Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Behavioral, contextual and biological factors associated with obesity during adolescence: A systematic review

  • Janine Narciso ,

    Roles Conceptualization, Methodology, Writing – original draft

    janine.narciso@ubi.pt

    Affiliation Department of Sports Sciences, University of Beira Interior, Covilhã, Portugal

  • António José Silva,

    Roles Funding acquisition, Writing – review & editing

    Affiliations Department of Sports, Exercise and Health Sciences, University of Trás-os-Montes e Alto Douro, Vila Real, Portugal, Research Center in Sports Sciences, Health Sciences and Human Development, CIDESD, Vila Real, Portugal

  • Vitor Rodrigues,

    Roles Writing – review & editing

    Affiliations Research Center in Sports Sciences, Health Sciences and Human Development, CIDESD, Vila Real, Portugal, Superior School of Health, University of Trás-os-Montes e Alto Douro, Vila Real, Portugal

  • Maria João Monteiro,

    Roles Funding acquisition, Writing – review & editing

    Affiliations Superior School of Health, University of Trás-os-Montes e Alto Douro, Vila Real, Portugal, Center for Health Technology and Services Research, CINTESIS, Porto, Portugal

  • António Almeida,

    Roles Writing – review & editing

    Affiliation Superior School of Health, University of Trás-os-Montes e Alto Douro, Vila Real, Portugal

  • Raquel Saavedra,

    Roles Writing – review & editing

    Affiliation Department of Sports, Exercise and Health Sciences, University of Trás-os-Montes e Alto Douro, Vila Real, Portugal

  • Aldo Matos Costa

    Roles Conceptualization, Funding acquisition, Methodology, Writing – review & editing

    Affiliations Department of Sports Sciences, University of Beira Interior, Covilhã, Portugal, Research Center in Sports Sciences, Health Sciences and Human Development, CIDESD, Vila Real, Portugal, Health Sciences Research Center, CICS-UBI, Covilhã, Portugal

Behavioral, contextual and biological factors associated with obesity during adolescence: A systematic review

  • Janine Narciso, 
  • António José Silva, 
  • Vitor Rodrigues, 
  • Maria João Monteiro, 
  • António Almeida, 
  • Raquel Saavedra, 
  • Aldo Matos Costa
PLOS
x

Abstract

Introduction/Objective

Adolescence is a critical period for the development of obesity. Obesity arises from a complex interaction between several factors, which are not yet fully understood. Therefore, the purpose of this review was to identify and assess the peer-reviewed scientific literature on the behavioral, contextual and biological factors associated with obesity in adolescents.

Methods

PubMed and Scopus were systematically searched to identify prospective cohort studies concerning the relation between behavioral, contextual and biological factors and obesity in adolescents aged 10 to 18 years.

Results

40 studies published between the year 2000 and 2018 were included. A positive consistent association between genetic factors and obesity during adolescence was found. Also, there is evidence to support the association between socioeconomic status and obesity. There was conflicting evidence for the contribution of dietary intake, physical activity, sedentary behavior, sleep, food store environment, school food environment. For the remaining factors no associations were found, or no conclusions could be drawn due to the limited number of studies identified.

Conclusions

Further prospective studies that assess multiple obesity determinants simultaneously and use state-of-art measures are warranted to aid in the development of effective strategies and interventions to prevent obesity during adolescence.

Introduction

Childhood obesity has been recognized as one of the major public health challenges of the 21st century.[1] The number of obese children and adolescents aged 5 to 19 years increased from 11 million in 1975 to 124 million in 2016 and, an additional 213 million were overweight in 2016.[2] Two recent studies show that the rise in obesity in children and adolescents appears to be levelling off in high-income countries—albeit at high prevalence levels—but persists in some low and middle-income countries.[2,3]

Children and adolescents are exhibiting obesity-related health conditions such as type 2 diabetes, insulin resistance, hypertension, dyslipidemia, obstructive sleep apnea and fatty liver disease, which were previously only seen in adults.[4,5] In addition, obesity diminishes adolescents’ quality of life[6] and obese or overweight adolescents have a higher risk of becoming obese and/or centrally obese in adulthood.[7]

Adolescence is one of three critical periods for the development of obesity.[8] This might be explained by the fact that adolescence is a period characterized by changes in the quantity and location of body fat, a decline in physical activity and diet quality and an increase in sedentary behaviors. Additionally, from a psychosocial perspective, it is considered a period of increased risk of depression and anxiety.[9] Therefore, primary prevention of obesity in adolescents should be a public health priority.

Obesity results from a long-term energy imbalance between nutritional intake and activity, with the last being affected by physical activity (PA) and sedentary behavior.[10,11] Hence, it is no surprise that these behavioral factors have been the major focus of childhood obesity prevention and treatment interventions.[12] However, obesity arises from a complex interaction between genetic, epigenetic, environmental and behavioral factors, which are not yet fully understood.[1315]

Due to the complexity of obesity, social ecological models have been proposed for its understanding, as they are useful in examining a wide range of factors that contribute to a complex health issue.[16] The conceptual framework adopted in this review was guided by both Faith and Kral conceptual model[17] and the Identifying Determinants of Eating and Activity study conceptual model for the etiology of childhood obesity.[16] Faith and Kral model[17] hypothesizes that genetic and social-environmental factors lead to the development of obesity through their independent influences on intermediary behavioral phenotypes (food intake and PA). These intermediary behavioral variables possibly will induce a positive energy balance that, if sustained, will promote obesity. The Identifying Determinants of Eating and Activity study conceptual model posits that obesity risk is most closely and directly affected by diet, activity-related and biological factors. It also posits that diet, activity and sedentary patterns impact both, biological factors and energy balance and, at the same time diet and activity levels are impacted by numerous contextual factors.[16]

There is a notable demand for a better understanding of the factors that influence obesity in adolescents to develop interventions to prevent and manage obesity in this crucial period of life. To the extent of our knowledge, no study has systematically reviewed the literature on the factors associated with the development of obesity focusing solely on adolescents. Thus, the purpose of the present systematic review was to identify and assess the peer-reviewed scientific literature on the behavioral, contextual and biological factors associated with obesity in adolescents aged 10 to 18 years.

Methods

Literature search strategy

PubMed and Scopus were systematically searched to identify relevant articles published between January 2000 and April 2018. The literature search was conducted between February and April 2018.

A combination of terms involving the population (adolescents), the exposure (biological, contextual and behavioral factors) and the outcome (overweight or obesity or body mass index (BMI)) was used. In PubMed, terms involving the study design (longitudinal, prospective or cohort or follow-up) were also combined with the previous terms, while in Scopus the limit to keyword feature was applied. The search terms are presented in Table 1.

thumbnail
Table 1. Search terms of the population, exposure and outcome used in PubMed and scopus.

https://doi.org/10.1371/journal.pone.0214941.t001

In addition, the reference lists of the selected studies and relevant reviews were hand searched to identify further relevant articles.

Inclusion and exclusion criteria

For inclusion, articles were required to (1) be published in peer-reviewed journals between January 2000 and April 2018, (2) be published in English, (3) have a prospective cohort design, (4) include only adolescents between the ages of 10 and 18 years old, (5) to measure both, the outcome and the exposure of interest during the age range mentioned on point 4 (with the exception of infant feeding practices) and (6) use BMI or overweight/obesity (assessed using BMI) as the outcome measure.

No restriction on the method for assessing the exposure was considered.

Studies were excluded if they used reported height and weight and if subjects presented any medical conditions. Studies were also excluded if children and adolescents were examined as a total group and no separate analysis for the adolescent population existed.

Quality assessment

The quality assessment of selected studies was performed by one reviewer (JN) using the Newcastle-Ottawa Scale[18] and revised by a second reviewer (AC). Studies were scored based on selection of study groups (four criteria), comparability of study groups (one criteria), and ascertainment of the outcome (three criteria). The final score ranged between 0 to 9 stars, with 9 representing the highest methodological quality. We considered the studies to be high, medium and low quality if the scores ranged between 7–9, 4–6 and 1–3, respectively. The quality assessment details for each study can be found in Table A in S2 Appendix.

