TABLE 3.
Summary of sleep and weight status studies1
| Key observations (reference) | n | Age range | Female | Region |
| y | % | |||
| Inverse relations between sleep and weight status | ||||
| Less than 10 h sleep/day was associated with increased risk of obesity (101) | 1031 | 5–6 | 49 | France |
| Risk of obesity was significantly greater with fewer hours of sleep (78) | 383 | 11–16 | 54 | U.S. |
| Less than 7 h sleep/night was associated with increased likelihood of obesity compared with ≥7 h sleep/night (102) | 3682 | 32–49 | 68 | U.S. |
| Overweight and obese subjects obtained less sleep than normal-weight counterparts (103) | 924 | 18–91 | 65 | U.S. |
| BMI was highest for those with < 6 h sleep/night and lowest for those with ≥9 h sleep/night (104) | 990 | ≥18 | 52 | U.S. |
| Those sleeping 5–6 h/night weighed more than those sleeping ≥7 hours/night (105) | 68,183 | 39–65 | 100 | U.S. |
| 8–10 hours of sleep/night was associated with significantly higher odds of obesity than 12–13 h sleep/night (109) | 422 | 5–10 | 50 | Canada |
| BMI was significantly higher with 5–6 h sleep/night compared with 7–8 h sleep/night (106) | 276 | 21–64 | 53 | Canada |
| Seven or fewer hours of sleep/night significantly increased the risk of obesity compared with >7 h sleep/night (80) | 5877 | 10–19 | 45 | Saudi Arabia |
| Less than 9 to 9.4 h sleep/night was associated with increased likelihood of obesity compared with ≥11 h sleep/night (111) | 1311 | 3–4 | 50 | China |
| Less than 6 h sleep/night and sleep fragmentation were both independently associated with higher BMI values (100) | 612 | 18–50 | 58 | U.S. |
| Short sleep duration was associated with increased weight status (112) | 1108 | 5–13 | 50 | U.S. |
| Compared with >11 h sleep/d, <9 h and 10–11 h sleep/d were associated with increased risk of obesity in a dose-dependent fashion (113) | 7867 | 2–9 | 49 | Europe |
| As total sleep time decreased, BMI increased (116) | 1231 | 6–10 | 50 | Sweden |
| Less than 6 h/night of time in bed was associated with increased BMI compared with 7–8 h/night of time in bed (107) | 250 | 18–25 | 66 | U.S. |
| Less than 6 h sleep/night was associated with increased BMI compared with 8–9 h sleep/night in adults in the sub-Arctic (108) | 6413 | 30–65 | 52 | Norway |
| Less than 7 h sleep/night was associated with increased BMI compared with 7–8 h sleep/night (124) | 34,852 | 30–60 | 10 | Japan |
| Influence of gender on sleep-weight status | ||||
| Inverse/males: short sleep duration was a significant predictor of overweight in males only (119) | 4486 | 13–18 | 51 | U.S. |
| Inverse /males: In males, <8–10 h sleep/night was associated with significant higher risk of obesity compared with ≥10 h sleep/night; no associations were observed in females (120) | 6324 | 7–15 | 49 | Australia |
| Inverse /males: a significant dose-dependent, inverse relation was observed between sleep duration and BMI in males but not in females (121) | 4793 | 17–83 | 51 | Hong Kong |
| Inverse /females: <8 h sleep/night was associated with higher values of BMI, total body fat, truncal fat, waist circumference, and hip circumference along with lower values of % LBM in females only (130) | 500 twins | 10–20 | 45 | China |
| Inverse /females: significant inverse correlation between sleep and BMI was observed in women, whereas a negative trend was observed in men (126) | 3473 | 33–45 | 54 | U.S. |
| U-shaped/males: <5 h and ≥9 h sleep/night was associated with increased BMI gain over 1 y in males; no longitudinal association was observed in females (124) | 34,852 | 30–60 | 10 | Japan |
| U-shaped/females: Compared with 7 h sleep/night, ≤4 h and ≥9 h sleep/night was associated with increased likelihood of obesity in females only; no association observed among males (131) | 23,579 | 14–18 | 51 | U.S. |
| Inverse /males: an inverse relation between sleep and BMI was observed in males but not in females (122) | 2006 | 26–28 | 52 | U.S. |
| Inverse /males: as sleep duration decreased, BMI, waist circumference, and subcutaneous fat area increased in males only; no associations were observed in females (123) | 6042 | 30–75 | 11 | Japan |
| Longitudinal observations of associations between sleep and weight status | ||||
