Sleep during Infancy and Associations with Childhood Body Composition: A Systematic Review and Narrative Synthesis

Abstract

Introduction:

Prevention of childhood overweight should start as early as possible preferably in “the first 1000 days of life.” Sleep is one of the modifiable health behaviors during this age period, besides dietary intake and physical activity. The aim of this systematic review is to summarize the existing literature regarding the association between sleep during infancy (age ≤24 months) and body composition measures during childhood (age ≤12 years).

Methods:

We registered the protocol of this systematic review (PROSPERO registration no. CRD42018087088) and conducted the review following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement. We searched for articles published until July 31, 2019 reporting on longitudinal associations with a minimal follow-up of 6 months. Methodological Quality was assessed and a narrative synthesis was performed.

Results:

We included 19 studies. Sleep was reported as sleep duration (n = 18) or sleep problems (n = 2). Sleep was assessed at least once before the age of 12 months in 14 out of the 19 studies. Methodological quality was rated as strong for five studies, moderate for five studies, and weak for nine studies.

Conclusion:

This narrative synthesis found inconsistent evidence that longer infant sleep duration during the first 2 years of life is associated with a healthier body composition during childhood.

Introduction

Childhood obesity is becoming a permanent problem in our society even at the toddler age.^1,2 Because childhood obesity often tracks into adulthood,³ prevention should start as early as possible, preferably in the first 1000 days of life: between conception and second birthday.⁴ Childhood obesity risk factors during the first 1000 days that are modifiable and identified in a recent systematic review include: higher maternal prepregnancy BMI, prenatal tobacco exposure, maternal excess gestational weight gain, and accelerated infant weight gain.⁵ Less consistent evidence was found for curtailed infant sleep, bedtime routine with bottle feeding at age 9 months, bottle feeding at age 24 months, and introduction of solid food intake before age 4 months.⁵ Curtailed infant sleep and bedtime routine have recently been under the attention of pediatricians and youth health care workers and healthy sleep is part of newly developed lifestyle interventions during infancy.^4,6–10 Infant sleep problems can be problematic and include sleep latency and night awakenings, both leading to reduced sleep duration, quality, and efficiency (% of time spent in bed that the child really sleeps). Systematic reviews and meta-analysis so far have looked at the association and dose–response effects of sleep duration at age 0–4 years^11,12 and age 0–18 years^13–15 and concluded that there is an inverse association between sleep duration and weight gain. From these reviews we cannot conclude if infant sleep duration is a determinant of childhood overweight, as these reviews did not investigate infant age separately. Also, new studies in infants have been published that are not included in the last reviews. Infant sleep is different from childhood sleep in timing (frequent day-time naps, nighttime wakening's), initiation (feeding, pacifiers, room or bedsharing), and optimal measurements (diaries vs. accelerometry detecting the duration of laying and sleeping in bed or other places).

Infancy (age 0–2 years) is a defining period in life and overweight prevention should ideally start as early as possible. Sleep is one of the modifiable health behaviors during this age period, besides dietary intake and physical activity. The aim of this systematic review is to summarize the existing literature regarding the association of sleep duration and sleep problems during infancy (age ≤24 months) with body composition outcomes during childhood (age ≤12 years).

Methods

Protocol and Registration

We registered the protocol of this systematic review with the International Prospective Register of Systematic Reviews (PROSPERO; Registration no. CRD42018087088; available from www.crd.york.ac.uk/PROSPERO/display_record.php?ID=CRD42018087088. We conducted the review following the Preferred Reporting Items For Systematic Reviews and Meta-analysis (PRISMA) statement for reporting systematic reviews and meta-analysis.¹⁶

Information Sources and Search Strategy

We searched Ovid MEDLINE and EMBASE for articles published from inception until July 31, 2019 in collaboration with a clinical librarian (J.G.D.). Detailed search terms and MeSH terms are provided in Supplementary Table S1. The search strategy can be displayed as the following Boolean combination of concepts: ([children 0–2 years] AND [sleep duration] AND [body composition and adverse cardiometabolic outcomes]) OR [relevant cohorts| -studies]. No language limitation or filter was applied. In addition, we reviewed the reference lists from the relevant articles and earlier published review articles.

Study Selection

Eligible studies were longitudinal studies with a minimal follow-up time of 6 months, reporting the determinant sleep (either duration or problems) at any age between 0 and 24 months and as outcome BMI, another indicator of weight status up until 12 years of age. We included all definitions of sleep problems or different measurements of sleep quality or sleep duration. Two independent reviewers (M.W.H. and T.G.M.V., I.H., or K.O.A.) screened titles and abstracts using Rayyan, a web-app for screening titles and abstracts.¹⁷ Reviewers were blinded for each other's decision and discrepancies were discussed until consensus was reached. Full texts were screened for eligibility by M.W.H. and in case of doubt screened by T.G.M.V. as well. If more than one study analyzed associations in the same children at the same age, we selected the study that reported the most complete analysis for that age period, preferring continuous outcomes over binary outcomes and panel data over linear regressions on single determinants.

Data Extraction

Data extraction was performed by M.W.H. and checked by T.G.M.V. or I.H. for all articles, using a predefined data extraction table. Discrepancies were discussed until consensus was reached. We chose to present the adjustment model with the most complete set of covariates, also if this included confounders that could be considered potential mediators in a causal pathway. We additionally report confounders that were considered, but not included in the final model. Table 1 presents associations with BMI or weight-for-age outcomes, Table 2 associations with other body composition measures.

Table 2.

Summary of Observational Studies Assessing the Association between Sleep Duration and Indicators of Body Composition Other Than BMI/Weight Sorted by Methodological Quality Assessment and Age at Exposure