Results

Study selection

The 5092 studies retrieved from both databases and the 54 studies retrieved from handsearching were exported to Mendeley (Version 1.17.13) where duplicates were excluded (n = 574). To identify relevant studies for the review a two-step process was used: first the information contained in the title and abstract was screened; and second if the abstracts did not provide enough information or were not available, the full-text articles were screened based on the inclusion-exclusion criteria previously described. One reviewer (JN) conducted the database search and study selection. However, if any doubts occurred during the selection process a second reviewer (AC) was consulted and a collaborative decision to include or exclude the study was made. Even though two studies[19,20] used the same sample to evaluate the relationship between neighborhood food environment and obesity, we decided to retain them as they measured the neighborhood food environment differently. We also decided to include one study[21] in which a minimal portion of the sample (n = 10) had height and weight measured by the parents using the investigators’ instructions. At the end of the selection process 40 articles were included in the review. A flow diagram of the selection process can be found in Fig 1.

Characteristics of the studies included

The studies included in the present review were published between 2001 and 2017, with most of studies being published during or after 2010 (n = 33). Of the 40 studies included, the majority were conducted in North America (n = 15), followed by Europe (n = 10) and Australia (n = 8). The sample size ranged from 103[22] to 7090[19] participants and the follow-up period from 1[23,24] to 18 years[25]. Two studies included only females[26,27] and one study only males.[25]

Out of the 40 studies included, 26 [2023,26,2848] used BMI as their outcome, 6[24,27,4952] used overweight/obesity, while 8 [19,25,5358] studies used both measures. In 13 studies, body fat either expressed in percentage [25,26,28,29,33,34,38,43,44,53,56] or in kg[31,32] was used as the outcome alongside BMI or overweight/obesity. Furthermore, measures of fat mass index, fat free mass index, waist circumference (WC) and skinfold thickness were also concurrently used with BMI or overweight/obesity.

Regarding quality assessment, 21[23,2529,31,33,35,36,40,41,4345,48,5155] studies were classified as high quality and 19[1922,24,30,32,34,3739,42,46,47,49,50,5658] as medium quality. Studies final scores ranged from 4 to 8 stars.

A summary of all the studies can be found in Table 2 and the description of individual studies can be found in Table A in S3 Appendix.

Relation between behavioral factors and obesity

Dietary intake.

In this review sixteen studies[23,26,2836,49,5356] examining the association between dietary intake and obesity during adolescence were included.

Cohen et al[26] examined whether changes in diet over a 2-year period were associated with changes in BMI and body fat in 265 girls and found that for every increase of 100 calories over the baseline consumed per day, girls gained 0.4 BMI percentiles points. However, no evidence of association between BMI and calories consumed from carbohydrates, protein and fat was found. Enes and Slater[23] found no significant associations between one-year change in BMI z-score and both, total energy and fat intake. However, a significant positive association was found for fatty food intake.

Gopinath et al[28] followed 1213 12 years old students for 5 years and reported that each 1-SD increase in dietary fiber intake was associated with a concurrent 0.44kg/m2 decrease in mean BMI in girls and a 1.45 cm decrease in WC in boys. Additionally, in girls each 1-SD increase in carbohydrates and dietary glycemic load was associated with an increase of 0.77kg/m2 in BMI. Conversely, Murakami et al[29] found that changes in glycemic load were not associated with concurrent changes in BMI or any other body composition measures in either boys or girls. In both studies[28,29], no association with baseline glycemic index and changes in BMI and percentage of body fat (%BF) were detected. However, Murakami et al[29] found that glycemic index was associated with increased change in fat mass index between the ages 12 and 15 years and that results regarding glycemic index and BMI (tertiles of baseline GI: 2.03(1.74, 2.32); 2.09 (1.80, 2.38) and 2.42 (2.13, 2.71) kg/m2; Ptrend = 0.06) appear to point in the same direction.

Research regarding the consumption of dairy products was conducted by Bigornia et al[53] and Lin et al[30]. The first study[53] examined the association between dairy consumption at 10 years of age and risk of excess adiposity 3 years later. The authors found no association between total and reduced-fat dairy intake and risk of overweight or excess fat mass. Even though, in many cases the results did not achieve significance they showed inverse relations between dairy intakes and risk of excess adiposity, particularly for full-fat dairy products. However, the highest versus the lowest consumers of full-fat dairy products had smaller gains in BMI during follow-up and appeared to have a reduced risk of overweight at 13 years of age (OR: 0.65; 95% CI:0.40, 1.06; P = 0.19). The second study[30] using imputed data for 5,968 Chinese adolescents with BMI z-score, reported that neither non-milk dairy products nor milk consumption at 11 years was associated with BMI z-score at 13 years of age.

Three[31,32,54] of the seven[23,28,33,55] studies that looked at the association between sugar-sweetened beverages (SSB) and BMI found significant associations. Ludwig et al[54] reported that for each additional serving of SSB consumed, both BMI and odds of being obese increased. Likewise, results from The Avon Longitudinal Study of Parents and Children [31] showed that increased SSB consumption from ages 10 to 13 years was associated with higher BMI, WC and total body fat mass at 13 years, with the effects being strengthened among plausible dietary reporters Feeley et al[32] found a positive association between SSB consumption and both, BMI z-score and fat mass in South African males. It is important to note that Gopinath et al[28] found that girls who consumed one or more soft drinks/day compared to those who never/rarely consumed soft drinks at baseline had a greater %BF, BMI (3.20 versus 1.96 kg/m2, P = 0.01) and WC (10.00 versus 6.46 cm, P = 0.004) after 5 years. However, the trend was nonsignificant for BMI and WC. Additionally, Laska et al[33] reported that among males there was some evidence for a positive association between SSB consumption and both %BF and BMI (β = 0.27, P = 0.008). However, the last was not significant as α = 0.003125 was used to indicate statistical significance.

Studies regarding fast food consumption and obesity in adolescents have yield mixed results. Fraser et al[56] reported that increased consumption of fast food at age 13 years was positively associated with BMI SD scores, body fat percentage and being obese at age 15 years. However, three[3234] other studies found no association between fast food or take-away food consumption and BMI.

Four studies[32,33,35,49] that examined the relationship between breakfast consumption and obesity in adolescence were identified. Elgar et al[35] found that skipping breakfast in Year 7 was positively associated with BMI in Year 11. Likewise, Wang et al[49] found that the odds of being overweight or obese was significantly higher for students that frequently skipped breakfast compared with those classified as double breakfast eaters. Two studies[32,33] found no association between breakfast consumption and BMI or fat mass.

No evidence of association was found between fruits/vegetables intake and BMI.[23,26,36]

Infant feeding practices.

Two studies[25,50] that examined the relationship between breastfeeding and obesity were identified in our search.

Shields et al[50] examined the association between the duration of breastfeeding and being overweight or obese at 14 years of age in 3698 adolescents. They found that the risk of obesity was reduced for those breastfed for longer than six months, however when early life and 14 years confounders were taking into consideration the effect size diminished and lost statistical significance (OR:0.8, 95% CI: 0.5, 1.3). Even though no significant results were reported for risk of overweight, there was a trend for breastfeeding for less than 4 months to be associated with an increased prevalence of overweight at 14 years of age. Victora et al[25] explored the relationship between the duration of breastfeeding and several anthropometric measures in 18-year-old males. Neither the duration of total breastfeeding nor that of predominant breastfeeding showed consistent associations with any anthropometric or body composition measures. However, after adjustment for confounding variables (family income, maternal education at birth, maternal BMI, skin color, birth weight, gestational age, maternal smoking during pregnancy, and current behavioral variables -smoking, alcohol drinking, type of diet and physical exercise) a linear trend for a reduction in obesity with increasing duration of predominant breastfeeding remained.

