Prerpints.org logo
Preprint
Review

Associations between Meal Patterns and Risk of Overweight/Obesity and Metabolically Unhealthy Obesity in Children and Adolescents: A Systematic Review of Longitudinal Studies and Randomised Controlled Trials

Altmetrics

Downloads

160

Views

70

Comments

0

A peer-reviewed article of this preprint also exists.

supplementary.zip (43.56KB )

Submitted:

31 July 2024

Posted:

31 July 2024

You are already at the latest version

Alerts
Abstract
Childhood overweight/obesity (OV/OB) is a major public health problem, of greater concern when accompanied with comorbidities such as hypertension and insulin resistance leading to metabolically unhealthy obesity (MUO). Aetiologic associations between meal patterns, OV/OB risk and MUO are limited. The aim of this systematic review was to explore associations between meal patterns and the risk of childhood OV/OB and MUO. Longitudinal studies and randomised controlled trials from PUBMED and Scopus published between January 2013 - April 2024 were retrieved. Twenty-eight studies were included, all of which reported on OV/OB risk; none on MUO risk. Regular consumption of breakfast and family meals and avoiding dining while watching TV may be protective factors against childhood OV/OB, whereas meal skipping (primarily breakfast) may be a detrimental factor. Mixed effects of meal frequency on OV/OB risk were observed; no effects of frequency of lunch or of fast-food consumption and of meals served at school were found. There was insufficient evidence to support the role of other patterns (meal timing, eating in other social contexts). Meals were mainly participant-identified, leading to increased heterogeneity. Research focusing on childhood MUO and improved methodological approach (e.g. harmonised definitions) regarding the assessment of meal patterns are highly warranted.
Keywords: 
Subject: Public Health and Healthcare  -   Other

1. Introduction

Childhood obesity remains one of the most significant global public health challenges of the 21st century. According to the Non-Communicable Diseases Risk Factor Collaboration, from 1990 to 2018 the age-standardised prevalence of childhood obesity increased in the vast majority of countries worldwide (93% of countries in girls and 98% in boys) [1]. The prevalence of overweight (including obesity) and obesity in Europe remains at alarmingly high rates (29% and 13% in boys aged 6-9 years; 27% and 9% in girls aged 6-9 years, respectively), with significant variations among countries [2] while in the U.S.A., a 20-year analysis highlighted a significant increase of about 30% in the prevalence of obesity and almost a 2-fold increase in severe obesity (defined as Body Mass Index (BMI) ≥120th percentile of U.S. Centers for Disease Control and Prevention growth charts in the U.S.A.) in children aged 2-19 years [3].
Research so far has focused on the causes and risk factors of obesity, such as early life exposures / perinatal factors, growth trajectories, socioeconomic factors, physical environment and lifestyle habits (diet, physical activity, sedentary activities, sleep) [4,5]. The role of diet in childhood overweight/obesity (OV/OB) has been variously studied in terms of nutrients and food groups intake or adherence to dietary patterns [6,7,8]. There is, however, an interest towards other approaches to examine relationships between diet and health or disease that also capture other dimensions of eating habits, such as consumption of meals and snacks. For example, the family meal has been identified as an entry point for intervention, as it constitutes a social setting with the potential to shape children’s eating behaviours from a young age [9]. A dietary approach in the context of meals, e.g. meal preparation, could complement dietary consultations and may be a more practical and easy way to convey clear messages to children and families.
A “meal pattern” has been used in the literature as a term to describe an individual’s dietary habits either at the level of meals or eating occasions [10,11]. The term “eating occasion” or “eating event” or “eating episode” describes any consumption of food based on specific characteristics, such as timing of the day, energy content, combination of foods, usually discerning eating occasions as main meals (i.e. breakfast, lunch, dinner) and snacks [10]. Meals have also been described in relation to their format (i.e. combination of foods or content of nutrients) and the context/environment the meals are consumed (i.e. with the presence of other people or while doing an activity) [11]. There is, however, heterogeneity in the assessment of meal patterns in the literature, as different definitions of a meal have been applied across studies. A meal can be participant-defined through self-reporting [12] or it is defined according to the time it is consumed within the day [13], according to the energy content [14] or a combination of both [10], which may affect the interpretation of the findings [10].
In relation to childhood OV/OB, published systematic reviews have highlighted the role of different aspects of meal patterns on the risk of OV/OB, but evidence derives primarily from cross-sectional studies. For example, a meta-analysis exploring associations between frequency of family meals and children’s health showed that having frequent family meals was associated with a lower BMI and better overall diet quality; nevertheless 67 out of the 75 included studies had a cross-sectional design [9]. Similarly, an earlier meta-analysis of 10 cross-sectional and one case-control studies showed an inverse association of eating frequency (defined as the total number of eating episodes per day) and childhood OV/OB status in boys, but not girls [15]. The systematic review conducted by Monzani et al. [16] included 37 articles, 32 of which were cross-sectional, and it reported that skipping breakfast was associated with an increased risk or prevalence of OV/OB [16]. A meta-analysis that examined the association of meal timing and adiposity showed weak associations between higher energy intake close to bedtime and evening meal skipping with adiposity, but 17 out of 20 included studies had a cross-sectional design [17]. To date, there is no systematic review to capture different dimensions of meal patterns and their associations with the risk of developing childhood OV/OB based on findings from longitudinal or randomized intervention studies that allow exploration of causality.
An emerging issue in relation to childhood OV/OB is the development of associated comorbidities, such as diabetes, hypertension, lipid abnormalities and liver dysfunction, which are often used to define metabolically unhealthy obesity (MUO) [5,18]. Children and adolescents with obesity have an increased prevalence ratio of 1.4 to develop prediabetes, 21.2 for cardiovascular disease and 26.1 for metabolic-associated steatotic liver disease, compared to children with normal weight [18]. Children with obesity are also more likely to live with obesity in adult life which is associated with comorbidities across the life course [5]. The role of diet, let alone of meal patterns, in relation to MUO in children and adolescents has been scarcely explored and very few interventions have evaluated markers of MUO [19,20,21].
Meal pattern consumption remains an interesting dimension of eating habits that could contribute in the engagement of optimal behaviours and lifestyle modification within the context of overweight and associated comorbidities prevention and management. The aim, therefore, of this review was to systematically gather all available evidence from longitudinal cohorts or randomised interventions exploring effects of meal patterns on the risk of developing OV/OB and MUO in children and adolescents.