Author	Observational/trial (n, age at sleep measurement, follow-up time, cohort/study name, country)	Exposure (definition, cutoff values, method of measurement)	Outcome (age, type and method of measurement)	Analysis and adjustment for confounders	Relevant outcome/main results β (95% CI) OR (95% CI)	Observed significant associations per study for narrative synthesis	Quality assessment outcome
Derks²⁹	Observational (n = 4144, 2, 6, and 24 months, Generation R, the Netherlands)	24 Hours sleep duration (parental report)	FMI and FFMI SD scores (DEXA, internal standardization measured at age 6 years)	Linear regression, multiple imputation. Confounders: maternal education level, BMI and psychiatric symptoms, ethnicity, birth weight, duration of breastfeeding, TV watching, and baseline BMI.	Sleep duration at: FMI-SD: age 2 months: β −0.013 (−0.022 to −0.004) age 6 months: −0.005 (−0.015 to 0.005) age 24 months: −0.012 (−0.0035 to 0.011) FFMI-SD: 2 months: −0.007 (−0.017 to 0.003) 6 months: −0.010 (−0.022 to 0.001) 24 months β −0.019 (−0.042 to 0.004)	$F M I \frac{1}{3}$ $F F M I \frac{1}{3}$	Strong
Derks⁴⁰	Nonrandomized quasi experimental trial (n = 336, 2, 4, 12 months, PEAS Kids Growth Study, Australia)	24 Hours sleep duration (24-hour diary)	FMI (measured at age 4, 6, and 10 years by multifrequency BIA)	Multivariable linear regression Confounders: child sex, date of birth, child place of birth (Australia vs. other), maternal education, gestational age, BMI at birth, neighborhood SES, maternal BMI at child age 4 years.	For sleep duration at different months of age: FMI at age 4 years: 2 months: β −0.02 (−0.09 to 0.06) 4 months: β 0.02 (−0.05 to 0.09) 12 months: β 0.01 (−0.06 to 0.08) FMI at age 6 years: 2 months: β −0.02 (−0.10 to 0.06) 4 months: β 0.03 (−0.05 to 0.11) 12 months: β −0.08 (−0.25 to 0.09) FMI at age 10 years: 2 months: β 0.01 (−0.16 to 0.19) 4 months: β 0.05 (−0.11 to 0.21) 12 months: β −0.08 (−0.25 to 0.09)	$F M I \frac{0}{9}$	Strong
Collings²⁸	Observational, (n = 1338, 6, 12, 19, 24, and 36 months, Born-in-Bradford-1000, United Kingdom)	24 Hours sleep duration in five categories, combined from parental response at age 6, 12, 19, 24, and 36 months.	Sum skinfolds (left triceps, subscapular and thigh); % body fat; and waist circumference (Holtain calipers, BIA, measured at age 6, 12, 19, 24, and 36 months)	Mixed effects regression analysis with repeated measurements. Confounders: baseline age, follow-up time, gender, SES, parity, gestational age, birth weight, season of birth, maternal pregnancy age, smoking during pregnancy, and follow-up BMI, TV viewing, unhealthy snacking, and fruit and vegetable intake. Sensitivity analysis: breast feeding history, age introduced to solids, physical activity, bedtime regularity, maternal early pregnancy BMI, TV viewing after 6 pm, and height.	Change in SD per 1 SD higher sleep duration (compared with 12–13 h/day at 24 months) stratified by ethnicity, because of significant interaction by ethnicity. South Asian (n = 765): % body fat: β −0.029 (−0.053 to −0.00042) Waist circumference: β −0.0072 (−0.038 to 0.023) Sum skinfolds: β −0.013 (−0.059 to 0.033) White: small and nonsignificant results. All models unchanged in sensitivity analysis.	$% b o d y f a t : \frac{1}{2}$ $W a i s t c i r c u m f e r e n c e \frac{0}{2}$ $S k i n f o l d s \frac{0}{2}$	Strong
Klingenberg³²	Observational (n = 311, 9 and 18 months, SKOT, Denmark)	24 Hours sleep duration (parental report)	Sum of subscapular and triceps skin-folds (mm), % body fat, body fat mass (kg). (Harpenden skin-fold caliper and DEXA measured at age 3 years)	Linear regression Confounders: birth weight, gestational age, duration of breastfeeding, maternal smoking during pregnancy, maternal BMI at 9 months of examination, household income and highest educational levels of both parents at time. Daytime napping and nocturnal awakenings had no putative effect on sleep duration.	Sum of skin folds 9 months: β 0.289 (−0.16 to 0.73). 18 months: β −0.039 (−0.24 to 0.16). % body fat 9 months: β −0.001 (−0.003 to 0.001). 18 months: β 0.00005 (−0.001 to 0.001)	$S k i n f o l d s \frac{0}{2}$ $% B o d y f a t \frac{0}{2}$	Moderate
Diethelm³⁰	Observational (n = 481, 18 and 24 months, DONALD, Germany)	24 Hours sleep duration [consistently short (<13 hours) vs. consistently Long (>13 hours), parental report]	FMI, and FFMI gain (derived from skinfolds, Holtain skinfold caliper, measured between age 2 and 7 years)	Linear regression (comparing linear trend between sleep duration group Confounders: sex, birth year, birth weight (<3000 g) and rapid weight gain (0–18 months of age). (high maternal education, maternal overweight, age, gestational age, birth order fully breastfeeding, smoking in the household were considered but did not meet requirements to be included as confounders)	FMI: β 0.03 (SE 0.11, p = 0.04) FFMI: β −0.12 (SE 0.11, p = 0.8)	$F M I \frac{1}{1}$ $F F M I \frac{1}{1}$	Moderate
Taveras²⁵	Observational (n = 915, 6, 12, and 24 months, project Viva, USA)	24 Hours sleep duration (<12 vs. ≥12 hours) (weighted average of three data points, parental report)	Sum subscapular and triceps skinfold (mm, Holtain skinfold calipers, measured at age 3 years)	Linear regression Confounders:, maternal education, income, prepregnancy BMI, marital status, prenatal smoking history, race/ethnicity, birth weight, sex breastfeeding duration, weight-for-length z-scores at age 6 months, age, daily TV viewing and active play.	β 0.79 (0.18–1.40)	$S k i n f o l d \frac{1}{1}$	Weak
Wake⁶	Randomized controlled trial, n = 193, 7, 10, and 12 months and 6 years, Kids Sleep Study, Australia	Nighttime sleep duration (single predictor from four measurements, parental report)	Waist circumference (cm) (tape measured at age 6 years)	Random effects linear regression Confounders: gender, maternal education level, family language English (yes/no), SES area code	For 1 hour increase in sleep time: β −0.001 (−0.07 to 0.07 p = 1.0) No intervention effect (intervention delivered at 8 months) on waist circumference.	$W a i s t c i r c u m f e r e n c e \frac{0}{1}$	Weak
Taylor⁵³	Randomized clinical trial (n = 232, 12 and 24 months, POI, New Zealand)	24 Hours sleep duration (accelerometry, waist worn)	FFMI, % body fat (DEXA scan, age 5 years)	Linear regression models using isometric log-ratio coordinates Confounders: randomized controlled trial group, sex, primiparous; maternal education, household deprivation, ethnicity, maternal BMI, and concurrent BMI z-score	FFMI: 12 months: β −0.46 (SE 0.32) 24 months: β −0.32 (SE 0.35) % body fat: 12 months: β −1.24 (SE 1.96) 24 months: β −0.07 (SE 2.26)	$F F M I \frac{0}{2}$ $% B o d y f a t \frac{0}{2}$	Weak

BIA, bioelectrical impedance; DEXA, dual energy X-ray absorptiometry; FMI, fat mass index; FFMI, fat-free mass index; SES, socioeconomic status.

Table 1.