Physical activity.

This review identified nine articles[23,26,27,3539,51] that examined the association between PA and obesity during adolescence.

Elgar et al[35] followed 355 adolescents aged 11 to 14 years for 4 years and found that the amount of time playing sports or exercising at baseline was negatively associated with changes in BMI, but it was not predictive of BMI at follow-up. Schuster et al[51] reported that in the bivariate analysis, overweight 5th graders were less likely to become obese at 10th grade if they performed more vigorous exercise. Furthermore, Bélanger et al[37] examined the association between the variation in the number of weekly moderate-to-vigorous PA (MVPA) sessions for 5 years and BMI, WC and skinfolds at follow-up in 756 adolescents. They found that in boys, higher PA fluctuation scores were associated with higher BMI and triceps skinfold thickness, while in girls higher PA fluctuation scores were associated with lower BMI, WC and skinfold thickness. Barnett et al[38] using the same sample as the previous study found that rates of decline in vigorous PA (VPA) were associated with declines in BMI during earlier adolescence in boys and slightly associated with declines in BMI during later adolescence in girls. However, MVPA and VPA were inversely associated with body fat percentage during earlier adolescence in girls and MVPA was inversely associated with body fat percentage during later adolescence in boys. de Souza et al[36] also contrary to what was expected, found that total PA had a positive association with girls’ BMI trajectories. Two studies[23,39] found no associations between PA and BMI.

Only two cohort studies[26,27] that used accelerometry to measure PA were identified. The first study[27] explored the associations between PA and 3 measurements of obesity (Centers for Disease Control and Prevention definition of obesity, the International Obesity Task Force reference BMI cut points for obesity in children and the sums of skinfold thickness—with the cohort ≥90th percentile as indicative of obesity) among 1148 black and white adolescent girls during a 2-year period. They reported a strong negative dose-response association between quartiles of accelerometer counts per day at 12 years of age and all the 3 measurements of obesity at age 14 years, but only for white girls. The second study[26] followed a cohort of adolescent girls for 2 years and found that every additional minute of increase in VPA per day was associated with a decline of body fat by approximately 0.1%. The association between BMI and VPA (β = -0.06, 95%CI: -0.28,0.16) pointed in the same direction.

Sedentary behavior.

Seven studies[23,26,35,3941,51] that examined the relationship between sedentary behavior (SB) and obesity were identified in our search.

Dumith et al[40] studied the effect of screen-time change (television (TV) + video game + computer) from age 11 to 15 years on several anthropometric measures at age 15 years and found that adolescents who increased their screen-time, had an increase in BMI, skinfold thickness and WC. Similarly, results from a cohort of Welsh adolescents[35] found that the number of hours per week watching TV or videos or playing computer games in Year 7 predicted BMI in Year 11. However, these sedentary activities did not predict change in BMI. Schuster et al[51] followed 3961 students for 5 years and reported that overweight 5th graders who watched more TV were more likely to become obese. Conversely, Fletcher et al[41] followed a cohort of 285 Australian adolescents for approximately 2 years and found no evidence of direct or indirect prospective associations for TV viewing, objectively measured total SB and average sedentary bout duration at baseline with BMI z-score at follow-up. Cohen et al[26] also, found no evidence of association between change in objectively measured SB and changes in BMI percentile, in girls. Furthermore, two studies found no significant association between self-reported screen time and BMI.[23,39]

Sleep.

This review identified five articles[24,36,4244] that explored the association between sleep and obesity.

Schafer et al[42] assessed the relationship between sleep duration trajectories and BMI, fat mass and fat-free mass at 18 years old in 3974 adolescents. They found that girls who reported inadequate sleep duration (<8hours/day) at age 11, but adequate sleep duration (≥8hours/day) at 18, had an increase of 0.39 z-scores and 0.30 z-scores in BMI and fat mass index, respectively compared to those who always had adequate sleep duration. Araújo et al[43] studied 1171 adolescents at both 13 and 17 years of age and found that reported sleep duration at age 13 was inversely associated with BMI z-scores and %BF 17 years in boys. In girls sleep duration at age 13 was positively associated with the changes in BMI z-scores between 13 and 17 years of age. However, after adjusting for adiposity at baseline, the prior associations were no longer statistically significant. Roberts and Duong[24] examined the reciprocal association between sleep restriction (≥6hours/night) and obesity in 3134 adolescents followed over one-year period and found that sleep restriction did not increase the future risk of obesity. It is worth mentioning that the odds for short sleep all week was 1.85, but the confidence interval was relatively large (95% CI: 0.91; 3.73).Moreover, Lytle et al[44] combining data from two American studies, showed that there was no statistically significant relationship between change in total sleep and change in BMI or %BF over time in both genders. The only longitudinal relationship that approach statistical significance was a positive association between sleep and %BF in females (b = 0.268, 95% CI: -0.02, 0.56, P = 0.068). Similarly, de Souza et al[36] did not find any significant association between sleep duration and BMI trajectories in either boys or girls.

Relation between contextual factors and obesity

Neighborhood food environment.

This review identified two studies[19,20] that assessed the relationship between the neighborhood food store environment and obesity during adolescence.

Chen and Wang[19] found, amongst a nationally representative cohort of American students followed from 5th to 8th grade, that girls living in neighborhoods with 2 or more supermarkets (based on ZIP code) had a lower BMI by 0.6 kg/m2 three years later, in comparison with those living in places without supermarkets. Also, for girls a greater quantity of limited-service restaurants in a neighborhood was associated with both, a greater BMI and greater odds of being obese 3 years later. On the contrary, Shier and Sturm[20] used the same sample as the previous study and found no association between two measures of the neighborhood food environment (counts of a specific type of food outlet per 1000 population and food environment indices) and BMI percentile in 8th grade or change in BMI percentile from 5th to 8th grade. Additionally, using a third measure of the neighborhood food environment (indicators for the presence of specific combinations of types of food stores) the authors reported that the existence of more types of food outlets in an area, including supermarkets, was associated with a higher BMI.

School environment.

The association between school environment and obesity was evaluated in four studies[26,46,55,58].

Cunningham and Zavodny[55] investigated the relationship between the availability and purchases of sweetened beverages at school and obesity among 5th and 8th graders. They found that the availability of sweetened beverages at school did not lead to heavier weight or greater risk of overweight or obesity among American adolescents. The second study[46] examined the association between school vending machines availability and BMI changes from 5th to 8th grade among students’ subgroups. An increase in vending machine availability was associated with an increase of 0.46 and 0.40 units in BMI for Hispanic males and for low-income Hispanic students, respectively.

Wardle et al[58] examined the impact of school physical education (PE) on BMI and WC change over 5 years and found no significant effects of the number of PE classes scheduled on changes in BMI or among overweight and obese adolescents when compared to their normal-weight counterparts. However, after adjusting for confounder variables, boys in schools with three weekly PE sessions had a. smaller increase in WC over the 5 years than those in schools providing one or two sessions. Differences in girls were in the same direction but did not achieve statistical significance. Cohen et al[26] following a cohort of girls, also found no association between participation in school PE classes and BMI changes over 2 years. Surprisingly, a positive association between school PE and %BF was found.

Socioeconomic status.

This review identified five studies[35,39,51,52,56] that explored the association between socioeconomic status (SES) and obesity.

