The Relationship Between Sleep Quantity, Sleep Quality, and Weight Loss in Adults: A Scoping Review

Abstract

Sleep is hypothesized to interact with weight gain and loss; however, modeling this relationship remains elusive. Poor sleep perpetuates a cascade of cardiovascular and metabolic consequences that may not only increase risk of adiposity, but also confound weight loss efforts. We conducted a scoping review to assess the research on sleep and weight loss interventions. We searched six databases for studies of behavioral weight loss interventions that included assessments of sleep in the general, non-clinical adult human population. Our synthesis focused on dimensions of Population, Intervention, Control, and Outcomes (PICO) to identify research and knowledge gaps. We identified 35 studies that fell into one of four categories: a) sleep at baseline as a predictor of subsequent weight loss during an intervention, b) sleep assessments after a history of successful weight loss, c) concomitant changes in sleep associated with weight loss, and d) experimental manipulation of sleep and resulting weight loss. There was some evidence of improvements in sleep in response to weight-loss interventions; however, randomized controlled trials of weight loss interventions tended not to report improvements in sleep when compared to controls. We conclude that baseline sleep characteristics may predict weight loss in studies of dietary interventions and that sleep does not improve because of weight loss alone. Future studies should enroll large and diverse, normal, overweight, and obese short sleepers in trials to assess the efficacy of sleep as a behavioral weight loss treatment.

INTRODUCTION

Obesity is a serious and costly chronic disease that continues to be a major public health concern. The World Health Organization (WHO) estimates that obesity affected 13% of the global population in 2016, representing nearly 650 million individuals.¹ The rates of obesity are considerably higher in the United States. According to the Centers for Disease Control and Prevention (CDC), nearly one-third of adults, representing almost 100 million Americans, meet the body mass index cut point for obesity.² Recent reports by the CDC show that the number of states where at least one-third of the population is obese has nearly doubled since 2018.³ The prevalence of obesity is expected to increase; projections indicate that nearly one-in-two Americans will meet the BMI classification for obesity by 2030, with even higher rates expected for women, African Americans, low-income adults, and for those in particular geographic regions.⁴

Given the large and increasing population affected, considerable attention has been placed on intervention efforts that focus on preventing and treating obesity. Substantial weight loss is possible across a range of treatment modalities, but long-term weight loss maintenance is much more challenging, and weight regain is typical. In a meta-analysis of 29 long-term weight loss studies, over half of the lost weight was regained within two years, and by five years more than 80% of lost weight was regained.⁵ Additionally, Fildes et al.⁶ found that nonsurgical obesity treatment strategies were failing to achieve sustained weight loss for a large majority of obese patients. Results from the study provide evidence that maintaining weight loss was rare and the probability of achieving normal weight was low for those with a BMI of 30 and greater.⁶

Short sleep has been associated with numerous chronic conditions including a constellation of cardiometabolic risk factors, inflammation, decreased psychomotor vigilance, and unintentional injuries.⁷ Data from the National Health Interview Survey (NHIS) found self-reported sleep duration decreased 10 to 15 minutes from 1985 to 2012, and the percentage of adults reporting less than 6 hours of sleep per night has increased from 22.3% to 29.2%.⁸ Epidemiological surveillance conducted by the Centers for Disease Control and Prevention (CDC), found short sleep duration was reported in over 35% of all adults in the United States.⁹ The prevalence of short sleep duration has paralleled the obesity epidemic. Both conditions affect approximately 33% of the US population lending credence to the suggestion that short sleep duration and obesity are associated phenomena¹⁰; nevertheless, this proposed mechanism linking short sleep and obesity remains elusive.

We conducted a scoping review to examine the relationship between sleep and weight loss, both summarizing sleep as a predictor of weight loss, as well as concurrent changes in sleep associated with weight loss. This scoping review employed a methodology to map, synthesize, and summarize the available data and make recommendations for future research based on the literature we reviewed. The main objectives of this scoping review were to: a) review of the published research on sleep and weight loss interventions; b) identify the characteristics and range of methodologies used in the peer-reviewed sleep and weight loss research; c) examine reported challenges and limitations of the stream of intervention; and d) propose recommendations for further research on sleep interventions and weight loss. The two main research questions for this review were:

1. How does sleep quality and quantity change in relation to weight loss among adults?

2. What is known from the existing literature about the relationship between sleep and weight loss in adults?

METHODS

We organized the review using the framework for conducting scoping reviews by Arksey and O’Malley¹¹ and the scoping review aims developed by Munn et al.¹² The steps included: a) defining the research question, b) identifying relevant studies, c) study selection, 4) charting the data, and 5) collating, summarizing, and reporting results. The protocol for our scoping review was registered with Open Science Framework (OSF) Registries.¹³ As no human subjects were enrolled in our review, we did not seek Institutional Review Board (IRB) approval for this study. We delimited our search to the following six databases: PubMed, MEDLINE, PsycInfo, Web of Science, Cochrane Central Register of Controlled Trials (CENTRAL), and Cumulative Index to Nursing & Allied Health Literature (CINHAL). We then developed our database search strategies using subject headings and title/abstract keywords related to sleep, weight loss, sleep quality, and sleep quantity. We conducted our searches in each database during May 2023. Once our search strategy was in place, we set up an RSS alert system to identify new articles pertinent to our search. We limited our search retrieval to peer-reviewed, completed research studies published in English between January 2006 and May 2023, in non-clinical, adult populations. As we sought to understand the role of sleep and weight loss in the general adult population, we excluded animal research; studies focused on clinical populations (e.g., diabetes, cancer, and/or fibromyalgia), including studies which enrolled subjects with clinically diagnosed sleep disorders (e.g., sleep apnea or other sleep disorders); and outpatient subjects that underwent surgical or pharmacological weight loss interventions (e.g., bariatric surgery). From a clinical significance perspective, weight loss interventions that result in 5–10% reduction¹⁴ from baseline weight and sleep interventions that shorten sleep onset latency and/or increase total sleep time by 30 minutes¹⁵ are considered efficacious. For the purposes of our review, we did not delimit our search to these weight loss or sleep outcomes due to the novelty of sleep as a potentially modifiable factor related to obesity.

We exported records from each database into a master Microsoft Excel© file. Two authors (AK & MO) independently screened each record across all four phases of data extraction. Disagreements were resolved by a third reviewer (JP). Titles of each record were reviewed during stage 1. Duplicates, review articles, study protocols, cross-sectional and formative research, as well as interventions which measured weight loss and sleep in clinical populations were removed during the stage 1 data extraction process. The remaining records were imported into EndNote© bibliographical software to review abstracts (stage 2) and full text (stage 3) of each article, resulting in the further removal of observational studies, interventions which did not target weight loss or measure sleep, and clinical populations. Finally, a descendent search (stage 4) was conducted using all articles extracted during stage 3 articles. Table 1 summarizes the databases and the disposition of the search strategy through all four stages of data extraction. In total, we included 35 studies published in the English language that focused on weight loss interventions conducted with adult human subjects where sleep was one variable investigated in the study.

Table 1.

Summary Table of the Data Extraction Process for the Sleep Quantity, Sleep Quality, and Weight Loss in Adults Scoping Review

Review Information			Distillation Process
Name of Study	The Relationship Between Sleep Quantity, Sleep Quality, and Weight Loss in Adults: A Scoping Review	Phase 1	PubMed (n=69)	Medline (n=135)	PsychInfo (n=44)	Web of Science (n=398)	CENTRAL (n=98)	CINAHL (n=70)		Phase 1 Total (n=898)
Boolean Search Logic/Key Terms	((overweight OR obes* OR adipos*)) AND ((sleep quantity* OR quality of sleep)) AND ((random* OR clinic* OR trial OR intervention OR evaluation OR experiment OR program* OR pilot OR feasibility OR treatment)) AND ((weight loss OR weight maint* OR weight reduc* OR weight management OR weight control OR obesity treat* OR obesity prev* OR overweight treat* OR overweight prev*)) NOT (surgical OR pharmacological weight loss) NOT (sleep disorder* OR sleep disturbance* OR sleep problem* OR insomnia) NOT (cardiovascular disease OR cardiac disease OR coronary heart disease OR cardiomyopathy) NOT (diabetes type 2 OR diabetes mellitus OR hyperglycemia) NOT (surgical OR pharmacological weight loss) NOT ((rats OR mice)) NOT ((child*))	Phase 2 Total (n=746)	Duplicate Articles (n=41)	Review Articles (n=21)	Study protocols/formative research (n=87)	Cross-sectional and observational studies (n=354)	Clinical populations (n=143)	Weight loss not targeted (n=63)	Sleep not measured (n=37)	Remaining Post-Phase 2 (n=152)
Registry	OSF Registries: The Relationship Between Sleep Quantity, Sleep Quality, and Weight Loss in Adults13	Phase 3 Total (n=118)	Cross-sectional and observational studies (n=42)	Clinical populations (n=39)	Weight loss not targeted (n=26)	Sleep not measured (n=11)				Remaining Post-Phase 3 (n=34)
		Phase 4	Descendant search (n=1)							Total in review (n=35)

RESULTS

Table 2 presents a summary description of the sample demographics and sampling location, primary study purpose, study design, measures collected, intervention assignments, variables assessed, behavior change frameworks, as well as intervention effects, baseline sleep measures, and quality assessment scores for the 35 articles included in this scoping review.

Table 2.