Summary of Studies Assessing the Association between Sleep and BMI or Weight for Age z-Score/Category Sorted by Methodological Quality Assessment, and Age at Exposure

Author	Observational/trial (n, age at sleep measurement, cohort/study name, country)	Exposure (definition, cutoff values, method of measurement)	Outcome (reference values, parental report/measured by trained staff, age)	Analysis and adjustment for confounders	Relevant outcome/main results β (95% CI) OR (95% CI)	Observed significant associations per study for narrative synthesis^a	Quality assessment
Kupers³³	Observational, (n = 2475, 4 months, GECKO, the Netherlands)	24 Hours sleep duration (parental report)	Weight-for-age z-score, (Dutch 1997 reference, measured between age 6–12 and 12–24 months).	Linear regression, multiple imputation for covariates, Confounders: paternal BMI, maternal prepregnancy BMI, age, diabetes and hypertension, educational level, gestational weight gain, smoking during pregnancy, gestational age, birth weight, gender, Dutch ethnicity, type of feeding at 3 months, child care and mother working at age 3 months, complementary feeding at 4 months, family screen time at 6 months, household income, possibility of unrestricted moving at 9 months, and one-parent family.	Weight-for-age z-score at: age 6–12 months: β −0.00011 (−0.00020 to −0.00002) age 12–24 months: β −0.00003 (−0.00008 to 0.00002)	$T o t a l \frac{1}{2}$	Strong
Derks²⁹	Observational (n = 4277, 2, 6, and 24 months, Generation R, the Netherlands)	24 Hours sleep duration (parental report)	BMI SD scores (Dutch reference, measured at age 6 years)	Linear regression, multiple imputation. Confounders: maternal education level, BMI and psychiatric symptoms, ethnicity, birth weight, duration of breastfeeding, TV watching, and baseline BMI.	Sleep duration at: age 2 months: β −0.018 (−0.026 to −0.009) (n = 3909) age 6 months: β −0.008 (−0.018 to 0.002) (n = 3293) age 24 months: β −0.008 (−0.027 to 0.011) (n = 4277)	$T o t a l \frac{1}{3}$	Strong
Derks⁴⁰	Nonrandomized quasiexperimental trial, (n = 314 2, 4, 12 months, PEAS Kids Growth Study, Australia)	24 Hours sleep duration (24-hour diary)	BMI score at age 4, 6, and 10 years (z-score in sensitivity analysis with similar results)	Multivariable linear regression Confounders: child sex, date of birth, child place of birth (Australia vs. other), maternal education, gestational age, BMI at birth, neighborhood SES, maternal BMI at child age 4 years.	For sleep duration at different months of age: BMI at age 4 years: 2 months: β −0.04 (−0.12 to 0.05) 4 months: β −0.01 (−0.09 to 0.06) 12 months: β 0.01 (−0.07 to 0.08) BMI at age 6 years: 2 months: β −0.04 (−0.13 to 0.06) 4 months: β 0.04 (−0.05 to 0.13) 12 months: β −0.02 (−0.12 to 0.09) BMI at age 10 years: 2 months: β 0.04 (−0.15 to 0.22) 4 months: β 0.10 (−0.06 to 0.27) 12 months: β −0.06 (−0.24 to 0.12)	$T o t a l \frac{0}{9}$	Strong
Zhou³⁷	Observational, (n = 799, 3, 6, 9, 12, 18, and 24 months, GUSTO, Singapore)	24 Hours sleep duration (BISQ)	BMI (measured at age 3, 6, 9, 12, 18, and 24 months)	Mixed model without random effects in which sleep duration and BMI values at all time points were entered, stratified by ethnicity. Confounders: sex, maternal education, household income, pregnancy smoking and diabetes, birth weight and length, gestational age, maternal BMI and height, breastfeeding duration, media use and physical activity at age 24 months, and age at measurements.	Overall β −0.014 (−0.030 to 0.001) Chinese: −0.006 (p > 0.05) Indian: 0.002 (p > 0.05) Malay: −0.042 (−0.071 to −0.012)	$T o t a l \frac{1}{4}$	Strong
Collings²⁸	Observational, (n = 1338, 6, 12, 19, 24, and 36 months, Born-in-Bradford-1000, United Kingdom)	24 Hours sleep duration, combined from parental report at age 6, 12, 19, 24 and 36 months.	BMI z-scores (U.K. reference, measured at age 6, 12, 19, 24 and 36 months)	Mixed effects regression analysis with repeated measurements. Confounders: baseline age, follow-up time, gender, SES, parity, gestational age, birth weight, season of birth, maternal pregnancy age, smoking during pregnancy, and follow-up BMI, TV viewing, unhealthy snacking, fruit and vegetable intake. Sensitivity analysis: breast feeding history, age introduced to solids, physical activity, bedtime regularity, maternal early pregnancy BMI, TV viewing after 6 pm, and height.	Change in SD BMI per 1 SD higher sleep duration (compared with 12–13 h/day at 24 months) stratified by ethnicity, because of significant interaction by ethnicity. South Asian (n = 756): β −0.030 (−0.061 to −0.00016) p ≤ 0.05 with significant interaction by follow-up time, suggesting that the inverse association strengthened over time. Reversed negative association with sleep duration as outcome. White: β −0.0047 (−0.042 to 0.033). Unchanged in sensitivity analysis.	$T o t a l \frac{1}{2}$	Strong
Sha³⁴	Observational (n = 519, 1, 3, 6, 8, and 12 months, China)	24 Hours sleep duration, sleep problems (hours; or yes/no by Zuckerman definition, parental report)	Weight-for-age z-score (WHO, measured, age 1, 3, 6, 8 and 12 months)	Pooled effects model. Confounders: (number of) breastfeeding (per day), (number of) formula feeding (per day), complimentary feeding, outdoor activities, taking vitamin D, birthweight, gender, maternal education level, household income, gestational age, maternal and paternal smoking during pregnancy.	Sleep duration: β −0.192 (SE 0.013, p < 0.001) Sleep problems: β −0.136 (SE 0.125, p = 0.278)	Not applicable	Moderate
Klingenberg³²	Observational (n = 311, 9 and 18 months, SKOT, Denmark)	24 Hours sleep duration (parental report)	BMI z-score (WHO, measured at age 3 years).	Linear regression Confounders: birth weight, gestational age, duration of breastfeeding, maternal smoking during pregnancy, maternal BMI at 9 months of examination, household income and highest educational levels of both parents at time. Daytime napping and nocturnal awakenings had no putative effect on sleep duration.	9 months: β −0.008 (−0.13 to 0.12) 18 months: β −0.010 (−0.07 to 0.05).	Not applicable	Moderate
Miyakoshi³⁵	Observational (n = 31.463, 18 months, Japan)	Nighttime sleep duration (parental report, categorized as short; middle (reference); long; or irregular)	Overweight (IOTF, measured at age 3 years)	Multiple logistic regression models Confounders: day care (maternal or relatives), smoking family members, older siblings, breastfeeding, introduction of solid foods, regular mealtime, use of precooked foods, consumption of cow's milk, sugar-sweetened beverages, sweet or snacks, maternal age at delivery, smoking and alcohol consumption during pregnancy, pregnancy-induced hypertension, gestational age, birth weight, sex of the child and past medical history of allergic diseases.	Short sleep duration OR 1.10 (0.95–1.28) Long sleep duration OR 0.79 (0.66–0.96) Irregular sleep duration OR 0.73 (0.51–1.06)	Not applicable	Moderate