Assunção et al[52] explored whether socioeconomic position (based on a list of 18 socioeconomic indicators) at age 11 years was associated with entry and exit from obesity between the ages of 11 and 15 years. They found that girls in the lowest tertile of socioeconomic position had a greater probability of becoming obese and that for boys, high socioeconomic position was a predictor of ceasing to be obese. Similarly, Schuster et al[51] reported that 5th graders with a lower household education were more likely to remain and become obese from 5th to 8th grade, however after adjusting for confounders the last result was no longer significant (OR:1.83, 95% CI: 0.88, 3.79). Results from The Avon Longitudinal Study of Parents and Children [56] also showed a positive relationship between increasing deprivation and both increased BMI and odds of being obese at age 15 years. Contrarywise, Elgar et al[35] reported that most SES variables in Year 7 did not predict a change in BMI over 4 years, except for the amount of money earned, while Aires et al[39] found no association between SES based on parent’s education level and risk of being overweight/obese over a period of 3 years.

Other factors.

Due to the limited number of studies, there was inconclusive evidence for parental feeding practices[21], self-control[22], weight loss dieting[52], home and neighborhood environment[45] and family environment[57].

Relation between biological factors and obesity

Genetic factors.

Three studies[47,48,51] that look at the association between genetic factors and obesity in adolescents were identified in the present review.

Cornes et al[47] examined whether there were qualitative and quantitative differences in genetic and environmental influences affecting BMI longitudinally in males and females using a sample of 1143 twin pairs at ages 12, 14 and 16 years. They found that most of the variation in BMI at all three ages was explained by additive and/or non-additive genetic influences for both genders. Liem et al[48] explored the association of common genetic variants near the INSIG2, in the FTO and near the MC4R genes and changes in BMI (3 time points) and skinfold thickness (2 time points) z-scores. Variation in the FTO gene was associated with BMI and sum of skinfolds longitudinally. In addition, variation near the MC4R gene was also associated with BMI. The last study[51] estimated the heritability of BMI between parents and their children and reported that overweight 5th graders were more likely to become obese if they had an obese parent at baseline. Furthermore, obese 5th graders were less likely to exit obesity if they had an obese parent.

Discussion

This review included 40 prospective studies that explored the association between behavioral, contextual and biological factors and obesity in adolescents aged 10 to 18 years. There is evidence to support the association between genetic factors and SES with obesity. There was conflicting evidence for the contribution of dietary intake, PA, SB, sleep, food store environment and school food environment. Moreover, due to the limited number of studies identified regarding parental feeding practices, self-control, weight loss dieting, neighborhood environment and family/home environment, no reasonable conclusions could be drawn. No associations were identified for the remaining factors.

There is a consensus that obesity results from a long-term positive energy balance where dietary intake surpasses energy expenditure for metabolic processes and physical activity.[59] However, we found inconsistent evidence for the link between obesity and both, dietary intake and physical activity during adolescence. Similarly, two systematic reviews, one focusing on children and adolescents[60] and the other on adolescents and young adults[61], found that the longitudinal evidence regarding the association between dietary intake and body fat yield mixed results. In the current review, the inconsistent results concerning obesity and dietary intake can be, in part, explained by the diversity of methods used to assess diet alongside with the exclusive use of self-reported methods across all studies. Inaccurate dietary assessment may comprise a serious obstacle to understand the impact of dietary intake on the development of obesity. Therefore, continued efforts towards improving self-reported methods, as well as the use of technological advances to develop new methods to improve dietary assessment are crucial in the field of nutrition research.[62,63]

Regarding PA, most of the studies relied on self-reported measures and reported inconsistent results. However, it’s important to note that the two cohort studies[26,27] that included only girls and used accelerometry to assess PA found significant results in the expected direction; and that both studies[26,38] that looked at the association between %BF and PA found a protective, albeit weak relationship. Furthermore, differential effects of PA by gender[3638] and across ethnicities[27] were found. Rauner et al[64] conducted a systematic review of cross-sectional and longitudinal studies and found that the evidence on the relationship between physical activity and overweight in adolescents was also inconsistent. On the other hand, Ramires et al[65] reviewed systematically 18 longitudinal studies on the association of PA and BF in adolescents and concluded that PA has some protective effect against BF with differences between the genders. Likewise, a systematic review by Pate et al[60] that examined several factors that predict the development of excessive fatness in children and adolescents concluded that there is enough evidence to support the negative association between objectively measured PA and excessive fatness. Given the controversial findings in the literature, it appears that the association between PA and the development of obesity in adolescence is not yet clearly understood. In the current review the mixed results might arise due to the inconsistency of methods used to assess PA, combined with an overreliance on self-reported measures. Additionally, failure to distinguish between the effect of different intensities of activity on adiposity may mitigate the possibility of detecting associations.[37]

SB can be defined as an immobile state of the body (sitting or reclining) during waking hours, resulting in energy expenditure close to the resting metabolic rate.[66,67] Research in this field has expanded exponentially since the early 2000s[67] and SB in youth has attracted attention as a potential risk factor for obesity, even though the evidence is far from conclusive.[68] In our review we found inconsistent evidence for the association between obesity and SB during adolescence. However, the studies that reported significant results[35,40,51] all pointed in the same direction, showing a small positive association between SB and BMI. On the other hand, the two studies[26,41] using objective measures of SB (accelerometry) yielded null associations. Similarly, a recent systematic review[67] including 29 reviews in children or adolescents under the age of 19 years concluded that claims for ‘clear’ associations between SB and adiposity in youth are premature or misguided, as there is limited evidence for causality and the associations range from small to very small. Once again, these findings may be due to the use of different measures of SB and the reliability on self-reported measures. All studies controlled for PA and most for some dietary factors, however future research in adolescents must take into account usual dietary intake, pubertal status and possibly sleep, as they are important potential confounders of any relationship between SB and adiposity.[67]

Indeed, sleep plays an important role in the growth, maturation and health of children and adolescents and sleep deprivation has emerged as a potential risk factor for the development of obesity.[44,69] However, we found conflicting evidence for the association between sleep and obesity during adolescence. Moreover, a gender difference appears to exist in the relationship between sleep and obesity during adolescence. Likewise, Chen et al[70] conducted a systematic review and meta-analysis and reported that findings from studies in adolescents were inconsistent and that, even though, most studies of adolescents gave significant results in the expected direction some reported no associations among girls. Contrarywise, a recent meta-analysis[71] of prospective studies concluded that short sleep duration is a risk factor for the development of obesity in infants, children and adolescents. However, these findings should be interpreted with caution as the analysis only included 3 cohort studies in adolescents. All studies included in this review were based on self-report measures of sleep duration and have not considered other sleep dimensions that could affect obesity. Only one study, addressed energy intake and PA as confounders. Therefore, more prospective studies using objective measures of sleep that consider lifestyle, behavioral and environmental confounding variables are necessary to understand the relationship between sleep and obesity in this age group.

Contextual factors exist at various levels, including the socio-environmental level that occur with families, peers and within other community environments (e.g. schools) and the physical environment (e.g. the access to and support for healthy eating and physical activity) that also operates at home, school and neighborhood environments.[16] However in the current review only studies regarding the SES, family/home environment, food and neighborhood environment, school environment, weight loss dieting, and self-control were identified and due to the limited number of prospective studies often identified for each factor, no consistent conclusions could be drawn.

The food environment constitutes one among several aspects of the obesogenic environment (obesity-promoting) and includes availability and accessibility to food as well as food advertising and marketing.[72] In the present review only two studies[19,20] regarding the food retail environment were identified. Recently, the retail food environment has been receiving increased attention due to its possible link with obesity-related behaviors[73] and it has been hypothesized that a greater access to fast food outlets and convenience stores is associated with a higher BMI, while a greater access to supermarkets might have the opposite effect due to its provision of numerous healthy products.[19,20] We found conflicting evidence regarding this hypothesis from two studies using the same sample. Additionally, a recent systematic review found no association between supermarket or grocery store availability and obesity in children, as the associations were null ≥90% of the time.[74] Further evidence using direct measures of food availability and data specifically designed to evaluate the association between food environment and obesity that considers additional confounding variables (e.g. pubertal status, dietary intake and PA) are necessary to understand the association between food environment and obesity in this age group.