Summary of the Studies Investigating the Relationship Between Sleep Quantity, Sleep Quality, and Weight Loss in Adults (n=35)

#	Study Name* / Sample Demographics / Sampling Location / Study ID†	Primary Study Purpose / Study Design / Measurements / Intervention Assignments	Outcome Variable(s) / Sleep Variables / Behavior Change Theoretical Framework	Intervention Effects / Baseline Sleep Measures / Quality Assessment Scores
1.	PATH Study 16 Littman et al., 2007 Overweight and obese (Mean [M] Body Mass Index [BMI] of treatment [Tx] group=30.4±4.1 kg/m2, M BMI of active control [ACnt] group=30.5±3.7 kg/m2), sedentary, post-menopausal women (n=173) M years of age of Tx=60.7±6.7; M years of age of ACnt=60.6±6.8 (100% female sample) Community sample recruited from Seattle, Washington, United States	Assess relationships among exercise, sleep, ghrelin, and leptin Randomized control trial Secondary analysis of cross-sectional and longitudinal trial data Data were collected at baseline, 3, and 12 months Treatment [Tx]: Exercise intervention (n=87) Active control [ACnt]: Stretching intervention (n=86)	Weight loss Sleep quality and quantity (Women’s Health Initiative Insomnia Rating Scale Questionnaire17,18) BMI (kg/m2) Biomarkers (Ghrelin, leptin) Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	No consistent associations between self-reported sleep measures and ghrelin or leptin A weak and nonsignificant positive association between sleep duration or napping and BMI Changes in sleep were not strongly associated with changes in weight, ghrelin, and leptin Observed trends were typically in the direction opposite of the study hypotheses (e.g., the weight-loss difference between Tx and ACnt was greater for those who slept less at follow-up than at baseline compared to those whose sleep duration did not change) 19.5% of Tx participants and 24.4% of ACnt self-reported short sleep 6.9% of Tx participants and 9.3% of ACnt self-reported poor sleep quality
2.	GH-IGF-I & Sleep 19 Rasmussen et al., 2008 Obese (M BMI=41±1 kg/m2 and non-obese (M BMI= 23±1 kg/m2) adults (n=16) M years of age of Tx=32±2 (5 females, 1 male); mean years of age of Cnt=27±2 (8 females, 2 males) Participants recruited from an outpatient hospital clinic for weight reduction in the Copenhagen, Denmark region	To investigate the interrelationships and the impact of weight loss on sleep in severely obese participants Quasi-experimental design Tx bodyweight was stable for 6 weeks prior to 24-hour polysomnographic assessments Tx: Weight loss intervention (n=6) Control [Cnt]: Non-obese controls (n=10)	Weight loss Anthropometrics (Body weight, waist/hip circumferences) BMI (kg/m2) Sleep duration (Polysomnography) Biomarkers (Growth hormone, IGF-1, leptin, cortisol, insulin) Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	Tx mean weight loss of 36±7 kg Total sleep duration was significantly lower in Tx compared to Cnt (p<.01) Following weight loss, Tx still had significantly shorter sleep duration compared to Cnt (p<.01) The amount of arousal periods, the duration of REM and non-REM periods were similar between Tx and Cnt and was unaffected by diet-induced weight loss Sleep duration correlated with total IGF-I in all participants (r=0.57; p<.01) Sleep duration, 24-h GH, and IGF-I levels decreased while 24-h leptin levels increased in obese subjects Baseline sleep duration: Tx=360±17, Cnt=448±15
3.	Sleep, Energy Metabolism and Diabetes Risk 20 Nedeltcheva et al., 2010 Overweight and obese (M BMI=27.4±2.0 kg/m2) adults (n=10) Mean years of age=41±5.0 (3 females, 7 males) Community sample recruited in Chicago, Illinois, United States and assessed at university clinic research center and sleep laboratory ID: NCT00720889	To determine whether experimental sleep restriction attenuates the effect of reduced-calorie diet on excess adiposity Randomized two-period two-condition crossover design Data collected at end of each 14-day intervention, with mean of 7±3 months washout period between interventions Tx: 8.5 hours, time-in-bed ACnt: 5.5 hours, time-in-bed	Weight loss Fat loss and fat-free mass (Body weight subtracted from body fat using dual X-ray absorptiometry) Sleep (Electroencephalography) Sleep quality (Pittsburgh Sleep Quality Index [PSQI]18) Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	Sleep curtailment decreased the fraction of weight lost as fat by 55% (1.4 vs. 0.6 kg with 8.5 vs. 5.5-h sleep opportunity, p=0.043) Sleep curtailment increased the loss of fat-free body mass by 60% (1.5 vs. 2.4 kg, p=.002) Sleep duration at baseline of all participants: 7.7±0.7 PSQI baseline score of all participants:3.0±2.0
4.	Milk Supplementation & Energy Balance 21 Chaput et al., 2012 Overweight and obese (M BMI=33.2±3.6 kg/m2) adults (n=123) M years of age=41.1±6.0 (57% men, 43% women) Community sample in Quebec, Canada ID: NCT00729170	Evaluate whether sleep predicts the magnitude of fat loss in adults subjected to moderate caloric restriction One-group, pretest posttest design Secondary analysis of control participants pooled together from three weight loss studies Data collected at baseline, 15 and 24-weeks (depending on study) Tx: Moderate calorie restriction (n=123)	Weight loss 24-hour food recalls22 BMI (kg/m2) Sleep quality (PSQI18) Self-reported sleep duration Weight loss behavior change theoretical framework: Behavioral counseling Sleep behavior change theoretical framework: None reported	Significant positive relationship between sleep duration and the loss of body fat, in both unadjusted and adjusted models (adjusted β=0.72; p<.05) Sleep quality was positively associated with both fat mass loss and percent body fat loss in unadjusted and unadjusted models, suggesting that a better sleep quality was associated with greater fat mass loss (adjusted β=−.19; p<.05; adjusted β =−0.21; p<0.01, respectively) 1-hour increase in sleep duration for every 0.7 kg decrease in fat mass, after adjustment for covariates, with sleep entered as outcome variable into the model PSQI baseline score of all participants: M=4.8±2.9 Sleep duration of all participants: 7.1±1.0 hours
5.	LIFE Study 23 Elder et al., 2012 Obese (M BMI=37.7±5.2 kg/m2) adults (n=472) M years of age=55.3±11.7 (17% men, 83% women) Sample recruited from a medical care organization in Portland, Washington, United States ID: NCT00526565	Phase I analysis of behavioral weight loss intervention using calorie restriction One-group, pretest posttest design Phase 1: Data collected at baseline and 6 months Tx: Behavioral weight loss	Weight loss BMI (kg/m2) Insomnia (Insomnia Severity Index)24 Stress (Perceived Stress Scale25) Self-reported sleep duration (range of hours) Patient Health Questionnaire-8 (PHQ-8)26 Weight loss behavior change theoretical framework: Social cognitive theory,27 transtheoretical model,28 behavioral self-management,27 motivational enhancement27 Sleep behavior change theoretical framework: None reported	Tx mean weight loss of 6.3±7.1 kg Participants self-reporting >6 and ≤8 hours of sleep were more likely to move to Phase 2 of the weight loss trial Neither sleep duration nor insomnia were baseline predictors of weight loss ISI baseline score of all participants: M=7.0±5.4 Sleep duration baseline: 48% between 7 to 8 hours
6.	Better Weight-Better Sleep 29 Logue et al., 2012 Overweight and obese (% overweight=21.7, % obese=78.2%) adults (n=46) 87.0% female, 13.0% men (50% Caucasian, 50% African American) Sample recruited from a primary care setting in Ohio, United States	To explore the feasibility of integrating sleep management interventions with dietary and exercise interventions for obesity Randomized controlled trial Data collected at baseline, 6, and 12 weeks Tx: Better Weight-Better Sleep (n=23) StCnt: Better Weight (n=23)	Percent weight change Stress (Perceived Stress Scale25) Chronic disease self-management (Patient Activation Measure30) Sleep quality (PSQI18) Sleep efficiency (Sleep Timing Questionnaire31) Diet (Food Propensity Questionnaire32) Appetite (Visual Analog Scale33) Physical activity (Stanford Five-City physical activity recall34) Sleep apnea (Berlin Sleep Questionnaire35) Restless legs syndrome (Screening interview36) Weight loss behavior change theoretical framework: Coping, diet, exercise, and activity self-efficacy37 Sleep behavior change theoretical framework: Sleep hygiene, cognitive behavioral therapy38	Significant difference in weight loss between groups (StCnt=2% vs. Tx= 5%, p=.04) Sleep efficiency improved in both intervention groups (p<.001) PSQI group baseline scores: Tx=6.9±2.6, StCnt=7.3±4.5 Sleep efficiency percent baseline scores: Tx=85.1±9.7, StCnt=82.6±19.8
7.	Food, Activity and Behavior Trial 39 Thomson et al., 2012 Overweight or obese (M BMI= 33.9±3.3 kg/m2) adult women (n=245) Mean years of age= 45.5±10.4 (100% female sample) Community samples enrolled at four study sites: University of California, San Diego; University of Arizona, Tucson; University of Minnesota, Minneapolis; and Center for Health Research, Kaiser Permanente Center Northwest, Portland, Oregon, United States ID: NCT00640900	Determine association between sleep characteristics and weight loss in overweight or obese women enrolled in a weight-loss program Multi-site randomized controlled trial Secondary analysis of trial data assessing a commercial weight loss intervention Sleep data collected at baseline and 6, 12, 18, and 24 months Tx: Successful weight loss (losing ≥10% baseline body weight by 6 months) (n=108) ACnt: Unsuccessful weight loss (<10% baseline body weight by 6 months) (n=137)	Weight loss Anthropometrics (Body weight, waist/hip circumferences) BMI (kg/m2) Sleep quality (PSQI18) Weight loss behavior change theoretical framework: Behavioral counseling to assist with weight loss Sleep behavior change theoretical framework: None reported	By 6 months, mean weight loss was 7.8 kg, with 44.1% of women showing a weight loss of ≥10% of their baseline weight ACnt associated with a lower likelihood of longer-term successful weight loss (RR, 0.80; 95% CI, 0.62–1.02 and RR, 0.72; 95% CI, 0.54–0.95 at 12 and 18 months, respectively); this association was attenuated by 24 months and no longer significant PSQI global scores below 5, decreased from 116 (47.4%) at baseline to 104 (42.5%) among all participants; reduced scores related to changes in sleep disturbance score and greater use of sleep medications (p=.01) PSQI group baseline scores: 116 (47.4%) women reported a PSQI Global Sleep Score <5