Diethelm³⁰	Observational (n = 481, 18 and 24 months, DONALD, Germany)	24 Hours sleep duration (consistently short (<13 hours) vs. consistently long (>13 hours), parental report)	BMI z-score and category (German reference data/IOTF, measured between age 2 and 7 years)	Linear and logistic regression Confounders: sex, birth year, birth weight (<3000 g) and rapid weight gain (0–18 months of age). (high maternal education, maternal overweight, age, gestational age, birth order, fully breastfeeding, smoking in the household were considered but did not meet requirements to be included as confounders)	Gain in BMI between age 2 and 7: Consistent short vs. consistent long: β 0.073 (SE 0.18, p = 0.3) Overweight at age 7 years for consistent short vs. consistent long: OR 1.5 (0.8–2.8)	Not applicable	Moderate
Tuohino³⁶	Observational (n = 889, 3, 8, and 18 months, CHILD-SLEEP birth cohort, Finland)	24 Hours sleep duration: short vs. normal/long (lowest quartile), BISQ [8 months subsample with accelerometry (thigh-placed, n = 350)]	Overweight, defined as >internal 90th % of age- and sex-adjusted BMI (Finnish national reference data), 24 months)	Logistic regression Confounders: age, birthweight, sex, maternal early pregnancy BMI, parental education level, maternal smoking during pregnancy, and breastfeeding (all Finnish-speaking mothers in an ethnical nondiverse area)	3 months: OR 1.56 (1.02–2.38) 8 months: OR 0.78 (0.49–1.27) 18 months: OR 0.87 (0.52–1.46)	Not applicable	Weak
Taveras²⁵	Observational (n = 915, 6, 12, and 24 months, project Viva, USA)	24 Hours sleep duration (<12 vs. ≥12 hours) (weighted average of three data points, parental report)	BMI z-score and category, (CDC, measured at age 3 years)	Linear and logistic regression Confounders:, maternal education, income, prepregnancy BMI, marital status, prenatal smoking history, race/ethnicity, birth weight, sex breastfeeding duration, weight-for-length z-scores at age 6 months, daily TV viewing and active play.	BMI z-score: β 0.16 (0.02–0.29) Obesity (≥P95 vs. P5–85): OR 2.04 (1.07–3.91)	Not applicable	Weak
Taveras²⁴	Observational (n = 1046, 6, 12, and 24 months, Project Viva, USA)	24 Hours sleep duration (<12 vs. ≥12 hours, weighted average of three data points, parental report)	BMI z-score (CDC, measured at age 7 year) Obesity = ≥95th % vs. normal weight (<85)	Linear and logistic regression. Imputation for covariates (1 imputed value for each missing value) Confounders: maternal age, education, BMI and parity; household income, race/ethnicity, child television viewing at mid childhood	β 0.15 (0.02–0.28) Obesity (≥P95 vs. P5–85) OR: 1.36 (0.84–2.21)	Not applicable	Weak
Wang³⁸	Randomized Clinical Trial (n = 1704, 6; and 14 months, BeeBOFT, the Netherlands)	24 Hours sleep duration (parental report)	BMI z-score (WHO, measured at age 14 and 36 months)	Linear regression Confounders: age at sleep and BMI measurement; maternal age, education level, prepregnancy BMI, and parity; child gender, ethnic background, birth weight, gestational age, duration of breastfeeding, and screen time at baseline; intervention groups; and baseline BMI z-score.	6 months: β −0.005 (SE 0.010) 14 months: β −0.002 (SE 0.015)	Not applicable	Weak
Alamian²⁶	Observational (n = 895, 6 and 15 months, SECCYD, USA)	Sleep problems (yes/no at 6 and/or 15 months according to three definitions, parental report)	BMI category (CDC, measured at age 12 years)	Multinomial logistic regression Confounders: maternal education and poverty, family structure, ethnicity, birth weight, sex, and breastfeeding.	Overweight (P85–95 vs. <P85) per definition of sleep problems: Zuckerman OR 1.68 (1.11–2.55) Richman: OR 1.76 (1.05–2.97) Lozoff: OR 1.07 (0.66–1.74) No association with obesity (≥P95)	Not applicable	Weak
Hiscock³¹	Observational (n = 2741), 8.7 months, LSAC, Australia)	24 Hours sleep duration (2 days activity diaries)	BMI z-score (CDC, measured at age 2–3 years)	Linear regression. Confounders: sex, weight for age adjusted for birth length.	β −0.0002 (−0.0005 to 0.0001)	Not applicable	Weak
Bolijn²⁷	Observational (n = 1658, 2 years, KOALA, the Netherlands)	Nighttime and daytime sleep duration (parental report)	BMI-z-score and category (Dutch references, parental report age 5, 6, 7, 8, and 9 years)	Linear and logistic general estimating equation, persistence of associations over time of follow-up was evaluated by testing for interaction by age at the BMI measurement. Missings in covariates were imputed from all other variables using regression with addition of residuals from random cases. Confounders: general population or alternative (anthroposophic), maternal country of birth, age at birth, educational level and hours/week of paid work, prepregnancy BMI, smoking during pregnancy, pregnancy weight gain, child gender, time at kindergarten, TV time, computer time, and time playing outside, day time sleep (for night time sleep analysis), exact age of child at BMI measurement, and BMI at age 2 years.	Nighttime: β −0.082 (−0.120 to −0.044) Daytime: β −0.011 (−0.065 to 0.043) overweight/obesity (≥P85): Nighttime: OR 0.86 (0.76–0.97) Daytime: OR 1.10 (0.91–1.33) No significant interaction by age of BMI measurement	Not applicable	Weak
Wake⁶	Randomized controlled trial, n = 193, 7;10; and 12 months and 6 years, Kids Sleep Study, Australia	Nighttime sleep duration (single predictor from four measurements, parental report)	BMI z-score and category (CDC/IOTF measured at age 6 years)	Random effects linear and logistic regression Confounders: gender, maternal education level, family language English (yes/no), and SES area code	For 1 hour increase in sleep time: BMI z-score: β 0.001 (−0.01 to 0.01 p = 1.0) OR overweight/obese vs. normal: 0.9 (0.3–2.4) p = 0.8) No intervention effect (delivered at 8 months) on BMI z-score or overweight	Not applicable	Weak
Taylor⁵³	Randomized controlled trial (n = 292, age 12 and 24 months, POI, New Zealand)	24 Hours sleep duration (accelerometry, waist-worn)	BMI z-score (WHO, measured at age 5 years)	Linear regression models using isometric log-ratio coordinates Confounders: randomized controlled trial group, sex, primiparous, maternal education, household deprivation, ethnicity, maternal BMI, and concurrent BMI z-score	12 months: β −0.24 (SE 0.24) 24 months: β −0.43 (SE 0.28)	Not applicable	Weak
Zhang³⁹	Randomized controlled trial (n = 153, age 20 months, GET UP!Study, Australia)	24 Hours sleep duration; nap duration; nighttime sleep duration; sleep variability (actigraphy, waist-worn)	BMI z-score (internal sample standardization, measured at age 32 months)	Linear mixed models Confounders: clustering effects (modeled as random intercept), randomized controlled trial group, baseline age, gender, SES, and baseline z-BMI	24 Hours sleep duration: β −0.01 (−0.13 to 0.10); nap duration: β 0.41 (0.14–0.68); nighttime sleep duration: β −0.08 (−0.20 to 0.03); sleep variability: β 0.08 (−0.19 to 0.35)	Not applicable	Weak

Only studies with strong quality are included in the narrative synthesis on the association between sleep duration and BMI- or weight-for-age z-scores.