School is a key social context in which children and adolescents spend approximately one-third of their waking time and in which many behaviors that can impact weight take place.[55,75] Children consume about 35 percent of their total daily energy intake from school meals[76] and spend approximately 50 percent of their daily energy expenditure at school (depending on the length of their school day).[77] However, despite its importance only four studies were identified that explored the association between the school environment and obesity during adolescence. This also occurred in Pate et al[60] review in 2013, where only one study regarding this subject was identified. Hence, further prospective studies are still required to understand the effects of the school environment on the development of obesity during adolescence.

SES may affect a population’s lifestyle, including their access to food and patterns of physical activity, which consequently will impact their energy balance.[78] In the current review four[35,51,52,56] of the five studies that look at the relationship between SES and obesity found an association, yet the direction of the association was not consistent across studies. Additionally, the strength of the association is weak in most studies. These findings are in accordance with the existing literature which shows that an association between obesity and SES exist; although the association seems to vary by gender, age, and country.[78] A comprehensive review that explored the impact of SES on obesity in children and adults was conducted in 1989. The authors observed that in developed countries, a clear negative association between SES and obesity was observed in women, while an inconsistent relationship in men, boys and girls existed. Contrarywise, in developing countries a positive relationship was observed in women, men, boys and girls.[79] A recent review focusing exclusively on the relationship between SES and childhood-adolescent weight status in rich countries reported that the relationship is mainly inverse (60.4%), while the positive relationship has nearly disappeared (1.1%). [80] Similarly, in our review only one study found a positive association between money earned and change in BMI in Wales[35]. The other study that showed a positive association points in the opposite direction since the exposure was the level of socioeconomic deprivation measured by the index of multiple deprivation (a non-linear measure of community socioeconomic status commonly used in the United Kingdom).[81] However in our review, the only study conducted in a developing country[52] reported that low-income girls were more likely to become obese when compared with their high-income counterparts. This might be explained by the fact that Brazil has been suffering a rapid social economic transition[78] and the burden of obesity has been shifting from the poor to the rich.[82] Hence, we conclude that an association between SES and obesity exists, however, the plethora of approaches and different measures of SES that were employed in the studies reviewed make it difficult to compare results across studies, and consequently understand the direction of the relationship between SES and obesity.

In what concerns biological factors, only three studies examining the relationship between genetic factors and obesity in adolescents were retrieved and all indicated that a positive association between genetics and obesity in this age group exists. Consistently, a systematic review by Pate et al[60] concluded that there is enough evidence to support the association between genetic factors and excessive fatness in children and adolescents. However, caution is necessary when considering our findings due to the limited number of studies and diversity of methods used to assess both, the exposure and outcome throughout the studies. Even though genetic factors acting in isolation are unlikely to explain the rapid increase of obesity prevalence during the last decades, it remains quite possible that genetic susceptibility coupled with environmental and behavioral factors might favor the development of obesity.[83,84] In addition, further biological factors that were not identified in our search, such as hormonal factors, endocrine disruptors and microbiological factors can also impact independently and/or synergistically the development of obesity.[85,86]

Regarding research quality, all studies included were classified as medium or high quality. Most studies performed poorly in “ascertainment of exposure” and “adequacy of follow-up” criteria. This occurred because some studies used non-validated self-reported measures to evaluate the exposure or were based on a very limited number of questions from a validated tool. Attrition bias was a frequent issue as, only approximately half of the studies had a follow up rate ≥80% or a description of those lost.

Overall, the 40 studies selected differed significantly in terms of the population, sample size and measurements used to assess both, the exposure and the outcome, which might, in part, explain the mixed results conveyed here. Additionally, several of the studies reviewed used already existing data (collected for other purposes other than studying the research question being considered) and examined the association between obesity and one factor at a time. Therefore, further prospective research that assesses multiple obesity risk factors simultaneously and uses state-of-art measures for both, the outcome and the exposure is necessary. Future research should also implement strategies to prevent loss to follow-up, as it often leads to bias and can severely compromise a study’s validity.[87,88]

Additionally, different statistical methods were used to analyze the relationship between the exposure and the outcome measure, with most studies using multivariate linear regression. However, as the exposure is often measured using different methods and metrics across the studies, the use of regression coefficients is not meaningful. Instead standardized regression coefficients should be used as an effect size estimate.[89] However, most studies fail to mention which measure was used, making it difficult to compare effects size across studies.

One strength of our review lies on the inclusion of studies that only used objectively measured weight and height, as adolescents’ self-reported anthropometric measures might lead to inaccurate estimation of overweight rates and there is bias by sex and weight status.[90,91] Nevertheless, self-reported data is valuable when objective measures are not available. The exclusive inclusion of prospective cohort studies represents a further strength, as this study design is considered the “gold standard” of observational research.[92] It is important to acknowledge that prospective cohort studies demonstrate only association[60] and in order to establish causal relationships experimental research is required.[93] However, not all obesity risk factors can be submitted to experimental manipulation, as researchers cannot randomize genes nor contextual factors such as income, education, food prices or social norms.[94] In addition, randomized controlled trials (the strongest study design for identifying causal relationships) are limited in the number of exposures they can evaluate at once[95] and their usefulness for studying the multifactorial etiology of obesity might be limited.

This review is not without limitations. First, it is important to note that this review comprises 40 studies, each one with their own limitations. Secondly, the exclusive use of English-language studies published during or after the year 2000 might have led to the exclusion of relevant evidence; Thirdly, the review might be prone to publication bias, as only the peer-reviewed literature was searched. Lastly, the use of BMI as the outcome measure constitutes an additional limitation, as BMI does not measure body fat directly nor does it distinguish fat mass from lean mass.[96,97] However, BMI is the most widely used and accepted measure of obesity and has shown strong correlation with gold standard measures of body fat.[96]

The authors would also like to acknowledge that even though no associations were found between breastfeeding or fruit/vegetables and obesity in adolescents, both exposures have been associated with numerous health benefits[98,99] Thus, promoting and supporting exclusive breastfeeding in the first 6 months of life[100] and the consumption of more than 400 grams of fruit and vegetables per day[101] should remain public health priorities.

Conclusions

In the current review we found a positive consistent association between genetic factors and obesity during adolescence, however these findings should be interpreted with caution due to the heterogeneity and small number of studies identified. Furthermore, there is evidence to support the association between SES and obesity. For the remaining factors the evidence was either conflicting, limited or non-existent. Hence, the association between most risk factors and obesity during adolescence remains unclear. We also highlighted the heterogeneity of the studies included in this review, particularly regarding the methods used to assess the exposure and the outcome, which makes comparisons across studies difficult. Therefore, these findings reinforce the critical need for further prospective studies that examine the effect of multiple factors concurrently on obesity using state-of-art measures. This is imperative, as a profounder understanding of the multifactorial contributors to obesity and possible interactions between them will aid in the development of effective strategies and interventions to manage and prevent obesity during adolescence.