8.	Hopkins POWER Trial 40 Rubin et al., 2013 Obese (M BMI=36.6±5.0 kg/m2) adults (n=415) with at least one cardiovascular risk factor Mean years of age=54.0±10.2 (63.6 % females, 36.4% males; 56.1% Caucasian, 41.0 % African American) Participants recruited from six primary care practices in the Baltimore metropolitan area, Maryland, United States ID: NCT00783315	Assess patient-reported outcome changes over 24 months and to test for differences between the intervention arms Randomized controlled trial Data collected at baseline, 6, and 24 months Tx 1: In-person weight-loss intervention (n=138) Tx 2: Remote weight-loss intervention (n=139) ACnt: Information-based information (n=138)	Weight loss Sleep quality (PSQI18) Health status (SF-12 Health Survey41) Quality of life (EuroQol-5 Dimensions42) Depression (Patient Health Questionnaire 826) Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	No significant differences in PSQI change scores across arms Higher baseline BMI was associated with less improvement in PSQI scores (effect size=.34) Although weight loss during the trial period was associated with all other patient reported outcomes, it was not significantly associated with any changes in sleep quality PSQI baseline score of all participants: 7.33±2.94
9.	Adipogenic-Mediated Mediator of Weight Gain 43 Verhoef et al., 2013 Overweight and obese (31.9±3.2) adults (n=98) Mean years of age=40±9 (74.5% females, 25.5% males) Community sample recruited from Maastricht, Limburg, Netherlands ID: NCT01015508	Assess relationship between changes in sleep duration and changes in body weight and body composition during weight loss over the long term Quasi-experimental design Participants from three intervention arms (high, moderate, and low predisposition for weight regain) were pooled together for analysis Tx: Very-low-energy diet Data collected before weight loss, after weight loss, 3-, and 10-months post-weight loss follow-ups	Weight loss Anthropometrics (Body weight, waist/hip circumference, % body fat via air-displacement plethysmography/deuterium dilution) BMI (kg/m2) Physical activity (Baecke’s questionnaire44) Eating behaviors (three-factor eating questionnaire [TFEQ]45) Sleep duration (self-report) Daytime sleepiness (Epworth Sleepiness Scale [ESS]46) Biomarkers (leptin) Weight loss behavior change theoretical framework: Eating behavior traits assessment45 Sleep behavior change theoretical framework: None reported	Average weight loss was 10% after 2 months of dieting and 9% and 6% after 3- and 10-month follow-ups Short and average sleepers increased their sleep duration during weight loss [.7±1.1 hours (n = 21; p<.01) and 0.2±0.5 hours (n=57; p<.01) Changes in BMI and sleep duration largest during weight loss; stabilized during the 3-mo follow-up After the 10-mo follow-up, no longer a significant correlation likely due to weight regain of some participants Daytime sleepiness and time to fall asleep decreased during weight loss Short (#7 h) and average (.7 to ,9 h) sleepers increased their sleep duration, whereas sleep duration in long sleepers ($9 h) did not change significantly during weight loss Sleep duration was negatively correlated with the change in BMI during weight loss and after the 3-mo follow-up and with the change in fat mass after the 3-mo follow-up Baseline sleep duration: 7.6±0.9 hours
10.	Québec Family Study 47 Filiatrault et al., 2014 Overweight and obese (M BMI=33.3±3.5 kg/m2) adults (n=150) M years of age=38.8±8.6 (55.3% women, 44.6% men) Community sample in Québec City, Canada	Examination of associations between eating behavior traits and weight loss according to sleep quality and duration in adults enrolled in common weight-loss interventions One-group, pretest posttest design Secondary analysis of control participants pooled together from four weight loss studies Data collected at baseline, 12 and 16-weeks (depending on study) Only participants in control groups were analyzed for this study	Weight loss Eating behaviors (TFEQ45) BMI (kg/m2) Sleep quality (PSQI18) Waist circumference Food diary Weight loss behavior change theoretical framework: Eating behavior traits assessment45 Sleep behavior change theoretical framework: None reported	Changes in flexible control and strategic dieting behavior were constantly negatively associated with changes in body weight and fat mass (p<0.05) for recommended duration sleepers The change in situational susceptibility to disinhibition was positively associated with the change in fat mass and body weight for those having healthy sleeping habits (p<0.05) For poor quality sleepers, the change in avoidance of fattening foods was negatively associated with changes in adiposity (p<0.05) PSQI baseline score of all participants: M=4.6±2.6 Sleep duration of participants: 7.2±0.9 hours
11.	Look AHEAD (2014) 48 Sawamoto et al., 2014 Overweight and obese (M BMI=31.1±4.7 kg/m2) women (n=90) Mean years of age= 47.9±12 (100% Japanese female sample) Community sample attending an outpatient weight loss intervention in Fukuoka, Japan	Determine if sleep duration and quality can predict the magnitude of weight reduction in a weight-loss intervention program for overweight and obese women One group, pretest-posttest design Secondary analysis of trial data assessing a cognitive behavioral therapy weight loss intervention Data collected at baseline and at 7 months Tx 1: Normal (<5) waking events group (n=46) Tx 2: High (≥5) waking events group (n=44)	Weight loss Anthropometrics (Body weight, waist/hip circumferences, % body fat via dual energy x-ray absorptiometry) Sleep parameters (Actigraphy49) Diet (Food log) Physical activity (Exercise log) Sleep quality (PSQI18) Anxiety (State-Trait Anxiety Inventory)50 Depression (CES-D Scale)51 Biomarkers (Ghrelin, leptin, insulin-like growth factor, insulin, growth hormone) Weight loss behavior change theoretical framework: Cognitive behavioral therapy for weight loss52 Sleep behavior change theoretical framework: None reported	Sleep efficiency and number of wake episodes were positively associated with reduction in BMI (sleep efficiency: r=.258, p=.015; wake episodes: r=.239, p= .03) Fewer wake episodes at baseline were associated with greater reductions in BMI (β=−.341, p=.002) Reduction in BMI was significantly different between Tx 1 (15.4%) and Tx 2 (11.7%) PSQI-assessed parameters failed to detect an association with the reduction in BMI Baseline sleep parameters: sleep duration=332±69 minutes; sleep efficiency=93±5.1%; waking episodes= 5.72±3.1; PSQI, global score not provided
12.	POWER-UP 53 Alfaris et al., 2015 Obese (M BMI=38.5±4.7 kg/m2) adults (n=391) M years of age=51.5±11.6 (20.3% men, 79.7% women) Primary care settings in Pennsylvania, United States	To examine the effect of weight loss on sleep duration, sleep quality, and mood in obese participants receiving one of three behavioral weight loss interventions Randomized controlled trial Data collected at baseline, 6, and 12 months Tx 1: Brief lifestyle counseling (n=131) Tx 2: Enhanced brief lifestyle counseling (n=129) Standard Control [StCnt]: StCnt (n=130)	Weight loss Patient Health Questionnaire-8 (PHQ-8)26 BMI (kg/m2) Sleep quality (PSQI18) Self-reported sleep duration Weight loss behavior change theoretical framework: Behavioral counseling for diet and physical activity Sleep behavior change theoretical framework: None reported	At 6 months, participants in all three groups reported increases in sleep duration from 6.6 to 12.6 minutes/night At 6 and 24 months, PSQI scores declined in all three groups (p<.001), but no differences between groups At 6 months, sleep duration (21.6 vs 1.2 minutes) and PSQI (−1.2 vs. −0.4) improved in those participants that lost ≥5% of initial weight, compared to those that lost <5% of initial weight; however, these changes were not retained at 24 months PSQI baseline score of all participants: M=8.3±3.4 Sleep duration of all participants: 6.4±1.3 hours
13.	Diabète Québec 54 Tremblay et al., 2015 Overweight and obese (M BMI=33.0±3.5 kg/m2) women (n=75) Mean years of age= 39±8 (100% female sample) Community sample recruited from Québec City metropolitan area, Canada ID: NCT00353054	Examine effects of a diet-based weight loss program on energy intake, resting metabolic rate, appetite sensations, eating behaviors and sleep duration and quality in obese women resistant to weight loss One-group, pretest posttest design Secondary analysis of active control participants pooled together Tx: Weight loss intervention (n=75)	Weight loss Anthropometrics (Body weight, % body fat via dual energy x-ray absorptiometry) BMI (kg/m2) Sleep quality (PSQI18) Eating behaviors (TFEQ45) Resting metabolic rate (Indirect calorimetry) Diet (Food log) Weight loss behavior change theoretical framework: Eating behaviors45 Sleep behavior change theoretical framework: None reported	Mean weight loss of Tx=3.3±2.8 kg, explained by 2.9±2.6 kg reduction in fat mass (p<.001) Mean weight loss was 6.2±1.6, 3.4±0.6. and 0.2±1.4 kg in participants classified in the high, middle and low treatment response tertiles, respectively PSQI score was significantly decreased by the end of the intervention indicating an improvement in sleep quality (p=.02) Participants in the high tertile reported a significant improvement in sleep quality and an increase in sleep duration compared with other tertiles PSQI baseline score of all participants: 4.6±2.9 Baseline sleep duration of all participants: 7.4±0.9 hours