BISQ, Brief Infant Sleep Questionnaire; CDC, Centers for Disease Control and Prevention; CFI, Comparative Fit Index; CI, confidence interval; IOTF, International Obesity Task Force; OR, odds ratio; RMSEA, root mean squared error of approximation; SE, standard error; SES, socioeconomic status; WHO, World Health Organization; +, significant association; 0, no significant association.

Methodological Quality Assessment

For the assessment of the methodological quality of the studies, we used a tool adapted from the “Effective public health practice project quality assessment tool,” see Supplementary Data.¹⁸ We choose this tool as a specific tool on quantitative analysis in the public health field. We customized our criteria list and the corresponding judgment rules so they were fitting longitudinal cohort and interventional studies. We assessed the following five dimensions: selection bias, confounding, measurement, study attrition, and data analysis. Each dimension was judged as strong, moderate, or weak based on two questions. For the global quality rating of an article, no weak ratings resulted in a strong rating; one weak rating in a moderate rating; and two or more weak ratings in a weak rating. For a strong quality of the dimension confounding, the analysis had to adjust for at least: maternal BMI, socioeconomic status/maternal education/income, ethnicity, and birthweight/dysmaturity/baseline BMI. Studies that were not adjusted for maternal BMI were qualified as weak, regardless of other covariates, as we value maternal BMI as an important predictor for childhood BMI.¹⁹

The customized tool was pretested by assessing four articles by three independent researchers and improved by specifying sleep duration measurement methods. At the same time, researchers also improved their quality assessment skills by discussing discrepancies in these first four articles. Thereafter, all articles were assessed by two independent researchers (M.W.H. and T.G.M.V. or I.H.) using the tool as displayed in Supplementary Table S2. If needed, references were checked for study protocol details and selection procedures. Inconsistencies were discussed and resolved between the two researchers.

Narrative Synthesis

A best evidence synthesis was applied to summarize the results for the association between sleep duration and body composition outcomes and draw conclusions regarding the level of evidence (Table 3). We based the rating system on that of Chinapaw et al.,²⁰ consisting of four levels (i.e., strong, moderate, inconsistent, and insufficient). The levels of evidence take into account the number, the methodological quality, and the consistency of outcomes of the studies:

Table 3.

Level of Evidence in Narrative Synthesis, Presented Per Study, for the Association between Sleep Duration and Body Composition Measurement

	Body composition measurement	Conclusion of narrative synthesis, based on found associations with sleep duration expressed by $\frac{p}{t}$
BMI and weight-for-age z-scores Narrative synthesis from strong-quality studies only	BMI	Inconsistent evidence for an association $\frac{1}{2} \frac{1}{3} \frac{1}{2} \frac{0}{9} \frac{1}{3}$
Other body composition measurements Narrative study from strong, (moderate and weak) quality studies	FMI	Inconsistent evidence for an association $\frac{1}{3} \frac{0}{9} (\frac{1}{1})$
	FFMI	No evidence for an association $\frac{1}{3} (\frac{0}{1}) (\frac{0}{2})$
	Sum skin folds	No evidence for an association $\frac{0}{2} (\frac{0}{2}) (\frac{1}{1})$
	% body fat	No evidence for an association $\frac{1}{2} (\frac{0}{2}) (\frac{0}{2})$
	Waist circumference	No evidence for an association $\frac{0}{2} (\frac{0}{1})$

$\frac{p}{t}$ with p standing for positive associations found in a study and t standing for total tested associations in a study.

$\frac{p}{t}$ : strong-quality studies.

$(\frac{p}{t})$ : moderate- or weak-quality studies.

Strong evidence: consistent findings in multiple (≥2) high-quality studies.

Moderate evidence: consistent findings in one high-quality study and at least one moderate-quality study, or consistent findings in multiple (≥2) moderate-quality studies.

Inconsistent evidence: inconsistent findings in multiple (≥2) studies.

Insufficient evidence: only one study available.

Following the directions of Chinapaw et al.,²⁰ results were considered consistent when at least 75% of the studies showed results in the same direction and when the results were significant (p < 0.05). If there were two or more high-quality studies, we disregarded the studies of low methodological quality in the evidence synthesis. Associations are considered significant per study if 75% of the tests found significant associations in the same direction per outcome.

Results

Description of Included Studies

We identified 7539 unique records through database searches and 4 records through hand search. After titles and abstracts were screened, 7471 records were excluded (Fig. 1). We assessed 77 full-text articles for eligibility and excluded 58 full texts (reasons displayed in Fig. 1). Four studies were excluded after comparison with other publications from the same cohorts as they studied similar associations, but with binary outcomes or single outcomes vs. panel data.^7,21–23

Figure 1.

PRISMA flow diagram for the identification, screening, eligibility, and inclusion of studies. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analysis.

We included 19 studies in the systematic review (Tables 1 and 2). Two studies were based on analyses from the same cohort; project Viva,^24,25 presenting different follow-up times. Data across studies involved 51,963 unique participants. Studies were published between 2008 and 2019, with 11 of the 19 published in 2016 or later. All studies had a longitudinal design: 14 observational studies^24–37 and 5 randomized clinical trials^6,38–40,53 were included.

Sleep was reported as sleep duration (n = 18) or sleep problems (n = 2). Sleep was assessed at least once before the age of 12 months in 14 out of the 19 studies. Sleep duration was measured by parental report of sleep duration (n = 8), Brief Infant Sleep Questionnaire (n = 2); 24 hours activity diary (n = 2); and accelerometry (waist worn n = 2, thigh worn n = 1). Eight studies reported a combined sleep duration (deprivation score) for an age range. Sleep duration was reported as 24 hours sleep duration (n = 14); nighttime sleep duration (n = 3); both night- and daytime sleep duration separately (n = 1); nap duration (n = 1); or sleep variability (n = 2). Sleep problems were defined by the Zuckerman definition by two articles.^26,34 Alamian compared three sleep definitions: Zuckerman⁴¹; Richman⁴²; and Lozoff.^26,43

Methodological quality was rated as strong for five studies, moderate for five studies, and weak for nine studies. Weak ratings were often due to selection bias, low response rates at baseline, and poor study attrition at follow-up (Supplementary Table S3).