Supporting information

S2 Appendix. Quality assessment of the prospective cohort studies included in the present systematic review.

https://doi.org/10.1371/journal.pone.0214941.s002

(DOCX)

S3 Appendix. Characteristics of the 40 studies included in the systematic review.

https://doi.org/10.1371/journal.pone.0214941.s003

(DOCX)

References

  1. 1. World Health Organization. Childhood overweight and obesity [Internet]. World Health Organization; 2017 [cited 2018 Mar 14]. Available from: http://www.who.int/dietphysicalactivity/childhood/en/
  2. 2. Abarca-Gómez L, Abdeen ZA, Hamid ZA, Abu-Rmeileh NM, Acosta-Cazares B, Acuin C, et al. 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 Dec;390:2627–42. pmid:29029897
  3. 3. Ng M, Fleming T, Robinson M, Thomson B, Graetz N, Margono C, et al. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet. 2014 Aug;384(9945):766–81. pmid:24880830
  4. 4. Kelsey MM, Zaepfel A, Bjornstad P, Nadeau KJ. Age-related consequences of childhood obesity. Gerontology. 2014;60(3):222–8. pmid:24434909
  5. 5. Güngör N. Overweight and Obesity in Children and Adolescents. J Clin Res Pediatr Endocrinol. 2014 Sep 5;6(3):129–43. pmid:25241606
  6. 6. Inchley J, Currie D, Jewell J, Breda J, Barnekow V. Adolescent obesity and related behaviours: trends and inequalities in the WHO region 2002–2014. 2017.
  7. 7. Craigie AM, Matthews JNS, Rugg-Gunn AJ, Lake AA, Mathers JC, Adamson AJ. Raised adolescent body mass index predicts the development of adiposity and a central distribution of body fat in adulthood: A longitudinal study. Obes Facts. 2009;2(3):150–6. pmid:20054219
  8. 8. Dietz WH. Critical periods in childhood for the development of obesity. Am J Clin Nutr. 1994 May 1;59(5):955–9. pmid:8172099
  9. 9. Alberga AS, Sigal RJ, Goldfield G, Prud’ homme D, Kenny GP. Overweight and obese teenagers: why is adolescence a critical period? Pediatr Obes. 2012 Aug;7(4):261–73. pmid:22461384
  10. 10. Butte NF, Christiansen E, Sørensen TIA. Energy imbalance underlying the development of childhood obesity. Obesity. 2007;15(12):3056–66. pmid:18198315
  11. 11. Campbell MK. Biological, environmental and social influences on childhood obesity. Pediatr Res. 2016 Jan 20;79(1–2):205–11. pmid:26484623
  12. 12. Seburg EM, Olson-Bullis BA, Bredeson DM, Hayes MG, Sherwood NE. A Review of Primary Care-Based Childhood Obesity Prevention and Treatment Interventions. Curr Obes Rep. 2015 Jun 14;4(2):157–73. pmid:26213643
  13. 13. Agurs-Collins T, Bouchard C. Gene-Nutrition and Gene-Physical Activity Interactions in the Etiology of Obesity. Obesity. 2008 Dec;16(Suppl 3):S2–4.
  14. 14. Ang YN, Wee BS, Poh BK, Ismail MN. Multifactorial Influences of Childhood Obesity. Curr Obes Rep. 2013 Mar 30;2(1):10–22.
  15. 15. Moreno LA, Pigeot I, Ahrens W. Epidemiology of Obesity in Children and Adolescents. 1st ed. Moreno LA, Pigeot I, Ahrens W, editors. New York, NY: Springer New York; 2011.
  16. 16. Lytle LA. Examining the Etiology of Childhood Obesity: The IDEA Study. Am J Community Psychol. 2009 Dec;44(3–4):338–49. pmid:19838791
  17. 17. Faith MS, Kral TVE. Social environment and Genetic Influences on Obesity and Obesity-Promoting Behaviors: Fostering Research Integration. In: Hernandez L, Blazer D, editors. Genes, Behavior, and the Social Environment: Moving Beyond the Nature/Nurture Debate. Washington, D.C.: National Academies Press; 2006. p. 237–8.
  18. 18. Wells GA, Shea B, O’Connell D, Peterson J, Welch V, Losos M TP. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses [Internet]. 2018 [cited 2018 Jul 1]. Available from: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp
  19. 19. Chen H-J, Wang Y. Changes in the Neighborhood Food Store Environment and Children’s Body Mass Index at Peripuberty in the United States. J Adolesc Heal. 2016;58(1):111–8.
  20. 20. Shier V, An R, Sturm R. Is there a robust relationship between neighbourhood food environment and childhood obesity in the USA? Public Health. 2012;126(9):723–30. pmid:22898435
  21. 21. Campbell K, Andrianopoulos N, Hesketh K, Ball K, Crawford D, Brennan L, et al. Parental use of restrictive feeding practices and child BMI z-score. A 3-year prospective cohort study. Appetite. 2010;55(1):84–8. pmid:20420869
  22. 22. Duckworth AL, Tsukayama E, Geier AB. Self-controlled children stay leaner in the transition to adolescence. Appetite. 2010;54(2):304–8. pmid:20004223
  23. 23. Enes CC, Slater B. Variation in dietary intake and physical activity pattern as predictors of change in body mass index (BMI) z-score among brazilian adolescents. Rev Bras Epidemiol. 2013;16(2):493–501. pmid:24142019
  24. 24. Roberts RE, Duong HT. Is there an association between adolescent sleep restriction and obesity. J Psychosom Res. 2015;79(6):651–6. pmid:26055094
  25. 25. Victora CG, Barros F, Lima RC, Horta BL, Wells J. Anthropometry and body composition of 18 year old men according to duration of breast feeding: birth cohort study from Brazil. BMJ. 2003;327(7420):901. pmid:14563746
  26. 26. Cohen DA, Ghosh-Dastidar B, Conway TL, Evenson KR, Rodriguez DA, Beckman R, et al. Energy balance in adolescent girls: The trial of activity for adolescent girls cohort. Obesity. 2014;22(3):772–80. pmid:23804512
  27. 27. White J, Jago R. Prospective associations between physical activity and obesity among adolescent girls: Racial differences and implications for prevention. Arch Pediatr Adolesc Med. 2012;166(6):522–7. pmid:22665029
  28. 28. Gopinath B, Flood VM, Rochtchina E, Baur LA, Louie JCY, Smith W, et al. Carbohydrate nutrition and development of adiposity during adolescence. Obesity. 2013;21(9):1884–90. pmid:23519919
  29. 29. Murakami K, Mccaffrey TA, Gallagher AM, Neville CE, Boreham CA, Livingstone MBE, et al. Dietary glycemic index and glycemic load in relation to changes in body composition measures during adolescence: Northern Ireland Young Hearts Study. Int J Obes. 2014;38(2):252–8.
  30. 30. Lin SL, Tarrant M, Hui LL, Kwok MK, Lam TH, Leung GM, et al. The Role of Dairy Products and Milk in Adolescent Obesity: Evidence from Hong Kong’s “Children of 1997” Birth Cohort. PLoS One. 2012;7(12).
  31. 31. Bigornia SJ, Lavalley MP, Noel SE, Moore LL, Ness AR, Newby P. Sugar-sweetened beverage consumption and central and total adiposity in older children: A prospective study accounting for dietary reporting errors. Public Health Nutr. 2015;18(7):1155–63. pmid:25166959
  32. 32. Feeley AB, Musenge E, Pettifor JM, Norris SA. Investigation into longitudinal dietary behaviours and household socio-economic indicators and their association with BMI Z-score and fat mass in South African adolescents: the Birth to Twenty (Bt20) cohort. Public Health Nutr. 2013;16(4):693–703. pmid:22801035
  33. 33. Laska MN, Murray DM, Lytle LA, Harnack LJ. Longitudinal associations between key dietary behaviors and weight gain over time: Transitions through the adolescent years. Obesity. 2012;20(1):118–25. pmid:21701567