2. Materials and Methods

The systematic review was conducted according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines [22,23] (Supplementary Materials S1: PRISMA 2020 Checklist). The study was registered in the PROSPERO international prospective register of systematic reviews, of the National Institute for Health and Care Research (Title: The association of meal patterns and risk of obesity and metabolically unhealthy obesity in children and adolescents; registration number CRD42023477708).

2.1. Search Strategy

A systematic search was conducted in May 2024 in two databases (MEDLINE/Pubmed, Scopus) to identify studies which evaluated the role of meal patterns in childhood and adolescent OV/OB risk and MUO risk, published within the last decade (in specific January 2013 – April 2024). Research questions and search keywords were guided from the “Population, Exposure, Comparator, Outcome” (PECO) model for epidemiological studies and the “Population, Intervention, Comparator, Outcome” (PICO) model for interventional studies (Table 1) [24]. Selection of exposures parameters was guided by Leech et al. [11] and included frequency of / omitting meals, consumption of meal within different contexts (e.g. while watching TV) and environments (e.g. at home) and meal quality. Outcome measures related to OV/OB included anthropometric indices, BMI/BMI z-score and body composition parameters. Indicators for MUO included blood pressure, blood lipids, glucose metabolism and metabolic comorbidities. Keywords were formulated according to the PECO/PICO model are available in Supplementary Materials S2.

2.2. Eligibility Criteria

Inclusion criteria reflected the research questions, i.e. the population, the exposures/interventions and the outcomes of interest, as well as the study design. Original, peer-reviewed articles, conducted in children and adolescents (age range 2-19 years old as defined by the World Health Organisation [25] for baseline and follow-up), published from 2013 to 2024, which evaluated outcomes related to OV/OB or MUO were included. Only longitudinal studies and randomised controlled clinical trials (RCTs) were included to ensure a better quality of methodological design that could also allow aetiological assumptions. A minimum follow-up of 12 months was also applied to both study designs, which was deemed adequate to observe meaningful effects of meal patterns on OV/OB and MUO risk. Search was also refined to include only studies conducted in countries with a westernised way of living and in populations of a similar genetic background (i.e. Caucasian origin; Europe, U.S.A., Canada, Oceania).
Studies conducted in animals and people <2 or >19 years old; studies conducted in Asia, Africa and South America; reviews, letters, editorials, review protocols and pre-prints; cross-sectional studies, in-vitro, in-vivo animals or in silico studies and non-randomised, uncontrolled clinical trials were excluded. Also, studies with multidisciplinary lifestyle observations or interventions with no clear analysis of the association between meal patterns and risk of OV/OB or MUO were also excluded. Due to the nature of the research question, studies evaluating meal patterns and MUO indicators in children without OV/OB at baseline and/or follow-up were also excluded.