14.	SHIELD Study 55 Kuehl et al., 2016 Overweight and obese (Tx M BMI =30.31 kg/m2; Cnt M BMI=29.21 kg/m2) adults (n=408) M years of age of Tx=44.3±9.67 (55.9% male, 44.1% female) M years of age of Cnt=41.6±9.37 (68.6% male, 31.4% female) Worksite wellness program recruited among police and sheriff departments in Oregon and Southwest Washington in the United States ID: 5R01OH009676–02	Test the feasibility of a worksite wellness team-based intervention to improve diet, physical activity, body weight, and sleep and reduce the effects of unhealthy stress and behaviors Cluster randomized controlled trial Tx: SHIELD (n=204) Cnt: No treatment (n=204) Data collected at baseline, 6, 12, and 24 months	Weight loss BMI (kg/m2) Blood pressure Stress, health eating, physical activity scales Sleep disturbance (National Institutes of Health Patient-Reported Outcomes Information System (PROMIS) sleep disturbance56) Sleep quality (PSQI18) Sleepiness (Karolinska Sleepiness Scale57) Qualitative interviews Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	Program effects for sleep, stress, and general health variables were observed at 6-month follow-up but not at the 12- and 24-month follow-ups Significant program effects were observed for sleep quality (p<.001) and quantity (p<.05) Improvements in sleep occurred for personnel who worked on day, swing, and night shift hours Baseline sleep measures were unclear, but reported effect sizes for 6-month effects were d ppc3 =.32 for sleep quality and d ppc3 =.23 for sleep quantity (effect size, d ppc3 , calculated as pooled pretest standard deviation
15.	CALERIE 2 58 Martin et al., 2016 Nonobese (M BMI= 25.1±1.6 kg/m2) adults (n=218) Mean years of age=37.9±7.2 (69.7% female, 30.3% men) Community sample assessed at Pennington Biomedical Research Center, Baton Rouge, Louisiana; Tufts University (Tufts), Boston, Massachusetts; Washington University School of Medicine, St. Louis, Missouri; and was coordinated by Duke Clinical Research Institute, Durham, North Carolina (United States) ID: NCT00427193	To test the effect of calorie restriction on mood, quality of life, sleep, and sexual function in healthy nonobese adults Group randomized controlled trial Data collected at baseline, 12, and 24 months Tx: Calorie restriction (n=145) StCnt: Ad libitum (n=75)	Weight loss Calorie restriction (25%) Mood disturbance (Profile of Mood States59) General health (Short Form Health Survey60) Depression (Beck Depression Inventory61) Stress (Perceived Stress Scale25) Sleep quality (PSQI18) Sexual function (Derogatis Interview for Sexual Function–Self–report62) Biomarkers (reproductive hormone assays in men) Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	Tx and StCnt lost 7.6±0.3 kg and 0.4±0.5 kg, respectively, at month 24 (p<.001) Greater percent weight loss in the CR group at month 24 was associated with increased vigor (=−.30) and less mood disturbance (ρ=.27), improved general health (ρ=−.27), and better sleep quality (ρ=.28) (all p<.01). Sleep duration worsened in StCnt compared with Tx at month 12 (between group difference, −0.26; 95% CI, −0.49 to −0.02; effect size, −0.32; p=.03) PSQI baseline score of all participants:3.39±0.26
16.	Women’s Health Project for Women Wanting to Lose Weight 63 Nantsupawat et al., 2016 Overweight and obese (M BMI=33.4±4.3 kg/m2) female adults (n=40) Mean years of age=43.0±10.4 (100% female sample) Employees recruited from Texas Tech Health Sciences Center, Physicians Pavillion, Internal Medicine Clinic in Lubbock, Texas, United States ID: NCT01671397	Determine if counseling to standard weight loss counseling would alter unhealthy sleep habits and facilitate weight loss Randomized controlled trial Data collected at baseline and 6 months follow-up Tx: dietary advice with exercise and sleep counseling (n=27) ACnt: dietary advice with exercise counseling	Weight loss BMI (kg/m2) Sleep quality (PSQI18) Sleep duration (Sleep logs and sleep habit questionnaire64) Blood pressure Gait speed (100-Foot Walk Test65) Strength and coordination (Get Up and Go test66) Biomarkers (Glucose, cholesterol, triglycerides) Quality of life (World Health Organization Quality of Life survey67) Weight loss behavior change theoretical framework: Dietary advice with exercise counseling for Tx from physician Sleep behavior change theoretical framework: Sleep hygiene counseling for Tx from physician	BMI decreased from 33.4±4.3 kg/m2 to 30.4±4.3 kg/m2 (p<.002) No differences between the 2 counseling groups due to small number of short sleepers 21 participants reported symptoms of depression, which could impact weight loss and sleep quality Sleep duration at baseline of all participants: 8.0±1.0 based on sleep habit questionnaire; 7.4±1.2 hours based on sleep logs
17.	National Weight Control Registry & Chronotype 68 Ross et al., 2016 Obese (M BMI=36.2±4.7 kg/m2) adults (n=75) enrolled in two behavioral weight loss interventions (mean years of age=55.7±10.4; 77.33% female, 22.67 female) Overweight (M BMI=26.4±5.1 kg/m2) adults (n=690) registered with the National Weight Control Registry (mean years of age=51.7±12.5; 72.61% female, 27.36% male) Weight loss intervention participants recruited from Weight Control and Diabetes Research Center in Providence, Rhode Island, United States ID: NCT01428687	Identifying chronotype and sleep habits of successful weight loss maintainers compared to overweight and obese individuals enrolled in two behavioral weight loss interventions Quasi-experimental design Cross-sectional analysis of Tx vs. StCnt Tx required to have a BMI >25 kg/m2 StCnt requirements included weight loss of ≥13.6 kg and duration of weight loss maintenance ≥1 year (M weight loss at time of study: −33.91 ± 17.21 kg) Tx: Adults enrolled in behavioral weight loss interventions (n=75) StCnt: Adults registered with the National Weight Control Registry (n=690)	Weight loss BMI (kg/m2) Chronotype (Morningness–Eveningness Questionnaire69) Sleep quality (PSQI18) Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	StCnt had lower BMIs relative to Tx (p<.001) StCnt reported an average weight loss of −33.91±17.21 kg and reported maintaining this loss for an average of 7.26±6.05 years More subjects in StCnt were morning types compared to Tx (p=.004) and fewer were evening types (p=.014) StCnt participants reported better sleep quality, longer sleep duration, and shorter latency to sleep onset compared to Tx (p<.05) Fewer StCnt participants reported <6 or <7 hours of sleep (p<.01) PSQI group baseline scores: Tx=5.02±.11, StCnt=6.05±.34
18.	Look AHEAD (2016) 70 Sawamoto et al., 2016 Overweight and obese (M BMI=31.1±4.7 kg/m2) women (n=90) Mean years of age= 47.9±12 (100% Japanese female sample) Community sample attending an outpatient weight loss intervention in Fukuoka, Japan	Assess changes in serum adiponectin level during weight loss intervention and to determine if sleep parameters affect serum adiponectin levels One group, pretest-posttest design Secondary analysis of trial data assessing a cognitive behavioral therapy weight loss intervention Data collected at baseline and at 7 months	Weight loss Anthropometrics (Body weight, waist/hip circumferences, % body fat via dual energy x-ray absorptiometry) Sleep parameters (Actigraphy49) Sleep quality (PSQI18) Biomarker (Adiponectin levels) Depression (Center for Epidemiologic Studies-Depression Scale71) Anxiety (State-Trait Anxiety Inventory50) Weight loss behavior change theoretical framework: Cognitive behavioral therapy for weight loss52 Sleep behavior change theoretical framework: None reported	Adiponectin level and sleep efficiency significantly increased (p=.009) and the PSQI apnoea subscale significantly decreased (p=.005) at the end of the weight loss phase Adiponectin was significantly associated with sleep minutes (β=.210, p=.043) and body fat % (β=−.317, p<0.001) Baseline sleep parameters: sleep duration=331.7±66.9 minutes; sleep efficiency=92.9±5.1%; waking episodes= 5.72±3.1; PSQI apnoea subscale=.3±.9 * Sleep duration and sleep efficiency data reported differently between 2014 and 2016 studies
19.	Women Weigh-In for Wellness 72 Shade et al., 2016 Overweight and obese (M BMI=34.6±4.2 kg/m2) women (n=221) in rural counties Mean years of age=54.5±7.0 (100% female sample) Rural community sample recruited from Omaha, Nebraska, United States ID: NCT01307644	To observe relationships between sleep, pain, and health-related factors One group, pretest-posttest design Secondary analysis of trial data assessing a weight loss intervention	Weight loss BMI (kg/m2) Sleep parameters (Actigraphy49) Sleep disturbance/Pain interference (Patient Reported Outcomes Measurement System73) Blood pressure Waist circumference Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	Associations between change in self-reported sleep disturbance score and change in weight from baseline to 6 months (r=.202, p<.05) and BMI (r =.211, p<.05) No association with change in BMI or weight in terms of change in total sleep time Associations between number of awakenings and changes in weight (r=.167, p<.05) and BMI (r=.171, p<.05) Higher weight loss was a significant predictor of lower self-reported sleep disturbance at 6 months For every pound lost, a corresponding drop of .163 in self-reported sleep disturbance score Baseline sleep duration: 455.1±42.8 Baseline number of awakenings: 4.8±2.2 Baseline sleep efficiency: 96.3±2.1%
20.	SHINE 74 Adler et al., 2017 Obese (M BMI=35.5±3.62 kg/m2) adults (n=194) 47±13 years of age (80% women, 20% men; 59% white) San Francisco, California, United States ID: NCT00960414	Secondary outcome analysis comparing the effects of a mindfulness-based weight-loss intervention with an active control on self-reported sleep quality among adults with obesity Randomized controlled trial Data collected at baseline, 6, 12, and 18 months Tx: Mindfulness-based eating and stress management practices (n=100) ACnt: Progressive muscle relaxation (n=94)	Weight loss Five Facet Mindfulness Questionnaire75 BMI (kg/m2) Sleep quality (PSQI18) Sleep apnea (Berlin Sleep Questionnaire35) Mindfulness (Five-Factor Mindfulness Questionnaire75) Weight loss behavior change theoretical framework: Mindful eating and progressive muscle relaxation practices Sleep behavior change theoretical framework: None reported	Tx and ACnt experienced reduction in BMI, but reduction was not associated with change in sleep quality Between-group analysis observed more sleep improvement in Tx, but with small differences in effect size that were not statistically significant (Cohen’s d=0.09 at 6 months, 0.20 at 12 months, and 0.15 at 18 month) In the Tx, average weekly minutes of meditation practice time was associated with improved sleep quality from baseline to 6 months PSQI baseline score of all participants: M=5.95±3.00 PSQI group baseline scores: Tx=5.81±3.4, ACnt=6.1±2.6 BSQ group baseline scores: Tx=66%, ACnt=69% Mean sleep per night, group baseline: Tx=7.22±1.22 hours, ACnt=7.35±1.57 hours