  34. 34. Gopinath B, Flood VM, Burlutsky G, Louie JCY, Baur LA, Mitchell P. Frequency of takeaway food consumption and its association with major food group consumption, anthropometric measures and blood pressure during adolescence. Br J Nutr. 2016;115(11):2025–30. pmid:27046032
  35. 35. Elgar FJ, Roberts C, Moore L, Tudor-Smith C. Sedentary behaviour, physical activity and weight problems in adolescents in Wales. Public Health. 2005;119(6):518–24. pmid:15826893
  36. 36. De Souza MC, Eisenmann JC, E Santos D V, De Chaves RN, De Moraes Forjaz CL, Maia JAR. Modeling the dynamics of BMI changes during adolescence. the oporto growth, health and performance study. Int J Obes. 2015;39(7):1063–9.
  37. 37. Bélanger M, O’Loughlin J, Karp I, Barnett TA, Sabiston CM, Belanger M, et al. Physical activity fluctuations and body fat during adolescence. Pediatr Obes. 2012 Feb;7(1):73–81. pmid:22434741
  38. 38. Barnett TA, Maximova K, Sabiston CM, Van Hulst A, Brunet J, Castonguay AL, et al. Physical activity growth curves relate to adiposity in adolescents. Ann Epidemiol. 2013;23(9):529–33. pmid:23958405
  39. 39. Aires L, Mendona D, Silva G, Gaya AR, Santos MP, Ribeiro JC, et al. A 3-year longitudinal analysis of changes in body mass index. Int J Sports Med. 2010;31(2):133–7. pmid:20027539
  40. 40. Dumith SC, Garcia LMT, Da Silva KS, Menezes AMB, Hallal PC. Predictors and health consequences of screen-time change during adolescence—1993 Pelotas (Brazil) birth cohort study. J Adolesc Heal. 2012;51(SUPPL. 6):S16–21.
  41. 41. Fletcher EA, Lamb KE, McNaughton SA, Garnett SP, Dunstan DW, Baur LA, et al. Cross-sectional and prospective mediating effects of dietary intake on the relationship between sedentary behaviour and body mass index in adolescents. BMC Public Health. 2017;17(1).
  42. 42. Schafer AA, Domingues MR, Dahly DL, Meller FO, Goncalves H, Wehrmeister FC, et al. Sleep Duration Trajectories and Body Composition in Adolescents: Prospective Birth Cohort Study. PLoS One. 2016;11(3):e0152348. pmid:27010312
  43. 43. Araujo J, Severo M, Ramos E. Sleep Duration and Adiposity During Adolescence. Pediatrics. 2012 Nov 1;130(5):e1146–54. pmid:23027175
  44. 44. Lytle LA, Murray DM, Laska MN, Pasch KE, Anderson SE, Farbakhsh K. Examining the Longitudinal Relationship Between Change in Sleep and Obesity Risk in Adolescents. Heal Educ Behav. 2013;40(3):362–70.
  45. 45. Crawford D, Cleland V, Timperio A, Salmon J, Andrianopoulos N, Roberts R, et al. The longitudinal influence of home and neighbourhood environments on children’s body mass index and physical activity over 5 years: The CLAN study. Int J Obes. 2010;34(7):1177–87.
  46. 46. O’Hara JK, Haynes-Maslow L. Examining the Association Between School Vending Machines and Children’s Body Mass Index by Socioeconomic Status. J Nutr Educ Behav. 2015;47(6):526–531.e1. pmid:26371992
  47. 47. Cornes BK, Zhu G, Martin NG. Sex differences in genetic variation in weight: A longitudinal study of body mass index in adolescent twins. Behav Genet. 2007;37(5):648–60. pmid:17896175
  48. 48. Liem ET, Vonk JM, Sauer PJJ, Van Der Steege G, Oosterom E, Stolk RP, et al. Influence of common variants near INSIG2, in FTO, and near MC4R genes on overweight and the metabolic profile in adolescence: The TRAILS (TRacking Adolescents’ Individual Lives Survey) Study. Am J Clin Nutr. 2010;91(2):321–8. pmid:20007308
  49. 49. Wang S, Schwartz MB, Shebl FM, Read M, Henderson KE, Ickovics JR. School breakfast and body mass index: a longitudinal observational study of middle school students. Pediatr Obes. 2017;12(3):213–20. pmid:26989876
  50. 50. Shields L, O’Callaghan M, Williams GM, Najman JM, Bor W. Breastfeeding and obesity at 14 years: A cohort study. J Paediatr Child Health. 2006;42(5):289–96. pmid:16712560
  51. 51. Schuster MA, Elliott MN, Bogart LM, Klein DJ, Feng JY, Wallander JL, et al. Changes in obesity between fifth and tenth grades: A longitudinal study in three metropolitan areas. Pediatrics. 2014;134(6):1051–8. pmid:25384494
  52. 52. Assunção MCF, Muniz LC, Dumith SC, Clark VL, Araujo CLP, Goncalves H, et al. Predictors of body mass index change from 11 to 15 years of age: The 1993 Pelotas (Brazil) birth cohort study. J Adolesc Heal. 2012 Dec;51(SUPPL. 6):S65–9.
  53. 53. Bigornia SJ, LaValley MP, Moore LL, Northstone K, Emmett P, Ness R, et al. Dairy intakes at age 10 years do not adversely affect risk of excess adiposity at 13 years. J Nutr. 2014;144(7):1081–90. pmid:24744312
  54. 54. Ludwig DS, Peterson KE, Gortmaker SL. Relation between consumption of sugar-sweetened drinks and childhood obesity: A prospective, observational analysis. Lancet. 2001;357(9255):505–8. pmid:11229668
  55. 55. Cunningham SA, Zavodny M. Does the sale of sweetened beverages at school affect children’s weight? Soc Sci Med. 2011;73(9):1332–9. pmid:21907477
  56. 56. Fraser LK, Clarke GP, Cade JE, Edwards KL. Fast food and obesity: A spatial analysis in a large United Kingdom population of children aged 13–15. Am J Prev Med. 2012;42(5):e77–85. pmid:22516506
  57. 57. MacFarlane A, Cleland V, Crawford D, Campbell K, Timperio A. Longitudinal examination of the family food environment and weight status among children. Int J Pediatr Obes. 2009;4(4):343–52. pmid:19922051
  58. 58. Wardle J, Brodersen NH, Boniface D. School-based physical activity and changes in adiposity. Int J Obes. 2007;31(9):1464–8.
  59. 59. Hill JO, Wyatt HR, Peters JC. Energy Balance and Obesity. Circulation. 2012;126(1):126–32. pmid:22753534
  60. 60. Pate RR, O’Neill JR, Liese AD, Janz KF, Granberg EM, Colabianchi N, et al. Factors associated with development of excessive fatness in children and adolescents: a review of prospective studies. Obes Rev. 2013 Aug;14(8):645–58. pmid:23601571
  61. 61. Schneider BC, Dumith SC, Orlandi SP, Assunção MCF. Diet and body fat in adolescence and early adulthood: a systematic review of longitudinal studies. Cien Saude Colet. 2017 May;22(5):1539–52. pmid:28538925
  62. 62. Labonté M-È, Kirkpatrick SI, Bell RC, Boucher BA, Csizmadi I, Koushik A, et al. Dietary assessment is a critical element of health research–Perspective from the Partnership for Advancing Nutritional and Dietary Assessment in Canada. Appl Physiol Nutr Metab. 2016 Oct;41(10):1096–9. pmid:27608060
  63. 63. Thompson FE, Subar AF, Loria CM, Reedy JL, Baranowski T. Need for Technological Innovation in Dietary Assessment. J Am Diet Assoc. 2010 Jan;110(1):48–51. pmid:20102826
  64. 64. Rauner A, Mess F, Woll A. The relationship between physical activity, physical fitness and overweight in adolescents: a systematic review of studies published in or after 2000. BMC Pediatr. 2013 Dec 1;13(1):19.
  65. 65. Ramires VV, Dumith SC, Gonçalves H. Longitudinal Association between Physical Activity and Body Fat during Adolescence: A Systematic Review. J Phys Act Heal. 2015 Sep;12(9):1344–58.