2.3. Selection of Studies and Data Extraction

All studies identified from databases were imported in the Zotero software (https://www.zotero.org). Following the removal of duplicates, studies were screened for eligibility in two stages. Initially, a title and abstract screening was performed independently by two researchers (GS and AK). After exclusion based on titles and abstracts, all remaining articles were considered for full-text review by both researchers, who applied the eligibility criteria for the final selection. Disagreements at any stage were resolved by a third researcher (EB).
Data extraction was performed by one researcher (GS), with a second researcher (AK) randomly checking a sample of the eligible reports. Any disagreement was resolved by a third researcher (EB). Extracted information included:
-
Study information: first author’s name, year of publication, acronym, country, setting, duration)
-
Population: sample size, baseline age and sex distribution
-
Exposure(s) where relevant: type, definition, assessment method
-
Intervention (where relevant): groups, randomization, components, mode of delivery, duration
-
Outcome(s): type, definition, assessment method
-
Statistical analysis: analyzed sample, statistical model, covariates
-
Study results: main findings
Throughout the process, in case of missing information or uncertainties, relevant information was sought in Supplementary Materials or directly from study investigators. Extracted information is presented according to study design (prospective epidemiological studies or RCTs).

2.4. Risk of Bias

One researcher (GS) assessed the quality of all included studies and a second researcher (AK) assessed a random 20% of the sample. The quality assessment of prospective epidemiological studies was conducted with the Risk Of Bias In Non-randomized Studies - of Exposures (ROBINS-E) tool (https://www.riskofbias.info/welcome/robins-e-tool) [26]. The ROBINS-E tool comprises of seven domains: bias due to confounding; bias arising from measurement of the exposure; bias in selection of participants into the study (or into the analysis); bias due to post-exposure interventions; bias due to missing data; bias arising from measurement of the outcome; bias in selection of the reported result. Each domain and the overall study are assessed as “low risk”, “some concerns”, “high risk” or “very high risk”.
The quality assessment of randomised controlled trials was conducted with the revised Cochrane risk-of-bias tool for randomized trials (RoB 2) tool (https://www.riskofbias.info/welcome/rob-2-0-tool/current-version-of-rob-2) [27]. The RoB2 tool comprises of five domains: bias arising from the randomization process; bias due to deviations from the intended interventions (effect of assignment to intervention or adhering to intervention); bias due to missing outcome data; bias in measurement of the outcome; bias in selection of the reported result. Each domain and the overall study are assessed as “low risk”, “some concerns” or “high risk”.

3. Results

The initial search yielded 3304 results (2493 from Scopus and 811 from MEDLINE/Pubmed). After removal of duplicates (722), 2505 articles were excluded following review of title and abstract and further 49 were excluded after full-text examination. Twenty-eight reports from 25 studies were included in this review (Figure 1). All reports examined associations between meal patterns and childhood OV/OB risk. No studies were found in relation to MUO risk.

3.1. Study Characteristics

Of the 28 included reports (25 studies), 13 studies were conducted in the U.S.A. [20,28,29,30,31,32,33,34,35,36,37,38,39], 12 in Europe (U.K. [40,41,42], Netherlands [43,44], Germany [45,46], Republic of Ireland [47], Spain [48], Norway [49], multicentre across different European countries [50,51]), one in Australia [52] and one in New Zealand [53]). Finally, one study presented data from independent studies in different countries (Germany, the Netherlands, U.K., U.S.A.) [54]. All but one studies had a prospective observational design, with a follow-up range between one and 10 years and one study employed a RCT design [39], with a follow-up of 2.5 years. Analytic sample sizes ranged from 116 to 23,307 participants. The majority of studies recruited children and adolescents from school settings [29,30,31,32,33,34,35,37,38,39,44,45,46,47,48,49,51,52,54], and fewer from clinics [20,28,32,43,53] or the general population [36,40,41,42,54], while one study did not provide relevant information [50]. Most studies focused on school-aged children (n = 9) [29,30,31,37,38,45,46,47,51], while others included both pre-schoolers and school-aged children (n = 7) [20,28,35,40,41,50,54], both school-aged children and adolescents (n = 5) [36,39,44,48,49], only pre-schoolers (n = 4) [32,42,43,53] and only adolescents (n = 3) [33,34,52].
Regarding bias, most prospective observational reports had an overall high risk of bias, mainly due to risk of bias arising from measurement of the exposure (Table 2). For example, in several studies, data on exposures derived from self-reported questionnaires, often completed by parents/guardians. It was also unclear how meals were defined and whether clear instructions were provided to the participants on what constituted a meal. Only four reports raised “some concerns” in relation to bias [20,29,32,52] and no report had a low risk of bias. Most reports had a low risk of bias in relation to selection of participants (28/28), post-exposure interventions (24/28) and selection of reported results (23/28). The intervention included in this review [39] had a low risk of bias in four domains (randomisation process, missing data, measurement of the outcome and selection of reported result) and concerns raised in relation to one domain (deviation from intended intervention) (Table 3).