Association between Sleep Problems and Weight-for-Age z-Score or BMI Category

Two studies reported sleep problems using the Zuckerman definition, that is, waking three or more times per night, waking event lasting at least 1 hour, or parental report of “severe” disturbance. Using this definition, Sha et al. found no significant association between sleep problems at age 1, 3, 6, 8, and 12 months and weight-for-age z-score at the same ages in a pooled effects model.³⁴ Alamian et al. found that infants with sleep problems at age 6 and/or 15 months had 1.7 times higher odds of overweight at age 12 years [confidence interval [95% CI] 1.05–2.97], but they found no association with obesity.²⁶ With only these two studies from moderate and low quality we concluded there is no evidence for an association between infant sleep problems and childhood body composition.

Narrative Synthesis from Studies on Sleep Duration

We decided that a meta-analysis of the association between sleep duration and body composition measurements was not in place, due to high levels of heterogeneity in outcomes and ages. Instead we performed a narrative synthesis on the association between sleep duration and BMI or weight-for-age z-score or category (Table 1) and other body composition measures (Table 2). We did not include studies that assessed sleep problems as determinant. Found associations are displayed per study in Table 3.

Sleep Duration and BMI or Weight-for-Age z-Score or Category

With more than two high-quality studies available, we used only the five high-quality studies of Table 1 for this evidence synthesis. Four studies report inconsistent findings, with only significant associations for certain subgroups,^28,37 follow-up times,³³ or ages.²⁹ Therefore, we concluded there is inconsistent evidence for an association between infant sleep duration and childhood BMI or weight-for-age z-score/category.

Kupers et al. found that per 1 hour increase in 24 hours sleep duration at age 4 months, weight-for-age between 6 and 12 months decreased with 0.00011 (95% CI −0.00020 to −0.00002), but associations were smaller and not significant at age 12–24 months.³³ Derks et al. found that in the Generation R cohort per 1 hour increase in 24 hours sleep duration at age 2 months, there was a significantly lower BMI SD score at age 6 years. [−0.018 (95% CI −0.026 to −0.009)], but for sleep duration at age 6 and 24 months there was no significant association after adjustment for confounders and baseline BMI.²⁹ The same first author found no association in the PEAS Kids Growth Study, where sleep duration was measured at 2, 4, and 12 months by 24-hour diary and BMI measured at age 4, 6, and 10 years.⁴⁰ Zhou et al. found a significant negative association between sleep duration and BMI in Malay infants in their longitudinal mixed model and no significant association in Chinese and Indian infants in Singapore. In the overall group they found no significant association.³⁷ Collings et al. only found a significant association after stratification by ethnicity. They measured sleep at age 6–36 months and found that per one SD higher sleep duration in South Asians, BMI at the same age range decreased with 0.030 SD (95% CI −0.061 to −0.00016). For white families this association was smaller and nonsignificant.²⁸

Sleep Duration and Other Body Composition Measures

With less than two high-quality studies for all except one outcome, we used all eight studies of Table 2 for the narrative synthesis on other body composition measurements. There was inconsistent evidence for an association between sleep duration and Fat Mass Index. Derks et al. found in Generation R a significant inverse association with Fat Mass Index at one out of three ages at which sleep duration was measured.²⁹ The same first author found no association in the PEAS Kids Growth Study at three different ages at which sleep was measured and three ages at which Fat Mass Index was measured.⁴⁰ Diethelm et al. compared consistent short sleepers at age 18 and 24 months with consistent long sleepers and found that short sleepers had a 0.03 higher Fat Mass Index gain between age 2 and 7 years (standard error 0.11, p = 0.04).³⁰

There was no evidence for an association between sleep duration and fat-free mass index, % body fat, sum skin folds, and waist circumference (Table 3).

Discussion

This systematic review summarizes the existing evidence on the longitudinal association between sleep duration and sleep problems during infancy (<24 months) and indicators of body composition during infancy or childhood. We assessed the methodological quality of each study and performed a narrative synthesis. There was inconsistent evidence for an association between infant sleep during the first 24 months of life and BMI, weight-for age z-scores and Fat Mass Index during childhood. The significant associations reflect small effects; each hour off additional 24-hour sleep during infancy was associated with a 0.0001 to 0.0300 point lower BMI. There was no evidence for an association with fat-free mass index, sum of skin folds; % body fat; and waist circumference.

Comparison with Earlier Reviews

This systematic review summarizes all longitudinal studies on infant sleep duration (age period 0–2 years) and childhood overweight and obesity. Prior reviews of longitudinal data for sleep during age 0–4 years^11,12 and age 0–18 years^13–15 concluded that there is an inverse association between childhood sleep duration and weight gain, but that the underlying explanatory mechanisms are still uncertain. We cannot confirm this association in infants. Evidence for an association with sleep duration during infancy could also come from interventional studies. So far, results from three intervention studies on promotion of sleep quality and sleep duration during the first months of life have been published. Two of them showed healthier weight at age 1–2 years, but not necessarily due to longer sleep duration.^8,53

The first explanation for the lack of consistent evidence for an association is that sleep duration at infant age may only have a temporary effect on body composition, but not a longer-term effect. If so, this opens opportunity to change sleep habits at an early age to improve later healthy sleep habits and body composition. A second explanation for our findings can be the lack of high-quality studies. Methodological quality was strong for only 5 out of 19 studies. These cohorts were more representative of the base population from the start and had lower study attrition during follow-up. We encourage more high-quality studies on infant sleep and growth, specifically in the first 6 months of life, as existing studies have considerable risk of bias. A third explanation may be the different methods of sleep measurement across studies.

Measurement of Sleep During Infancy

The measurement of sleep duration and quality during infancy is complex due to a combination of longer nighttime stretches of sleep and shorter daytime naps and subtle differences in movements between awake stages and sleeping stages. The three most common methods of measurement are: parental report by questionnaire (e.g., Brief Infants Sleep Questionnaire); parent-reported 24-hour activity diaries; and accelerometry. A recent study by Tikotzky examined how well these three agree in the assessment of nocturnal wakefulness, but also sleep duration. They found that parents are only aware of nocturnal awakenings that involve signaling (e.g., crying), which makes parental report less valid after the age of 6 months when infants increasingly learn to self-soothe. They advise to use both objective and subjective measures of sleep, especially when measuring nocturnal wakefulness after the age of 6 months. They also found that sleep duration was more than 30 minutes longer based on diary report vs. accelerometry at the ages of 3, 6, 12, and 18 months. They additionally conclude that parents report infant sleep more accurately by diary than by the Brief Infants Sleep Questionnaire.⁴⁵ Accelerometry can provide an objective estimate of movement, but does not detect the difference between sleep and wakefully lying still in small infants. It is considered a valid method to estimate sleep duration in older children,^46,47 but under the age of 2 years, it is so far only valid for nighttime sleep duration.