  66. 66. Heinonen I, Helajärvi H, Pahkala K, Heinonen OJ, Hirvensalo M, Pälve K, et al. Sedentary behaviours and obesity in adults: the Cardiovascular Risk in Young Finns Study. BMJ Open. 2013 May 1;3(6):e002901. pmid:23794543
  67. 67. Biddle SJH, García Bengoechea E, Wiesner G. Sedentary behaviour and adiposity in youth: a systematic review of reviews and analysis of causality. Int J Behav Nutr Phys Act. 2017 Dec 28;14(1):43. pmid:28351363
  68. 68. Coombs NA, Stamatakis E. Associations between objectively assessed and questionnaire-based sedentary behaviour with BMI-defined obesity among general population children and adolescents living in England. BMJ Open. 2015 Jun 18;5(6):e007172–e007172. pmid:26088807
  69. 69. Eisenmann JC. Insight into the causes of the recent secular trend in pediatric obesity: Common sense does not always prevail for complex, multi-factorial phenotypes. Prev Med (Baltim). 2006 May;42(5):329–35.
  70. 70. Chen X, Beydoun MA, Wang Y. Is Sleep Duration Associated With Childhood Obesity? A Systematic Review and Meta-analysis. Obesity. 2008 Feb;16(2):265–74. pmid:18239632
  71. 71. Miller MA, Kruisbrink M, Wallace J, Ji C, Cappuccio FP. Sleep duration and incidence of obesity in infants, children, and adolescents: a systematic review and meta-analysis of prospective studies. Sleep. 2018 Apr 1;41(4).
  72. 72. Lake A, Townshend T. Obesogenic environments: exploring the built and food environments. J R Soc Promot Health. 2006 Nov 30;126(6):262–7. pmid:17152319
  73. 73. Wilkins EL, Morris MA, Radley D, Griffiths C. Using Geographic Information Systems to measure retail food environments: Discussion of methodological considerations and a proposed reporting checklist (Geo-FERN). Health Place. 2017 Mar;44(January):110–7.
  74. 74. Cobb LK, Appel LJ, Franco M, Jones-Smith JC, Nur A, Anderson CAM. The relationship of the local food environment with obesity: A systematic review of methods, study quality, and results. Obesity. 2015 Jul;23(7):1331–44. pmid:26096983
  75. 75. Ip P, Ho FK-W, Louie LH-T, Chung TW-H, Cheung Y-F, Lee S-L, et al. Childhood Obesity and Physical Activity-Friendly School Environments. J Pediatr. 2017 Dec;191:110–6. pmid:28987751
  76. 76. Briefel RR, Wilson A, Gleason PM. Consumption of Low-Nutrient, Energy-Dense Foods and Beverages at School, Home, and Other Locations among School Lunch Participants and Nonparticipants. J Am Diet Assoc. 2009 Feb;109(2):S79–90.
  77. 77. Institute of Medicine. Preventing Childhood Obesity: Health in the Balance [Internet]. Koplan JP, Liverman CT, Kraak VA, editors. Washington, D.C.: The National Academies Press; 2005. Available from: https://www.ncbi.nlm.nih.gov/books/NBK83833/
  78. 78. Wang Y, Lim H. The global childhood obesity epidemic and the association between socio-economic status and childhood obesity. Int Rev Psychiatry. 2012 Jun 24;24(3):176–88. pmid:22724639
  79. 79. Sobal J, Stunkard AJ. Socioeconomic status and obesity: A review of the literature. Psychol Bull. 1989;105(2):260–75. pmid:2648443
  80. 80. Barriuso L, Miqueleiz E, Albaladejo R, Villanueva R, Santos JM, Regidor E. Socioeconomic position and childhood-adolescent weight status in rich countries: a systematic review, 1990–2013. BMC Pediatr. 2015 Dec 21;15(1):129.
  81. 81. Pujades-Rodriguez M, Timmis A, Stogiannis D, Rapsomaniki E, Denaxas S, Shah A, et al. Socioeconomic Deprivation and the Incidence of 12 Cardiovascular Diseases in 1.9 Million Women and Men: Implications for Risk Prediction and Prevention. Woodward M, editor. PLoS One. 2014 Aug 21;9(8):e104671. pmid:25144739
  82. 82. Monteiro CA, Conde WL, Popkin BM. Income-Specific Trends in Obesity in Brazil: 1975–2003. Am J Public. 2007;97(10):1808–12.
  83. 83. Bhadoria A, Sahoo K, Sahoo B, Choudhury A, Sufi N, Kumar R. Childhood obesity: Causes and consequences. J Fam Med Prim Care. 2015;4(2):187.
  84. 84. Schwartz MW, Seeley RJ, Zeltser LM, Drewnowski A, Ravussin E, Redman LM, et al. Obesity Pathogenesis: An Endocrine Society Scientific Statement. Endocr Rev. 2017 Aug 1;38(4):267–96. pmid:28898979
  85. 85. Cardel M. Contributors to Pediatric Obesity in Adolescence: More than just Energy Imbalance. Open Obes Journal,. 2011 Mar 1;3(1):17–26.
  86. 86. Varnaccia G, Zeiher J. Factors influencing childhood obesity–the establishment of a population-wide monitoring system in Germany. J Heal Monit. 2017;2(2):85–97.
  87. 87. Dettori J. Loss to follow-up. Evid Based Spine Care J. 2011 Feb 10;2(1):7–10. pmid:22956930
  88. 88. Kristman V, Manno M, Cote P. Loss to follow-up in cohort studies: how much is too much? Eur J Epidemiol. 2004;19(8):751–60. pmid:15469032
  89. 89. Nieminen P, Lehtiniemi H, Vähäkangas K, Huusko A, Rautio A. Standardised regression coefficient as an effect size index in summarising findings in epidemiological studies. Epidemiol Biostat Public Heal. 2013;10(4):1–15.
  90. 90. Sherry B. Accuracy of Adolescent Self-report of Height and Weight in Assessing Overweight Status. Arch Pediatr Adolesc Med. 2007 Dec 1;161(12):1154. pmid:18056560
  91. 91. Galfo M, Censi L, D’Addezio L, Martone D, Roccaldo R. Validity of self-reported weight, height and BMI in Italian adolescents for assessing prevalence of overweight/obesity. Clin Nutr Metab. 2018;1(1):1–7.
  92. 92. Thiese MS. Observational and interventional study design types; an overview. Biochem Medica. 2014;24(2):199–210.
  93. 93. Kirk RE. Experimental Design. In: Ritzer G, editor. The Blackwell Encyclopedia of Sociology. John Wiley & Sons, Ltd; 2015.
  94. 94. Auld MC, Grootendorst P. Challenges for Causal Inference in Obesity Research. Cawley J, editor. The Oxford Handbook of the Social Science of Obesity. Oxford University Press; 2011.
  95. 95. Nicholson J, Sanson A, Rempel L, Smart D, Patton G. Longitudinal studies of children and youth Implications for future studies [Internet]. 2004 [cited 2018 Sep 4]. Available from: https://aifs.gov.au/publications/childrens-health-and-development/3-longitudinal-studies-children-and-youth-implications
  96. 96. Adab P, Pallan M, Whincup PH. Is BMI the best measure of obesity? BMJ. 2018 Mar 29;360(March):k1274.
  97. 97. Mahadevan S, Ali I. Is body mass index a good indicator of obesity? Int J Diabetes Dev Ctries. 2016 Jun 7;36(2):140–2.
  98. 98. Hoddinott P, Tappin D, Wright C. Breast feeding. BMJ. 2008 Apr 19;336(7649):881–7. pmid:18420694
  99. 99. Slavin J, Lloyd B. Health Benefits of Fruits and Vegetables. Adv Nutr. 2012;3(4):506–16. pmid:22797986
  100. 100. World Health Organization. Exclusive breastfeeding for six months best for babies everywhere [Internet]. World Health Organization; 2011 [cited 2018 Sep 10]. Available from: http://www.who.int/mediacentre/news/statements/2011/breastfeeding_20110115/en/
  101. 101. World Health Organization. Increasing fruit and vegetable consumption to reduce the risk of noncommunicable diseases [Internet]. World Health Organization; 2018 [cited 2018 Sep 10]. Available from: http://www.who.int/elena/titles/fruit_vegetables_ncds/en/