| Inverse: the highest increase in BMI over 8–10 y was associated with 2–4 h sleep/night whereas the smallest increase in BMI was associated with ≥10 h sleep/night (8–10 y study duration) (102) | 3682 | 32–49 | 68 | U.S. |
| Inverse: <7 h sleep/night was associated with significantly higher increases in weight over a 16-y period of time (∼0.3–0.5 kg additional gain/y) (16-y study duration) (105) | 68,183 | 39–65 | 100 | U.S. |
| Inverse: irregular sleep habits at ages 2–4 y were associated with increased BMI at 21 y (21-y study duration) (117) | 2494 | 0–1 | 50 | Australia |
| Inverse: shorter sleep duration at 8 y of age was associated with overweight status at 11 y of age (3-y study duration) (110) | 785 | 8 | 50 | U.S. |
| U-shaped association: 5–6 and 9–10 h sleep/night were both associated with higher amounts of weight gain over 6 y compared with 7–8 h sleep/night (6-y study duration) (106) | 276 | 21–64 | 53 | Canada |
| Inverse: short-duration sleep at 5–11 y of age was associated with increased BMI at 32 y of age (32-y study duration) (118) | 972 | 0–1 | Not reported | New Zealand |
| Inverse: <12 h sleep/day was associated with increased BMI, skinfold thickness measures, and odds of overweight (3-y study duration) (133) | 915 | 0–1 | 50 | U.S. |
| No association: no longitudinal associations between sleep and BMI observed (15–20-y study duration) (100) | 612 | 18–30 | 58 | U.S. |
| Inverse for 0- to 4.9-y-olds: short sleep duration was associated with increased BMI 5 y later; no association for 5- to 13-y-olds (5-y study duration) (112) | 1930 | 0–13 | 50 | U.S. |
| U-shaped association for <40-y-olds: ≤5 and ≥8 h sleep/night was associated with elevated gains in adipose tissue and BMI over 6 y (6-y study duration) (134) | 522 | 18–81 | 58 | U.S. |
| No association: sleep duration earlier in life did not predict BMI 2 y later (2-y study duration) (114) | 3045 | 0–1 and 4–5 | 49 | Australia |
| Inverse: short-duration sleepers experienced greater increases in BMI and fat mass over 6 y compared with those sleeping 7–8 h/night; correction of short-duration sleep attenuated this change (6-y study duration) (84) | 216 | 18–64 | 50 | Canada |
| Influence of age on sleep-weight status | ||||
| Inverse for 11-y-olds; no association for 8-y-olds (3-y longitudinal study) (110) | 785 | Third- to sixth-graders | 50 | U.S. |
| Inverse for 6- to 7-y-olds; no association for 0- to 1-, 2- to 3-, or 4- to 5-y-olds (114) | 3045 | Birth–7 | 49 | Australia |
| Inverse for 10- to 13-y-olds; no association for 14- to 17-y-olds (115) | 723 | 10–16 | 52 | U.S. |
| Inverse for 7- to 11-y-olds; no association for 32-y-olds (118) | 1037 | 5–11 baseline, 32 at follow-up | 48 | New Zealand |
| Inverse for 32- to 49-y-olds; no association for 50- to 67-y-olds; no association for 68- to 86-y-olds (102) | 9588 cross-sectional; 15,054 longitudinal | 32–49 at baseline | 68 | U.S. |
| Associations between sleep and energy balance | ||||
| Inverse relation/intake: sleep <8 h/d associated with increased intake of high-fat and -sugar foods; no association: no associations between sleep and energy output or balance (152) | 1976 | 5–10 | 50 | Portugal |
| Inverse relation/intake: reduced quality of sleep associated with higher total fat intake (132) | 300 | 10–17 | 49 | U.S. |
| Inverse relation/intake: longer sleep associated with lower kcal intake (83) | 41 | 2–5 | 58 | U.S. |
| Positive association/PA: each hour of sleep disturbance decreased daily PA by 3% (78) | 383 | 11–16 | 54 | U.S. |
| Positive association/expenditure: <9 h/night of sleep associated with increased sedentary time; no association/PA: no association between sleep and moderate/vigorous PA (173) | 519 | 7 | 51 | New Zealand |
| Inverse relation/intake: short sleep duration and late timing of sleep were associated with increased consumption of calories and fast foods (155) | 52 | 18–71 | 48 | U.S. |
| No association/expenditure: no difference in RMR, TEE, PA, or diet-induced thermogenesis between 2 wk of 8.5 h sleep/night and 5.5 h sleep/night (68) | 11 | 35–49 | 45 | U.S. |
| Inverse relation/intake: <8 h/night of sleep was associated with increased kcal intake from fat and snacks (153) | 240 | 16–19 | 52 | U.S. |