Sleep patterns and habits during infancy and toddlerhood differ across cultures.^48,49 According to the National Sleep Foundation, appropriate sleep duration for healthy newborns (0–28 days after birth) ranges from 14 to 17 hours; for infants (0–12 months) from 12 to 15 hours; and for toddlers (1 and 2 years of age) from 11 to 14 hours.⁵⁰ The cutoff values for short sleep duration that are used in the studies in this systematic review vary and might have altered the findings of individual studies and therewith of our review and narrative synthesis.

Strengths and Limitations

A strength of this review is that we have performed a very thorough literature search guided by a clinical librarian (J.G.D.). Our search identified 19 longitudinal studies, of which only 6 were included in prior reviews. Our search for outcomes included all different body size measurements and body composition measurements. Another strength is that we selected longitudinal studies with a follow-up of at least 6 months, so we could study the sequence of events. Reversed causation was minimized by adjusting for baseline body composition in many studies. We could not investigate causal mechanisms due to a lack of studies that investigated these. We selected the most complete reported model of each study for data extraction, even if this included confounders that could be regarded as potential mediators in the causal pathway of an association between sleep and body composition outcomes. Ideally these covariates should be taken out of all models across studies, something that could be done in a future pooled analysis. A weakness of this review and narrative synthesis is heterogeneity of the studies. The measurements of the determinant (continuous 24-hour sleep duration) and outcomes (age- and sex-adjusted BMI or weight-for-age scores) are very similar in all five strong-quality studies in the narrative review, but not the age at which they were measured and the analyses did not consider the same confounders.

Implications

Despite the lack of evidence for an association between infant sleep and childhood body composition, attention to adequate sleep is important. Both parents and professionals need to be aware that adopting healthy sleep habits might be easier during early infancy than in later childhood and adulthood. The inconsistent finding in our narrative synthesis might be due to a lack of strong-quality studies and due to heterogeneity. There is a need for more strong-quality observational studies. We recommend future observational and interventional studies to incorporate other sleep measurements than sleep duration alone in their study design and to measure sleep at different ages (starting between age 1 or 6 months) with a combination of objective and subjective measures.^45,46,51 Objective measures are accelerometry or video recording⁴⁶ and subjective measures are 24-hour activity/sleep diaries and validated questionnaires, in which parents report sleep duration and regularity of bedtime as well as sleep quality, and sleep problems.⁵²

We assessed the methodological quality of each study with an adapted tool. Using this we have rated the validated questionnaires stronger than accelerometry, no matter of the specific outcome measurement. We recommend future researchers to qualify accelerometry in combination with parental report more accurate than parental report alone when nighttime or 24-hour sleep duration is measured after infant age of 6 months.⁴⁵ For the measurement of infant sleep problems, we recommend standardized questionnaires (e.a. Brief Infant Sleep Questionnaire).

Conclusion

This systematic review provides a thorough overview of studies investigating the longitudinal association of sleep during infancy (<24 months) and body composition during childhood. There was inconsistent evidence for a longitudinal association between infant sleep duration and child body composition.

Footnotes

Funding Information

This research was supported by the municipal Amsterdam Healthy Weight Program (Amsterdamse Aanpak Gezond Gewicht).

Author Disclosure Statement

No competing financial interests exist.

Supplementary Material

References

Hales

, Fryar

, Carroll

, et al. Trends in obesity and severe obesity prevalence in US youth and adults by sex and age, 2007–2008 to 2015–2016. JAMA, 2018; 319:1723–1725.

Worldwide trends in body-mass index, underweight, overweight, and obesity from 1975 to 2016: A pooled analysis of 2416 population-based measurement studies in 128.9 million children, adolescents, and adults. Lancet, 2017; 390:2627–2642.

Singh

, Mulder

, Twisk

, et al. Tracking of childhood overweight into adulthood: A systematic review of the literature. Obes Rev, 2008; 9:474–488.

Reilly

, Martin

, Hughes

. Early-life obesity prevention: Critique of intervention trials during the first one thousand days. Curr Obes Rep, 2017; 6:127–133.

Woo Baidal

, Locks

, Cheng

, et al. Risk factors for childhood obesity in the first 1,000 days: A systematic review. Am J Prev Med, 2016; 50:761–779.

Wake

, Price

, Clifford

, et al. Does an intervention that improves infant sleep also improve overweight at age 6? Follow-up of a randomised trial. Arch Dis Child, 2011; 96:526–532.

Taylor

, Gray

, Galland

, et al. Targeting sleep, food, and activity in infants for obesity prevention: An RCT. Pediatrics, 2017; 139. pii: e20162037.

Savage

, Birch

, Marini

, et al. Effect of the INSIGHT responsive parenting intervention on rapid infant weight gain and overweight status at age 1 year: A randomized clinical trial. JAMA Pediatr, 2016; 170:742–749.

Santos

, Bassani

, Matijasevich

, et al. Infant sleep hygiene counseling (sleep trial): Protocol of a randomized controlled trial. BMC Psychiatry, 2016; 16:307.

10.

Wasser

, Thompson

, Suchindran

, et al. Family-based obesity prevention for infants: Design of the “Mothers & Others” randomized trial. Contemp Clin Trials, 2017; 60:24–33.

11.

Chaput

, Gray

, Poitras

, et al. Systematic review of the relationships between sleep duration and health indicators in the early years (0–4 years). BMC Public Health, 2017; 17:855.

12.

Miller

, Kruisbrink

, Wallace

, et al. Sleep duration and incidence of obesity in infants, children, and adolescents: A systematic review and meta-analysis of prospective studies. Sleep, 2018; 41. DOI: 10.1093/sleep/zsy018.

13.

Ruan

, Xun

, Cai

, et al. Habitual sleep duration and risk of childhood obesity: Systematic review and dose-response meta-analysis of prospective cohort studies. Sci Rep, 2015; 5:16160.

14.

Fatima

, Doi

, Mamun

. Longitudinal impact of sleep on overweight and obesity in children and adolescents: A systematic review and bias-adjusted meta-analysis. Obes Rev, 2015; 16:137–149.

15.

Felso

, Lohner

, Hollody

, et al. Relationship between sleep duration and childhood obesity: Systematic review including the potential underlying mechanisms. Nutr Metab Cardiovasc Dis, 2017; 27:751–761.

16.

Moher

, Liberati

, Tetzlaff

, et al. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. J Clin Epidemiol, 2009; 62:1006–1012.

17.

Ouzzani

, Hammady

, Fedorowicz

, Elmagarmid

. Rayyan-a web and mobile app for systematic reviews. Syst Rev, 2016; 5:210.

18.

Thomas

, Ciliska

, Dobbins

, Micucci

. A process for systematically reviewing the literature: Providing the research evidence for public health nursing interventions. Worldviews Evid Based Nurs, 2004; 1:176–184.

19.

Manios

, Birbilis

, Moschonis

, et al. Childhood Obesity Risk Evaluation based on perinatal factors and family sociodemographic characteristics: CORE index. Eur J Pediatr, 2013; 172:551–555.