21.	Physical Activity and Sleep in the Elderly: Pilot study 76 Goerke et al., 2017 Normal and overweight (M BMI=25.15±0.66 kg/m2) elderly adults (n=22) M years of age=68.36 (55% female, 45% male) Sample recruited from an elder care community dwelling in Magdeburg, Germany	To evaluate the effect of physical exercise on BMI and to assess sleep duration as a potential mediator of BMI reduction in the elderly Quasi-experimental design Tx: 12-week aerobic endurance training ACnt: progressive muscle relaxation Data collected at baseline and 12 weeks	Weight loss Physical activity (IPAQ77) BMI (kg/m2) Sleep quality (PSQI18) Sleep logs Biomarkers (lipid levels, HbA1c) Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	Overweight subjects reported a 50-min shorter sleep duration than normal weight subjects Reduction in BMI was only observed in Tx subjects who slept less than 7.5 hours per night BMI remained unchanged for Tx subjects that slept more than 7.5 hours per night and all ACnt subjects Mean sleep duration 452.47 ± 10.14 minutes Excluded participants with PSQI>5 Researchers seemed to consider 7.5 hours of sleep per night as “short sleep”
22.	MEASUR-UP 78 Payne et al., 2018 Obese (M BMI=36.9±6.3 kg/m2), elderly adults (n=67) with mild to moderate physical impairment Mean years of age= 68.2±5.6 (79% females, 21% males) Sample recruited from elder care community dwellings in Durham, North Carolina, United States ID: NCT01715753	Investigate bi-directional associations of a weight loss intervention with quality of life and mental health in obese older adults with functional limitations One group, pretest-posttest design Secondary analysis of trial data Data collected at baseline, 3, and 6 months Tx: high protein ACnt: RDA-level protein	Weight loss BMI (kg/m2) PSQI18 (*daytime dysfunction component score removed from global score calculation due to instrument error) Physical impairment (Short Physical Performance Battery) Depression (Center for Epidemiologic Studies Depression Scale79) Mood disturbance (Profile of Mood States59) Stress (Perceived Stress Scale25) Quality of life (Satisfaction with Life Scale80) General health (Short Form Health Survey60) Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	Longer baseline sleep onset latency was associated with less improvement in physical impairment score (β = −0.81, t1,46=−2.87, p=0.006) Sleep duration improved at 3 months (p=.005) and sleep efficiency improved at 3 and 6 months (p=.002) Baseline sleep duration of all participants: 0.6±1.0 Baseline sleep efficiency of all participants: 1.1±1.3
23.	WORDS 81 Wang et al., 2018 Overweight and obese (Tx M BMI=35.1±5.1; ACnt M BMI=31.3±3.3) adults (n=36) M years of age of Tx=45.3±6.0 (19.1% male, 80.9% female; 57.1% African American, 38.1% Caucasian) M years of age of ACnt=45.0±5.7 (20.0% male, 80.0% female; 66.7% African American, 33.3% Caucasian) Community sample recruited from Columbia and surrounding areas in South Carolina, United States ID: NCT02413866	Examine moderate sleep restriction on body weight, body composition, and metabolic variables in participants undergoing caloric restriction Randomized controlled trial Data collected at baseline and 8 weeks posttest Tx: caloric and sleep restriction (n=21) ACnt: caloric restriction (n=15)	Caloric restriction (Decreased to 95% of their resting metabolic rate) Sleep restriction (Total time in bed by reduced by 90 minutes 5 days a week) Sleep parameters (Actigraphy49) Anthropometrics (whole-body total mass, fat mass, lean mass, and body fat % via dual-energy x-ray absorptiometry) Biomarkers (Ghrelin and leptin) Metabolic rate (Indirect calorimetry) Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	Tx (−3.3±3.2kg) and ACnt (−3.2±2.5) lost similar amounts of weight (p=.87) Proportion of total mass lost as fat was significantly greater (p=.016) in the ACnt group Fasting leptin concentration was reduced only in Tx (p=.029) Approximately 1 hour of sleep restriction, five nights a week led to less proportion of fat mass loss in Tx vs. ACnt, despite similar weight loss Sleep restriction may adversely affect changes in body composition and “catch-up” sleep may not completely reverse it Group-by-race interaction for lean body mass change (p=.048), suggesting Tx and ACnt had different effects in African American and Caucasian participants Baseline total time in bed: Tx=393±53, ACnt=387±31
24.	Zumba Training 82 Muhammad et al., 2019 Normal and overweight (25.7±0.2 kg/m2) women (n=29) Mean years of age=21.1±0.4 Community sample recruited from Yogyakarta, Indonesia	Explore the effect of Zumba training on body composition and its compensatory effect on dietary intake and sleep in sedentary overweight women Crossover design Data collected at baseline and 12 weeks Tx: Zumba training (n=28) ACnt: Education (n=29)	Weight loss BMI (kg/m2) Anthropometrics (Body weight, waist/hip circumferences, % body fat via bioelectrical impedance analyzer) Global physical activity questionnaire Semi-quantitative83 food frequency questionnaire84 Sleep quality (PSQI-Indonesian translation85) Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	Tx was associated with the reduction of body fat (p=.023) but not body weight (p=.783) Tx improved sleep quality (p<.001) and duration (p=.047), relative to ACnt PSQI baseline score of all participants:8.3±0.5 Sleep duration at baseline of all participants: 5.3±0.2
25.	mDiet 86 Chevance et al., 2020 Overweight and obese (M BMI=32.7±3.4 kg/m2) English and Spanish speaking adults (n=278) M years of age=41.7±11.1 (77% women, 23% men; 41% Hispanic) Community sample in San Diego County, California, United States ID: NCT01171586	Examine between-person associations of health behaviors in adults with obesity participating in a weight loss study, as well as the covariations between these behaviors within-individuals across the study One-group, pretest posttest design Secondary analysis of control group participants enrolled in a randomized controlled trial Data collected at baseline, 6, and 12 months Tx: m-Diet (n=278)	Weight loss Self-reported health behaviors: physical activity, sedentary behavior, sleep, diet Self-reported sleep duration BMI (kg/m2) Weight loss behavior change theoretical framework: Behavioral counseling for diet and physical activity Sleep behavior change theoretical framework: None reported	Sleep duration was not related to other health behaviors assessed for the between-participant and within-participant networks Sleep duration of participants: 7.8±1.0 hours
26.	Move, Eat and Sleep 87 Duncan et al., 2020 Overweigh and obese (M BMI=31.7±3.9 kg/m2) adults (n=116) M years of age=44.5±10.5 (70.6% women, 29.3% men) Community sample in Newcastle area, New South Whales, Australia ID: ACTRN12617000735358	Compare the efficacy of two multi-component interventions on body weight cardiovascular risk factors, lifestyle behaviors, and mental health Randomized controlled trial Data collected at baseline, 6, and 12 months Tx 1 and Tx 2 pooled for assessment Tx 1: Enhanced (n=39) Tx 2: Traditional (n=41) Wait-list Control (n=36) [WLCnt]: HWL (n=121)	Weight loss BMI (kg/m2) Anthropometrics (Waist circumference) Biomarkers (HbA1c) Sleep parameters (Actigraphy49) Sleep quality (PSQI18) Insomnia (Insomnia Severity Index)24 Sleep Timing Questionnaire (STQ)31 Sedentary behavior (Workforce Sitting Questionnaire88) Diet (Australian Eating Survey89) Physical activity (Cardiovascular and resistance training self-report) Chronotype (Morningness–Eveningness Questionnaire69) Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	Compared with WLCnt, average body weight of Tx 1 and Tx 2 group did not differ at 6 (between-group difference=−0.92, (95% CI=−3.33, 1.48)) or 12 months (0.00, (95%=CI −2.62, 2.62)) Compared with WLCnt, Tx 1 and Tx 2 improved insomnia symptoms at 12 months (−2.59, (−4.79, −0.39)) and improved bedtime variability at 12 months (−1.08, (−1.86, −0.29)) PSQI baseline score of all participants: M=7.0±3.0 ISI baseline score of all participants: M=8.7±4.8 Accelerometer baseline score of all participants: M=6.5±0.9 hours
27.	Healthy US-Style Eating Pattern 90 Hudson et al., 2020 Overweight and obese (M BMI= 32.6±0.5 kg/m2) adults (n=51) with poor sleep as assessed by global PSQI18 score M years of age=47±1.0 (84.3% female, 15.7% male) Community sample in Lafayette, Indiana, United States ID: NCT03174769	Assess effects on sleep indices on consuming recommended or higher amounts of USDA-recommended, animal-based protein-rich foods during energy restriction Randomized controlled trial Tx: High protein quantity (n=21) ACnt: Recommended protein quantity (n=30) Data collected at baseline, 6, and 12 weeks	Weight loss BMI (kg/m2) Sleep parameters (Actigraphy49) Sleep quality (PSQI18) Daytime sleepiness (ESS46) Biomarkers (Salivary melatonin concentrations) Weight loss behavior change theoretical framework: None reported Sleep behavior change theoretical framework: None reported	Body mass decreased among all participants (−6.2±0.4 kg) Objective measure of sleep efficiency improved in both groups (1±1%; p=0.026) Subjective measures of perceived sleep quality (−2.7±0.4) and daytime sleepiness (−3.8± 0.4) improved in both groups (p<.001) Group of participants transitioned from poor quality sleepers (PSQI>5) to good quality sleepers (PSQI≤5) PSQI baseline score of all participants: 7.9±0.5 PSQI post-intervention score of all participants: 4.0±0.6 * PSQI change scores reported differently in abstract vs. results
28.	PREVIEW 91 Adam et al., 2021 Prediabetic, overweight and obese (M BMI=35.26 kg/m2 (95% CI=33.93 kg/m2, 36.59 kg/m2) adults (n=2,184) 25 to 70 years of age (32.3% men, 67.7%% women) Community sample recruited from Australia, Denmark, Finland, Hungary, the Netherlands, New Zealand, Spain, and the United Kingdom ID: NCT01777893	Assess associations of psycho-behavioral variables measured during 3-year weight loss maintenance period Secondary analysis of a multicenter randomized controlled trial Assessments at baseline, 6 months, 1 year, 2 years, and 3 years 2×2 diet-by-physical activity intervention Tx 1: High-protein diet Tx 2: Low protein diet Tx 3: Moderately intense physical activity Tx 4: High intensity physical activity	Weight loss Insulin resistance (Homeostatic Model Assessment of Insulin Resistance92) BMI (kg/m2) Sleep quality (PSQI18) Sleepiness (ESS46) Sleep duration (Actigraphy49) Stress (Perceived Stress Scale25) Mood disturbance (Profile of Mood States59) Weight loss behavior change theoretical framework: Eating behavior traits assessment45 Sleep behavior change theoretical framework: None reported	BMI, homeostatic model assessment of insulin resistance, and the psycho-behavioral variables were not statistically different between the four intervention groups from baseline to year 3 Sleepiness and poor sleep quality were consistently and positively associated with BMI over the 3-year period Sleep duration and low sleep quality scores were positively associated with HOMA-IR in men (p<.001) In all participants, daytime sleepiness was decreased compared with baseline at all time points PSQI baseline score: M=6.49 (95% CI=6.17, 6.81); 3-year posttest: M=6.19 (95% CI=5.86, 6.52) ESS baseline score: M=7.81 (95% CI=7.07, 8.56); 3-year posttest: M=7.29 (95% CI=6.54, 8.04) Sleep duration baseline: M=477.9 (95% CI=470.0, 484.9) minutes; 3-year posttest: M=488.5 (95% CI=480.4, 496.7) minutes