3.2. Main Exposures

A number of different exposures in relation to meal patterns were identified. The most commonly studied exposure was frequency of consumption of/ skipping specific meals, such as breakfast [20,29,31,33,34,36,37,41,43,44,45,46,47,51,54], dinner/evening family meals [20,29,34,43,51], and lunch at home or at school [29,30,36,43,51]. In most studies, meals were self-reported without evidence of a clear, objective definition on the timing or the content of a meal consumed. One study provided a definition of breakfast [31], that is “the first meal in the morning consisting of any solid food, beverages, or both and named by the respondent as ‘breakfast’”. In one study [34], participants were asked how often they had breakfast “which was more than a glass of milk or fruit juice”, whereas in another study [33] they were asked what they usually have for breakfast “on a weekday morning”. Regarding lunch, two studies [29,30] defined school-provided lunch as “a full meal including salad, soup, a sandwich”. Relevant information about how meals were defined and assessed are included in Table 4 and Table 5.
Included studies also examined the association of the consumption of other meal patterns with childhood OV/OB, such as snacking [36,47] (one study [47] provided examples of snacks), eating fast foods [20,34,36,47,52] (three studies provided examples [34,47,52] of fast foods) and having sugary drinks in between meals [40]. Other exposures related to meal patterns included frequency of family meals [28,36], family meal interpersonal quality [28], regular timing of meals [40], eating meals while watching TV [20,34,36], eating while doing homework [36], eating alone [36], eating with friends [36], first and last eating events [32], mealtime setting [36,42] and patterns of breakfast location (combination of the variables “frequency” and “setting” in relation to breakfast) [38]. Some studies examined the association of meal frequency/ eating occasion in childhood obesity, however the definition of meal/eating occasion varied significantly among studies. In Jaeger et al. [50], an eating occasion was defined as any occasion where food or beverages are consumed. In Taylor et al. [53] a separate eating occasion was defined as “the start of the next meal or snack that had to be more than 15 min after the end of the previous meal or snack (i.e. separated by at least four five-minute blocks)”. Stea et al. [49] defined regular breakfast/lunch/dinner/evening meal consumers if eating all meals every day. Two studies assessed meal frequency as the combined frequency of breakfast and evening meals [30,35]. In one study the frequency of eating meals with family was evaluated on a continuous scale from 0 (never) to 8 (>7 meals per week); however, it was unclear how family meals were defined [28].
Data on exposure(s) derived primarily from self-reported questionnaires completed by either parents/caregivers [20,29,30,41,42,43,45,46,47,49,51,54] or children/adolescents [20,33,34,38,44,48,52]. Two studies used dietary recalls [31,32], three studies used dietary records/diaries [36,50,53] and three studies collected data via interviews with parents/caregivers [35,37,40]. All but one studies analysed the exposure of interest as a categorical variable; Jaeger et al. [50] explored meal patterns as the amount of energy intake (kcal) in predefined time slots.
The only cluster-RCT included in this review assessed the effect of eating breakfast in the classroom and of providing breakfast-specific nutrition education in comparison to having breakfast in the school cafeteria, on overweight and obesity among urban children in low-income communities. Data on exposure were collected by teachers (intervention arm) and cafeteria staff (control arm) [39].

3.3. Main findings

Findings are presented in three axes, according to the definition of meal constructs as described by Leech et al. [11]:
a)
meal patterning, including frequency of eating occasions, regularity of meals, meal skipping, meal timing, and spacing of eating occasions
b)
meal format, referring to food type, food combinations, or food sequencing;
c)
meal context, related to the presence of others at a meal, eating while doing activities, meal location.