| Inverse relation/intake: 1 night of sleep deprivation was associated with increased kcal intake; 1 night of sleep deprivation was not associated with hunger changes; inverse relation/PA: 1 night of sleep deprivation was associated with increased PA and activity energy expenditure (156) | 12 | 19–25 | 0 | France |
| Inverse association/intake and expenditure: 5 h sleep/night (vs. 9 h) for 5 nights was associated with increased TEE and kcal intake (158) | 16 | 17–27 | 50 | U.S. |
| Inverse relation/intake: 4 h sleep/night (vs. 7–9 h) for 4 nights was associated with increased fat and kcal intake; no association/expenditure: no differences in RMR and TEE were observed between conditions (157) | 30 adults | 30–49 | 50 | U.S. |
| Inverse relation/intake: ≤6 h of sleep/night was associated with increased kcal intake from alcohol (84) | 703 | 18–64 | 57 | Quebec, Canada |
| Inverse relation/intake: shorter sleepers (≤6 h/night) had higher eating disinhibition and greater risk of overeating and weight gain (162) | 276 | 21–64 | 58 | Quebec, Canada |
| Inverse relation/ghrelin: <8 h sleep/night was associated with increased ghrelin concentrations; positive association/leptin: <8 h sleep/night was associated with decreased leptin concentrations (128) | 1024 | 30–60 | 46 | U.S. |
| No association/intake and PA: no differences in kcal intake or PA were observed with 4 h of sleep/night (vs. 8 and 12 h) for 6 nights; positive association/leptin: 4 h sleep/night was associated with reduced leptin concentrations (147) | 11 | 18–27 | 0 | Belgium |
| Inverse relation/hunger: 4 h sleep/night (vs. 10 h) for 2 d was associated with decreased leptin concentrations, increased ghrelin concentrations, and increased ratings of hunger; intake and PA were held constant (164) | 12 | 20–24 | 0 | U.S. |
| Inverse relation/intake and leptin: gradually decreasing sleep duration from 7 to 4 h of sleep/night (vs. >8 h) for 4 nights was associated with increased kcal intake, body weight, and concentrations of leptin and thyroid hormones; no association/expenditure: no differences were observed in RMR, PA, and TEE (165) | 14 | 23–38 | 100 | Germany |
| Inverse relation/leptin: 3 h sleep for 1 night elevated next-day leptin concentrations but did not affect hunger and craving scores; kcal intake was not measured (166) | 15 | 18–25 | 100 | Canada |
| Inverse relation/leptin: 1 night of total sleep deprivation increased leptin concentrations but did not affect hunger (167) | 21 | 18–30 | 52 | U.S. |
| Inverse relation/leptin: 4 h sleep/night for 5 nights was associated with increased leptin concentrations (168) | 136 | 22–45 | 49 | U.S. |
| Positive association/expenditure: 1 night of total sleep deprivation decreased RMR and diet-induced thermogenesis (170) | 14 | 21–23 | 0 | Germany |
| Positive association/PA: 4.25 h sleep/night (vs. 8.25 h) for 2 nights was associated with decreased PA and PA intensity; no association/intake: sleep restriction was not associated with changes in kcal intake, feelings of hunger, appetite, or ghrelin or leptin concentrations (171) | 15 | 20–40 | 0 | Germany |
| Positive association/ghrelin: 1 night of total sleep deprivation was associated with lower concentrations of ghrelin (172) | 10 | 25–31 | 0 | Germany |
| No association/PA: no relation between self-reported sleep duration and PA observed (105) | 68,183 (NHANES) | 39–65 | 100 | U.S. |
| Positive association/PA: <7 h sleep/night was associated with not meeting vigorous PA requirements (174) | 1203 | 20–92 | 76 | U.S. |
| Positive association/RMR: 2 wk of a kcal-restricted diet plus 5.5 h sleep/night (vs. 8.5 h) was associated with reduced RMR; no association/TEE: sleep and kcal restriction were not associated with changes in leptin, ghrelin, or TEE (175) | 10 overweight | 35–49 | 30 | U.S. |
| Inverse relation/PA: 2 nights of fragmented sleep was associated with increased PA and activity energy expenditure; no association/TEE and RMR: fragmented sleep not associated with changes in TEE or RMR (176) | 15 | 20–27 | 0 | U.S. |
| No association/RMR: 5 h sleep/night (vs. 10 h) for 7 nights was not associated with changes in RMR (5) | 20 | 20–35 | 0 | U.S. |
1
LBM, lean body mass; PA, physical activity; RMR, resting metabolic rate; TEE, total energy expenditure.