29.	Healthy Families Healthy Forces 93 Das et al., 2021 Overweight and obese (M BMI=34.7 kg/m2) adults (n=238) M years of age=41 (97.4%, 99.2% women in Tx and ACnt) Community sample delimited to dependents of active-duty or retired military personnel from US military installations in Colorado, Connecticut, Kentucky, Massachusetts, and New York, United States ID: NCT02348853	Evaluate the efficacy of two lifestyle interventions for weight loss in overweight and obese adults Randomized controlled trial Data collected at baseline, 6, and 12 months Tx: Healthy weight for living (n=121) Standard Care Control (StCnt): Current best practice (n=117)	Weight change, kg Sleep quality (PSQI18) Cardiometabolic risk factors (Cholesterol, blood pressure, glucose) Self-reported health (Short Form Health Survey,60 Beck Depression Inventory61) Diet (24-hour dietary recall94) Physical activity (Stanford Five-City physical activity recall34) Eating behaviors (TFEQ45) Weight loss behavior change theoretical framework: Goal setting theory, social cognitive theory,27 transtheoretical model,28 eating behaviors45 Sleep behavior change theoretical framework: None reported	38% of Tx and 30% of StCnt completers achieved 10% or more weight loss (p=0.32) Improvements in sleep quality were similar between Tx and StCnt (p=.78) PSQI group baseline scores: StCnt=7.7±3.7, Tx=7.5±3.6; 12-months posttest: StCnt=7.7±3.7, Tx=7.5±3.6 PSQI mean group difference, 12-months posttest: M=0.14 (−1.07, 1.35)
30.	National Health and Examination Survey 95 Knell et al., 2021 M years of age of unsuccessful long-term weight loss participants=42.8±0.7 (46.5% female, 53.5% male) M years of age of successful long-term weight loss participants=44.8±0.5 (42.8% female, 57.2% male) Population-based sample (n=1,735) in the United States	Cross-sectional study evaluating the joint relations of sleep, physical activity, and sedentary behaviors with successful long-term weight loss Observational study of secondary data (2005–2006 wave of the National Health and Examination Survey)	Weight loss BMI (kg/m2) Physical activity/sedentary behavior (Actigraphy49) Sleep outcomes (Functional Outcomes of Sleep Questionnaire96) Weight loss behavior change theoretical framework: None reported Sleep-based behavior change theoretical framework: None reported	A total of 525 participants (30.26%) successfully maintained at least 5% weight loss for 12 months After adjustment for relevant factors, there were no significant associations between physical activity, sedentary time, sleep, and successful long-term weight loss Average sleep time per night was dichotomized into “ideal sleepers” (≥7 hours of sleep per night) and “short sleepers” (<7 hours of sleep per night)
31.	Time-Restricted Eating 97 Park et al., 2021 Normal weight (22.7±2.7 kg/m2) adults (n=34) Mean years of age= 22.5±2.8 (76.4% female, 23.6% male) Community sample recruited in Seoul, South Korea ID: PRE20210–002	Examine the effects of 8 hours of time-restricted eating on body weight and cardiometabolic risk factors in young adults mainly active at night Crossover design Feasibility study Data collected at baseline, 2, and 4 weeks Tx: Weight loss group (n=23) ACnt: Weight gain group (n =10)	Weight loss Body composition (Bioelectrical impedance) BMI (kg/m2) Sleep quality (Selected items from PSQI18) Chronotype (Morningness–Eveningness Questionnaire69) Weight loss behavior change theoretical framework: None reported Sleep-based behavior change theoretical framework: None reported	All participants showed late-shifted sleeping patterns, but no significant differences in sleep duration or sleep quality between Tx and ACnt Tx experienced significant changes in body weight (−1.0 ±1.4 kg), body mass index (−0.4±0.5 kg/m2), and body fat (−0.4±1.9%) after 4 weeks Participants experienced significant changes in body weight (−1.0±1.4 kg), body mass index (−0.4±0.5 kg/m2), and body fat (−0.4±1.9%) after 4 weeks of Tx Baseline sleep quality of all participants: very good (15.2%), fairly good (51.5%), fairly bad (33.3%)
32.	ICECAP Trial 98 Peos et al., 2021 Resistance-trained (M weight=77.2±16.1) kg; M body fat %=25.5±6.1) adults (n=61) Mean years of age=28.7±6.5 (52.5% females, 47.5% males) Community sample recruited from Perth, Western Australia, Australia ID: ACTRN12618000638235p	Determine if energy restriction improve fat loss and fat-free mass retention compared with continuous energy restriction in resistance-trained adults Randomized controlled trial design Data collected at baseline, 7, and 15 weeks Tx: Intermittent energy restriction (n=30) ACnt: Continuous energy restriction (n=31)	Weight loss Weight, fat mass, fat-free mass (Dual-energy x-ray absorptiometry) Sleep parameters (Actigraphy49) Sleep quality (PSQI18) Daytime sleepiness (ESS)46 Weight loss behavior change theoretical framework: None reported Sleep-based behavior change theoretical framework: None reported	After accounting for baseline differences, Tx did not result in lower fat mass or body weight, or greater fat-free mass, compared with ACnt after energy restriction Daytime sleepiness scores increased in response to both interventions but did not differ between Tx and ACnt Sleep quality scores were not affected by Tx or ACnt PSQI group baseline scores: Tx=5.36, StCnt=5.34 ESS group baseline scores: Tx=7.46, StCnt=5.26
33.	Energy Restricted High-Volume Resistance Training 99 Roth et al., 2021 Normal weight (Tx M BMI=24.68±2.19; Cnt M BMI= 24.55±2.54) male college students (n=28) Mean years of age of Tx= 26.57±4.20; mean years of age of Cnt=25.29±2.97 (100% males to control for hormone fluctuations) Participants recruited from local sports clubs and university courses in Frankfurt, Germany ID: DRKS00017263 (https://trialsearch.who.int/Trial2.aspx?TrialID=DRKS00022146)	Investigate whether a high-protein, moderately energy-restricted diet preserves lean body mass in resistance-trained males in the absence of resistance training, while examining influence of moderator variables (e.g., sleep) Randomized controlled trial design Data collected at baseline, weeks 1, 2, 3, 4, 5, and 6 Sleep assessed at weeks 1, 3, and 5 Tx: Energy restriction group (n=14) Cnt: Eucaloric control group (n=14)	Body composition (Bioelectrical impedance) Muscle contraction and quality (Tensiomyography; MyotonPRO) Sleep quality (PSQI18) Mood disturbance (Profile of Mood States59) Weight loss behavior change theoretical framework: None reported Sleep-based behavior change theoretical framework: None reported	Relative to Cnt, Tx revealed greater reductions in body mass (−3.22 kg vs. −1.90 kg, p<.001, partial η2=.360), lean body mass (−1.49 kg vs. −.68 kg, p<.001, partial η2=.152), body cell mass (−.85 kg vs. −.59 kg, p<.001, partial η2=.181), and body fat percentage (−1.74% vs. −1.22%, p<.001, partial η2=433) Sleep onset, sleep duration did not change (p > 0.05) PSQI score (1–1.43 vs. 1–0.64, p=0.006, partial η2=.176) decreased significantly in both groups PSQI group baseline scores: Tx=5.14±1.75, Cnt=5.07±2.23
34.	PREFER 100 Teong et al., 2021 Overweight or obese (M BMI= 32.9±4.4 kg/m2) adult women (n=46) Mean years of age=50±9 (100% female sample) Community sample recruited from in Adelaide, South Australia, Australia ID: NCT01769976	Compare the effects of energy matched intermittent fasting versus calorie restriction on eating behaviors, mood, sleep quality, quality of life, and cognition Randomized controlled trial Secondary analysis of intervention completers of two of the four arms of the trial Data collected at baseline and 8 weeks Tx: Intermittent fasting (n=25) ACnt: Calorie restriction (n=26)	Weight loss Anthropometrics (Body weight, waist/hip circumferences, % body fat via dual energy x-ray absorptiometry) BMI (kg/m2) Sleep quality (PSQI18) Weight loss behavior change theoretical framework: None reported Sleep-based behavior change theoretical framework: None reported	Greater body weight loss (p=.032) and fat mass loss (p=.039) in Tx compared to ACnt Perceived eating behaviors, mood, sleep quality and cognitive performance were not changed by an acute 24-hour fast within Tx (p>.05) PSQI group baseline scores: Tx 1=4.9±2.6, Tx 2=5.0±2.7
35.	Timing of Exercise Initiation Within a Lifestyle Weight Loss Program 101 Creasy et al., 2022 Overweight or obese (M BMI= 34.4±4.3 kg/m2) adults (n=156) Mean years of age=40±9 (16% males, 84% females) Community sample recruited in Denver, Colorado, United States ID: NCT01985568	Compare sleep quality and quantity outcomes with dietary and physical activity adherence Data collected at baseline, 6, 12, and 18 months Secondary analysis of behavioral weight loss randomized controlled trial Tx 1: Standard behavioral intervention Tx 2: Sequential behavioral intervention	Weight loss BMI (kg/m2) Sleep parameters (actigraphy,49 bedtime, waketime, midpoint of sleep, regularity, sleep efficiency, sleep duration, time in bed, wake after sleep onset, sleep onset latency, number of awakenings) Physical activity (METs) Diet (Food log) Weight loss behavior change theoretical framework: Group-based behavioral support Sleep behavior change theoretical framework: None reported	Weight loss occurred at 6 (M=7.7±5.4 kg), 12 (M=8.4±7.9 kg), and 18 (M=7.1 ± 9.0 kg) months Lower sleep efficiency, higher wake after sleep onset, more awakenings, and higher sleep onset latency were associated with reduced weight loss after controlling for covariates (p<.05) Sleep duration group baseline=411.0±68.2 minutes