20.

Chinapaw

, Proper

, Brug

, et al. Relationship between young peoples' sedentary behaviour and biomedical health indicators: A systematic review of prospective studies. Obes Rev, 2011; 12:e621–e632.

21.

Fairley

, Santorelli

, Lawlor

, et al. The relationship between early life modifiable risk factors for childhood obesity, ethnicity and body mass index at age 3 years: Findings from the Born in Bradford birth cohort study. BMC Obes, 2015; 2:9.

22.

Gillman

, Rifas-Shiman

, Kleinman

, et al. Developmental origins of childhood overweight: Potential public health impact. Obesity, 2008; 16:1651–1656.

23.

Taylor

, Gray

, Heath

ALM

, et al. Sleep, nutrition, and physical activity interventions to prevent obesity in infancy: Follow-up of the Prevention of Overweight in Infancy (POI) randomized controlled trial at ages 3.5 and 5 y. Am J Clin Nutr, 2018; 108:228–236.

24.

Taveras

, Gillman

, Pena

, et al. Chronic sleep curtailment and adiposity. Pediatrics, 2014; 133:1013–1022.

25.

Taveras

, Rifas-Shiman

, Oken

, et al. Short sleep duration in infancy and risk of childhood overweight. Arch Pediatr Adolesc Med, 2008; 162:305–311.

26.

Alamian

, Wang

, Hall

, et al. Infant sleep problems and childhood overweight: Effects of three definitions of sleep problems. Prev Med Rep, 2016; 4:463–468.

27.

Bolijn

, Gubbels

, Sleddens

, et al. Daytime sleep duration and the development of childhood overweight: The KOALA Birth Cohort Study. Pediatr Obes, 2016; 11:e1–e5.

28.

Collings

, Ball

, Santorelli

, et al. Sleep duration and adiposity in early childhood: Evidence for bidirectional associations from the born in Bradford Study. Sleep, 2017; 40:01.

29.

Derks

IPM

, Kocevska

, Jaddoe

VWV

, et al. Longitudinal associations of sleep duration in infancy and early childhood with body composition and cardiometabolic health at the age of 6 years: The Generation R Study. Child Obes, 2017; 13:400–408.

30.

Diethelm

, Bolzenius

, Cheng

, et al. Longitudinal associations between reported sleep duration in early childhood and the development of body mass index, fat mass index and fat free mass index until age 7. Int J Pediatr Obes, 2011; 6:e114–e123.

31.

Hiscock

, Scalzo

, Canterford

, Wake

. Sleep duration and body mass index in 0-7-year olds. Arch Dis Child, 2011; 96:735–739.

32.

Klingenberg

, Christensen

, Hjorth

, et al. No relation between sleep duration and adiposity indicators in 9–36 months old children: The SKOT cohort. Pediatr Obes, 2013; 8:e14–e18.

33.

Kupers

, L'Abee

, Bocca

, et al. Determinants of weight gain during the first two years of life—The GECKO Drenthe Birth Cohort. PLoS One, 2015; 10:e0133326.

34.

Sha

, Yan

, Gao

, et al. Association between sleep and body weight: A panel data model based on a retrospective longitudinal cohort of chinese infants. Int J Environ Res Public Health, 2017; 14:25.

35.

Miyakoshi

, Yamamoto

, Mishina

, et al. Childcare environment and Japanese children who are overweight in early childhood. Child Obes, 2018; 14:197–206.

36.

Tuohino

, Morales-Munoz

, Saarenpaa-Heikkila

, et al. Short sleep duration and later overweight in infants. J Pediatr, 2019; 14:14.

37.

Zhou

, Aris

, Tan

, et al. Sleep duration and growth outcomes across the first two years of life in the GUSTO study. Sleep Med, 2015; 16:1281–1286.

38.

Wang

, Jansen

, Boere-Boonekamp

, et al. Sleep and body mass index in infancy and early childhood (6–36 mo): A longitudinal study. Pediatr Obes, 2019; 14:e12506.

39.

Zhang

, Pereira

, Sousa-Sa

, et al. The cross-sectional and prospective associations between sleep characteristics and adiposity in toddlers: Results from the GET UP! Study. Pediatr Obes, 2019:e12557.

40.

Derks

IPM

, Gillespie

, Kerr

, et al. Associations of infant sleep duration with body composition and cardiovascular health to mid-adolescence: The PEAS Kids Growth Study. Child Obes, 2019; 15:379–386.

41.

Zuckerman

, Stevenson

, Bailey

. Sleep problems in early childhood: Continuities, predictive factors, and behavioral correlates. Pediatrics, 1987; 80:664–671.

42.

Richman

. A community survey of characteristics of one- to two- year-olds with sleep disruptions. J Am Acad Child Psychiatry, 1981; 20:281–291.

43.

Lozoff

, Wolf

, Davis

. Sleep problems seen in pediatric practice. Pediatrics, 1985; 75:477–483.

44.

Paul

, Savage

, Anzman-Frasca

, et al. INSIGHT responsive parenting intervention and infant sleep. Pediatrics, 2016; 138. pii: e20160762.

45.

Tikotzky

, Volkovich

. Infant nocturnal wakefulness: A longitudinal study comparing three sleep assessment methods. Sleep, 2019; 42. DOI: 10.1093/sleep/zsy191.

46.

Galland

, Meredith-Jones

, Gray

, et al. Criteria for nap identification in infants and young children using 24-h actigraphy and agreement with parental diary. Sleep Med, 2016; 19:85–92.

47.

Werner

, Molinari

, Guyer

, et al. Agreement rates between actigraphy, diary, and questionnaire for children's sleep patterns. Arch Pediatr Adolesc Med, 2008; 162:350–358.

48.

Mindell

, Sadeh

, Kwon

, Goh

. Cross-cultural differences in the sleep of preschool children. Sleep Med, 2013; 14:1283–1289.

49.

Mindell

, Sadeh

, Kwon

, Goh

. Relationship between child and maternal sleep: A developmental and cross-cultural comparison. J Pediatr Psychol, 2015; 40:689–696.

50.

Hirshkowitz

, Whiton

, Albert

, et al. National Sleep Foundation's updated sleep duration recommendations: Final report. Sleep Health, 1:233–243.

51.

Sadeh

, Lavie

, Scher

, et al. Actigraphic home-monitoring sleep-disturbed and control infants and young children: A new method for pediatric assessment of sleep-wake patterns. Pediatrics, 1991; 87:494–499.

52.

Sadeh

. A brief screening questionnaire for infant sleep problems: Validation and findings for an Internet sample. Pediatrics, 2004; 113:e570–e577.

53.

Taylor

, Haszard

, Meredith-Jones

, Galland

, Heath

ALM

, Lawrence

, et al. 24-hour movement behaviors from infancy to preschool: cross-sectional and longitudinal relationships with body composition and bone health. Int J Behav Nutr Phys Act, 2018; 15:118.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB