficient sleep in the SR and SR + WR groups and the corresponding time during sleep in the CON group for workweek-1 and workweek-2; (B) baseline study segment; (C) workweek-1 study segment; (D) Workweek-2 study segment. Solid blue and red lines represent significant differences (p < 0.05) for workweek-1 (combined SR plus SR + WR groups combined) and workweek-2 versus the CON group. *P < 0.05 for workweek-1 (SR plus SR + WR groups combined) and workweek-2 (SR and SR + WR groups separately) versus the CON group. CON, control group; SR, sleep restriction group, SR + WR, sleep restriction with weekend recovery sleep group; WW-1, workweek-1; WW-2, workweek-2; Min, minute. Data are mean ± SEM.

Figure 4.

Mean and hourly leptin and ghrelin between groups. (A) Mean leptin by study segment; (B) baseline hourly leptin; (C) workweek-1 hourly leptin; (D) workweek-2 hourly leptin; (E) mean ghrelin by study segment; (F) baseline hourly ghrelin; (G) workweek-1 hourly ghrelin; (H) workweek-2 hourly ghrelin. CON, control group; SR, sleep restriction group; SR + WR, sleep restriction with weekend recovery sleep group; BL, baseline; WW-1, workweek-1; WW-2, workweek-2. Data are mean ± SEM. See Figure S2 for within group analyses of leptin and ghrelin.

Figure 5.

Individual differences in 24 h energy expenditure, 24 h energy intake, and 24 h energy balance. (A) Change from baseline in energy expenditure control group; (B) change from baseline in energy expenditure sleep restriction group; (C) change from baseline in energy expenditure sleep restriction with weekend recovery sleep group; (D) change from baseline in energy intake control group; (E) change from baseline in energy intake sleep restriction group; (F) change from baseline in energy intake sleep restriction with weekend recovery sleep group; (G) energy balance control group; (H) energy balance sleep restriction group; (I) energy balance sleep restriction with weekend recovery sleep group. Each column shows data for an individual participant. ICC’s show the consistency in outcomes during workweek-1 and workweek-2. Data points for energy balance at baseline are provided for descriptive purposes to illustrate changes from baseline. BL, baseline; WW-1, workweek-1; WW-2, workweek-2.

Figure 6.

Individual differences in pre-dinner and after-dinner snack energy intake. (A) Pre-dinner snack control group; (B) pre-dinner snack sleep restriction group; (C) pre-dinner snack sleep restriction with weekend recovery sleep group; (D) change from baseline in after-dinner snack EI control group; (E) change from baseline in after-dinner snack EI sleep restriction group; (F) change from baseline in after-dinner snack EI sleep restriction with weekend recovery sleep group; Each column shows data for an individual participant. ICC’s show the consistency in outcomes during workweek-1 and workweek-2. Data points for pre-dinner snack EI at baseline are provided for descriptive purposes to illustrate changes from baseline. BL, baseline; WW-1, workweek-1; WW-2, workweek-2; EI, energy intake.