Notes.

^*Study names derived from reviewed publication and/or based on name of study provided on clinicaltrials.gov.

^†Study ID linked to trial registry and/or grant number.

Abbreviations. Tx, treatment group; Cnt, control group; ACnt, active treatment concurrent control; StCnt, standard care control; WLCnt, Wait-list Control; BMI, body mass index; ESS, Epworth Sleepiness Scale; IPAQ, International Physical Activity Questionnaire; PSQI, Pittsburg Sleep Quality Index; TFEQ, Three-Factor Eating Questionnaire.

Study Acronyms. LIFE, Randomized Trial of Tapas Acupressure for Weight Loss Maintenance; PREVIEW, Prevention of Diabetes Through Lifestyle Intervention and Population Studies in Europe and Around the World; SHINE, Supporting Health by Integrating Nutrition and Exercise; POWER-UP, Practice-based Opportunities for Weight Reduction trial at the University of Pennsylvania; mDiet, A Text Message Intervention for Weight Loss; HF2, Healthy Families Healthy Forces; SHIELD, Safety & Health Improvement: Enhancing Law Enforcement Departments; PATH, Physical Activity for Total Health; CALERIE 2, Comprehensive Assessment of Long-term Effects of Reducing Intake of Energy Phase 2; MEASUR-UP, Measuring Eating, Activity and Strength: Understanding the Response-Using Protein; GH-IGF-I & Sleep, Growth Hormone-Insulin-Like Growth Factor-I and Sleep; Hopkins POWER, Practice-based Opportunities for Weight Reduction; Look AHEAD, Action for Health in Diabetes; PREFER, Effects of Periodic Fasting Versus Daily Energy Restriction on Metabolic Health; FAB, Food, Activity and Behavior Trial; WORDS, Weight Outlooks by Restriction of Diet and Sleep.

Study Characteristics

The most common study location was the United States,^{16,20,23,29,39,40,53,55,58,63,68,72,74,78,81,86,90,93,95,101} but studies were reported from Australia,^87,91,98,100 Canada,^21,47,54 Denmark,¹⁹ Germany,^{76 99} Indonesia,⁸² Japan,^48,70 Korea,⁹⁷ the Netherlands,⁴³ and from multiple countries.⁹¹ Settings included hospital clinics,¹⁹ universities,^20,39,58,99 primary care,^29,40,53 and a law enforcement work setting.⁵⁵ Majority of the reviewed studies used the Pittsburgh Sleep Quality Index (PSQI)^{18,20,21,29,39,40,47,48,53–55,58,63,68,70,74,76,78,82,87,90,91,93,97–100} to measure sleep whereas other studies used measures such as the Sleep Timing Questionnaire (STQ),^31,87 PROMIS (perceived sleep quality, depth, and restoration),^56,72 Insomnia Severity Index (ISI),^24,87 Epworth Sleepiness Scale (ESS),^43,46 Functional Outcomes of Sleepiness (FOSQ),^95,96 Women’s Health Initiative Insomnia Rating Scale Questionnaire,^16–18 and actigraphy.^{48,70,72,81,87,90,91,95,98} Most of the studies did not report employing a weight loss and/or sleep behavior change theoretical framework. Among those that used behavior change theoretical frameworks, social cognitive theory,^23,27,93 transtheoretical model,^23,28,93 behavioral counseling/support,^{21,39,63,86,101} cognitive behavioral therapy,^52,70 eating behaviors^{43,47,54,74,91,93} and self-efficacy^29,93 were applied to facilitate weight loss, while sleep hygiene^29,63 and cognitive behavioral therapy²⁹ were used to modify sleep behaviors.

Nine studies^{16,39,48,54,63,70,72,82,100} included only women and those that focused on both men and women tended to enroll a larger proportion of women relative to men in their samples. Most studies related to sleep were secondary analyses of intervention data coming from trials in which weight loss was the primary outcome variable under observation.^{16,21,23,39,47,48,54,70,72,78,86,91,100,101} In some of these instances,^21,47,54,86 although the primary interventions applied parallel-group, random assignment, the findings related to sleep were based on analysis of control group data.^21,86 Two studies used cross-over designs,^82,97 a more common feature of dietary interventions. No power calculations were reported for sleep outcomes, which points to the potential for under-powered designs. Studies tended to include smaller sample sizes, with most enrolling less than 200 participants.^{16,19–21,29,43,47,48,54,63,70,74,76,78,81,82,87,90,97–100} Furthermore, many of the reviewed studies which used the PSQI as their primary sleep-related outcome variable, reported mean global sleep quality scores within, or near the cut point, for the “good sleep quality” (global scores less than 5) category at baseline.^{20,21,47,54,58,68,74,98–100}

The role of sleep as a variable tended to fall into four major categories: a) sleep at baseline as a predictor of subsequent weight loss during an intervention, b) sleep assessments after a history of successful weight loss, c) concomitant changes in sleep associated with weight loss, and d) experimental manipulation of sleep and resulting weight loss.

Sleep at Baseline as a Predictor of Weight Loss

Studies assessing sleep quality and/or duration as a predictor of weight loss in response to interventions yielded varying patterns of results depending on the type of sleep assessment and weight loss intervention. Four studies demonstrated that baseline sleep was significantly predictive of improved weight loss outcomes following dietary and/or physical activity interventions. More specifically, prospective associations existed between: self-reported sleep quality and weight loss;³⁹ self-reported sleep duration and weight loss;²³ self-reported sleep duration and sleep quality and loss of body fat;²¹ and actigraphy-based sleep parameters and attenuated weight loss.¹⁰¹ However, a fifth study failed to detect significant associations between self-reported sleep duration or quality and changes in BMI, although it did observe a significant, negative association between objective measures assessed via actigraphy (e.g., number of wake episodes) and BMI.⁴⁸

In contrast, two studies examining weight loss following participation in physical activity interventions demonstrated the opposite pattern of results.^16,76 Specifically, Goerke et al.⁷⁶ reported that reductions in BMI were observed only among participants with short sleep duration (<7.5 hours per night). Similarly, Littman et al.¹⁶ reported that the greatest differences in weight loss across intervention and control conditions were observed among participants whose sleep duration decreased from baseline to follow-up. However, it should be noted that this physical activity intervention yielded relatively small changes in body weight (M=1.3 kg). Overall, these studies point to only modest evidence to suggest that better sleep quality and duration predict positive outcomes for dietary interventions, whereas individuals with shorter sleep duration may benefit more from physical activity interventions.

Sleep After History of Weight Loss

Two studies examined associations between a history of successful weight loss and sleep using unique study designs. Ross et al.⁶⁸ compared sleep quality of individuals with a history of successful weight loss and maintenance relative to individuals presenting to begin a weight loss program. Results showed that participants with a history of successful weight loss reported better sleep duration, quality, and shorter sleep latency relative to those initiating the weight loss intervention. It should be noted, however, that individuals with a history of weight loss were recruited from a national registry and time between successful weight loss and the assessment of sleep was not reported and thus potentially varied widely across the study group. In contrast to Ross et al.’s findings, Rasmussen et al.¹⁹ compared objective sleep measures obtained from participants soon after successful completion of a weight loss intervention relative to healthy, nonobese community controls. Results indicated that controls had significantly longer sleep duration than those with recent successful weight loss. Given that these studies did not include repeated assessments within intervention and control conditions, it is difficult to draw any firm conclusions from their findings.

Additionally, three studies also examined associations between a history of successful weight loss and sleep by looking at weight loss maintenance over a long period of time. Sawamoto et al.⁷⁰ reported that sleep efficiency, but not duration, significantly increased from baseline to the end of the weight-loss phase. However, change in adiponectin was significantly associated with change in sleep duration and change in body fat. Adam et al.⁹¹ found that sleepiness and poor sleep quality were consistently and positively associated with BMI over the 3-year time period of weight loss maintenance. Sleep duration and low sleep quality scores were associated with homeostatic model assessment of Homeostatic Model Assessment for Insulin Resistance (HOMA-IR) for men only. In contrast, Kenll et al.⁹⁵ found that a greater proportion of unsuccessful long-term weight loss maintainers slept the ideal number of hours (7 or more per night) than those successfully maintaining long-term weight loss; however, these differences were not statistically significant (p=0.35). After an adjustment of relevant factors, there were no significant associations between any of the independent body weight loss behaviors (e.g., sleep) and successful long-term weight loss.

Taken together, there is no clear evidence to suggest that long-term weight maintenance is associated with sleep quality and/or duration, however, the direction of these correlational connections is unclear given limitations in study designs.

Changes in Sleep Associated with Weight Loss

Twenty-two studies investigated longitudinal changes in sleep quality and/or duration during weight loss interventions. Several single-arm intervention studies provided some evidence that sleep duration and/or quality improved over the course of a behavioral weight loss intervention,^47,54,78 cognitive behavioral interventions,^39,70 and a very low-energy diet.⁴³ Shade et al.⁷² observed improvements in sleep disturbance and number of awakenings, but not sleep duration, were associated with weight loss over the course of a web-based behavioral intervention. A recent study by Das et al.⁹³ found weight loss was accompanied by marked improvements in sleep as well as other health areas. Improvements in sleep were seen across both weight loss groups. Each of these studies lacked a control group, making it difficult to draw conclusions regarding the efficacy of these interventions to improve sleep.

Studies comparing changes in sleep across active treatment conditions and control groups yielded more equivocal findings. Within an aging population, a diet-based intervention appeared to prevent decreasing sleep duration for those in the intervention, but not in the control condition.⁵⁸ Short-term improvements in sleep quality and duration were also observed within a wellness intervention targeted at law enforcement officers relative to a control condition, however long-term improvements were not maintained.¹⁰² Alfaris et al.⁵³ found significant differences favoring counseling conditions over the usual care group with regards to sleep duration at a long-term follow up, but enhanced counseling did not outperform brief counseling. Similarly, Nantsupawat et al.⁶³ did not observe significant differences in sleep duration across standard and enhanced dietary counseling conditions. Relative to control conditions, several other interventions also resulted in an absence of changes in sleep quality from pre- to post-intervention across a mindfulness-based intervention,⁷⁴ behavioral weight loss intervention,⁴⁰ cognitive behavioral interventions,²⁹ and dietary intake focused on altered protein content.⁹⁰ Chevance et al.⁸⁶ also failed to find associations between sleep duration and other health behaviors (e.g., physical activity, fruit and vegetable consumption), either within or across weight loss intervention conditions. However, it should be noted that several of these interventions were secondary analysis of trials that initially powered to observe other variables (e.g., weight loss) beside sleep. For example, the mean value of sleep quality in the Adler et al.⁷⁴ mindfulness-based study, for example, was 5.95(SD=3.00), where scores ≥5 indicate poor sleep quality. In this study, the cut-point, though clinically relevant, had a standard deviation of 3.00, indicating a sizeable portion of the sample did not have poor sleep quality. Thus, a sample size of 194 may not have been able to detect statistically significant treatment effects. Inclusion criteria requiring all participants to fall into the poor sleep quality range may have increased power for detecting changes in sleep outcomes.

More recent studies have focused on dietary interventions such as time restricted/intermittent fasting or energy restricted diets while investigating changes in sleep. Two studies observed different energy restricted diets while measuring other variables such as sleepiness, sleep patterns, and sleep quality. Park et al.⁹⁷ found significant individual changes in body weight and percent of body fat after an 8-hour time-restricted eating intervention. No significant difference was seen in participants’ sleep patterns between weight loss group and weight gain group after 4 weeks. Additionally, both Peos et al.⁹⁸ and Roth et al.⁹⁹ examined energy restricted diet effects on weight loss and measured sleep as one of the study’s variables. Peos et al.⁹⁸ reported an increase in sleepiness for both energy restricted groups. Roth et al.⁹⁹ found that while the energy-restricted group had greater reductions in body composition, there was no significant difference found for sleep in hours per night and time to fall asleep (p>0.05). Teong et al.¹⁰⁰ examined both calorie restriction vs intermittent fasting and reported that neither calorie restriction or intermittent fasting had altered perceived sleep quality. It was also observed that the 24-hour fast did not impact perceived sleep quality. Trials that focused on physical activity interventions produced similarly mixed findings. One intervention resulted in improved sleep quality and duration within the treatment condition, but not among controls.⁸² Other physical activity-based interventions did not observe such differences.⁷⁶

Manipulation of Sleep and Weight Loss

Three studies attempted to study changes in weight associated with manipulations in participant sleep. The first study attempted to improve participant sleep via an enhanced weight loss intervention including sleep hygiene.⁸⁷ There were no significant group differences in weight loss between the traditional weight loss intervention and the enhanced intervention, but the enhanced intervention group did show significantly lower waist circumference measurements relative to the traditional intervention group. Similarly, interventions focused on sleep restriction yielded results implicating sleep in body composition, but not weight loss. Specifically, Wang et al.⁸¹ did not observe significant differences in weight loss between calorie restriction vs. calorie restriction + sleep restriction interventions, they did observe significant reductions in the proportion of mass lost from fat among the calorie restriction group (p=.004), but not the calorie restriction + sleep restriction group. Finally, in a lab-based sleep restriction experiment, Nedeltcheva et al.²⁰ demonstrated similar weight loss across sleep restricted vs. ad libitum conditions, but observed significantly more weight lost from fat within the ad libitum condition relative to the sleep restriction condition. Further, participants reported increased hunger during the sleep restriction condition relative to the ad libitum condition. Given this study’s extreme level of control over caloric intake and sleep, in addition to some support from home-based sleep manipulation studies, there is some evidence to suggest that changes in body composition are more favorable with increased sleep duration, despite the absence of differences in body weight related to changes in sleep duration.

DISCUSSION

Efforts to prevent and treat obesity have increased focus on interventions to improve diet and physical activity and to examine, including sleep as a factor. This review assessed 35 peer-reviewed articles in the literature on the relationship between sleep and weight loss with the primary objective to examine prospective associations between sleep and weight loss as well as longitudinal changes in sleep during weight loss interventions. The reports varied considerably in terms of specific research questions addressed and study design and methods used, including the timing of sleep assessment and weight loss interventions. We also sought to identify new research targets in the area of sleep as related to weight loss. Previous research has shown that insufficient sleep can contribute to increased energy intake^103,104 and is increasingly prioritized in addition to physical activity and diet in weight loss interventions. Our results provide some evidence that baseline sleep characteristics are predictive of weight loss during dietary interventions,^21,39,48 but not for interventions that focused solely on increasing physical activity.^16,76 The solution to weight loss, thus, may not be simply focusing on diet or physical activity, but also calling attention to sleeping habits prior to engaging in dietary interventions, particularly when designing weight-reduction programs.

Despite several single-arm intervention trials showing that various aspects of sleep improved in response to weight-loss interventions,^{43,47,54,70,72,78} the results were equivocal among studies that included a comparative control (or comparison) group. Further, the results of randomized controlled trials with interventions with the general population generally failed to detect differences in changes in sleep across groups, with the exception of studies focused on specific populations (e.g., aging adults and law enforcement officials).^{29,40,53,58,63,74,86,90,102} These mixed results, along with the variable quality and generally low power of studies, limit the conclusions that can be drawn about the effectiveness of weight loss interventions and resulting weight loss to improve sleep. The literature reviewed does not support the assumption that sleep will improve concomitantly with weight loss resulting from behavioral interventions focused on diet and exercise. Studies, and specifically randomized controlled trials that recruit larger samples and enroll participants with adequately powered sleep parameters at baseline, that employ a longitudinal design with validated and reproducible long-term outcome assessments of changes in weight and sleep are required to draw firmer conclusions than the literature presently affords.

Studies focused on altering sleep patterns and resulting weight loss consistently showed that, despite similar weight loss outcomes, participants with reduced sleep lost less body mass in the form of fat.^20,81 Whereas, an intervention focused on improving sleep quality and duration observed larger changes in waist circumference, but not weight loss among those who received a weight loss intervention. This included a focus on sleep relative to those who received the weight loss intervention only.⁴⁷ Results from these rigorous experimental study designs strongly suggest that better sleep is associated with improved body composition outcomes and may provide evidence that points to the importance of targeting sleep in order to promote positive outcomes in weight loss interventions. Most studies we reviewed included a large proportion of female participants. Out of the 35 studies reviewed and analyzed, 9 reported only focusing on women^{16,39,48,54,63,70,72,82,100} and several others included samples that predominantly comprised women. With a larger focus on women’s sleep in weight loss interventions, generalizability of these results may be limited. Moreover, the studies we analyzed had specific inclusion and exclusion criteria that limited the number of participants, creating more limitations regarding generalizability of results.

Limitations and Strengths

Several limitations deserve mention. First, this review included the selection of only English-language, peer-reviewed publications, which may have excluded relevant studies published in other languages or in the grey literature. Furthermore, we did not include or fully review articles that reported studies of specific populations such as individuals with diabetes, cancer, fibromyalgia. Second, most of the studies reviewed employed subjective, questionnaire-based measures of sleep quality and/or quantity (e.g., the PSQI), which are susceptible to recall biases and issues of validity). Third, many of the studies enrolled participants with adequate sleep parameters at baseline (e.g., global PSQI scores in the good quality sleep range; sleep durations within epidemiologically healthy ranges), which could lead to issues of investigations being underpowered to detect statistically significant changes in sleep outcomes. Most studies reported no use of behavior change theoretical frameworks, which may reduce potential treatment effects. Finally, a large portion of studies reviewed and assessed focused on populations that were obese or overweight and the use of unblinded data may have increased the risk of bias in the review. While the articles focused on weight loss in study participants with sleep apnea were excluded, we could not discern how undiagnosed mild to moderate sleep apnea may have been a confounder in the studies that were reviewed. Despite these limitations, our study provides an in-depth review of the literature dating back 17 years in which we screened a high number of published research articles. Our review also provides details of the study process, including study selection, data extraction and data analysis. Additionally, the validity of our review is strengthened by having applied a priori inclusion and exclusion criteria in the selection of candidate articles and corroborating findings through independent review by three researchers. Finally, our findings suggest that practitioners should consider the duration and quality of sleep when implementing weight loss programs.

Future Research Directions

Future studies must be adequately powered to detect statistically significant interactions between weight loss and sleep outcomes. In this context, we advise researchers power their studies using clinically significant benchmarks: 1) weight loss of 5–10% from baseline¹⁴ and 2) shortened sleep onset latency and/or increased total sleep time of 30 minutes¹⁵ Our review did not operationalize sleep and weight loss according to these standardized outcomes of clinical efficacy. Instead, we classified sleep and weight loss based on the design types applied in the reviewed studies: (1) sleep at baseline as a predictor of weight loss; (2) sleep after history of weight loss; (3) changes in sleep associated with weight loss; and (4) manipulation of sleep and weight loss. While these categories facilitated some level of inquiry, there is a need for prospective, randomized controlled trials that include no treatment control groups to model causal relationships between sleep and weight loss. In addition to robust, adequately powered designs, we recommend researchers include objective assessment of sleep through actigraphy, in conjunction with standardized sleep diaries, to elucidate relationships between sleep and weight loss. Interventions focused on modifying lifestyle factors should be rooted in behavioral theory, with the theoretical constructs operationalized and measured. Adopting this approach will allow researchers to model the relationships between theoretical constructs and empirical outcomes. Finally, we have included a Supplementary Appendix containing quality assessment data on each of the studies we reviewed that may help the reader to better understand the state of the rigor of the designs we encountered in our review The recommendations we propose will provide the foundation future investigations will require to estimate the effect sizes in weight loss and sleep studies—the next logical step for researchers seeking to conduct systematic reviews and meta-analyses.

CONCLUSION

Adequate sleep confers preventive benefits for mental health, physical health, and quality of life, and plays an important role in an individual’s day-to-day functional wellbeing. This scoping review provides synthesized evidence that baseline sleep characteristics may predict weight loss in studies of dietary interventions, thus pointing to the role that sleep can play in weight loss. However, few studies, most of which are based on small sample sizes and are underpowered, have focused on altering sleep to understand its role in weight loss. Additionally, there is evidence that sleep does not improve because of weight loss alone, and participants who complete or maintain successful weight loss do not compare to controls. Finally, experimental manipulations of sleep might alter body composition, with some evidence implicated in weight loss. Larger randomized control trials that focus specifically on sleep outcomes in weight loss interventions with diverse populations and longer-term follow-up to assess the impact of longer-term sleep hygiene on weight loss are needed to get a better and more comprehensive picture of the role sleep plays within weight loss.