3.3.1. Meal Patterning

Meal patterning essentially refers to the frequency or timing of eating occasions, either examining the number or the distribution of meals/snacks within the day or focusing on the regularity/skipping of a specific eating occasion, with special interest in breakfast. In relation to meal frequency, with different definitions been used to describe the term, most studies showed no association with adiposity parameters. One study found that higher meal frequency (5 meals per day compared to 3 or fewer meals per day) at baseline was associated with a smaller increase of BMI-z score, a smaller increase in waist-to-height ratio and lower odds of developing obesity at follow up [48]. However, no association of meal frequency with obesity risk [29] or change in BMI-z score at follow up [35,49,53] was found in other studies. A dose-response association was observed between the regularity of mealtimes at age 3 and the risk of developing obesity at age 11 [40]. Specifically, compared to children who always had regular mealtimes, those who usually had regular mealtimes experienced a 23% reduction in the odds of obesity, while those who rarely or never had regular mealtimes had a 38% reduction [40]. Moreover, timing of the first eating episode at baseline was not associated with fat mass, fat-free mass or body fat% in children 3-5 years old at follow-up one year later [32]; however, the same study showed later timing of the last meal of the day at baseline to be associated with increased fat mass and body fat% at follow-up one year later [32].
Breakfast constitutes the most studied meal within meal patterning. Daily consumption of breakfast at baseline compared to less frequent consumption (<7 times/week) had a favourable association with BMI [20] and body fat mass % [20,43] at follow-up, whereas no association with odds of OV/OB [43] was observed. In the de la Rie et al. study [54], such benefit (association with lower BMI) was evident in two of the four cohorts examined. Other studies showed no association of breakfast frequency with OV/OB incidence or/and prevalence [44,47] or change in BMI-z score [34,47]. Moreover, breakfast eating habits have been also assessed as skipping breakfast, with mixed results in relation to obesity risk, and a varying definition of the term across studies. One study used a yes/no variable [41], two studies assessed meal skipping as consuming it “never/rarely” compared to “often/always” [45,46], one study included the category “frequent skippers” as having a meal 0-3 times/week [38], another named a stable breakfast skipping pattern as “eating breakfast <7 times/week” consistently over time [43], and one study assessed meal skipping if participants reported they “did not eat” that meal (option 0) on a 0-7 times/week scale [33]. In most studies, skipping breakfast was associated with weight gain [41,45], increased waist-to-height ratio [45], increased BMI percentile [45], abdominal obesity [45] , increased % of body fat mass [43], and increased overweight risk [45] and OV/OB risk [38]. However, skipping breakfast (yes vs. no) was also associated with decreasing weight trajectory, which was defined as a change in BMI category over time i.e. from overweight at baseline to normal weight at follow-up) [41]. In three studies no association of skipping breakfast with overweight risk (in males) [33], obesity risk [33,45] or abdominal obesity risk [46] was observed.
Child’s sex may also play a role, with some findings indicating a beneficial effect of breakfast consumption especially in girls. In a study with both sexes, breakfast consumption was associated with waist circumference, trunk fat mass and trunk to peripheral fat mass ratio only in girls [20]. Another study conducted only in girls, following them from childhood through adolescence, identified eating breakfast at age 9 (without specifying frequency) as a significant protective predictor against adiposity at age 19 [36]. Skipping breakfast (no breakfast consumption in any day of the week) was also associated with increased OV/OB risk only in girls in one study, but not in boys [33].
Consumption of lunch (as a yes/no variable) and frequency of its consumption were not associated with any parameter of adiposity [30,43,51], but skipping lunch at 9 years of age was identified as a significant risk predictor of adiposity in females at age 19 in one study [36]. Regarding dinner, eating dinner <6 times/week was not associated with changes in fat mass or BMI over time compared to eating dinner every day [43].

3.3.2. Meal Format

The energy and macronutrient content and consumption of specific foods have been examined within the meal format, with a special focus on snacks and fast-food consumption. One study investigated the time-of-day energy intake in relation to obesity risk and found that energy and macronutrient intake distributed in different eating occasions throughout the day, was not significantly associated with BMI-z score [50]. Regarding the content of breakfast, higher frequency consumption of Ready-to-Eat-Cereals (RTEC, 3 vs. 1 times per week of breakfast RTEC) in fourth grade was associated with a decrease of about 2 percentiles of children’s BMI in sixth grade [31].
Children who abstained from sweet and savoury snacks at baseline had decreased OV/OB prevalence, decreased OV/OB risk and decreased change in BMI-z score at follow-up, compared to regular eaters on a daily basis [47]. On the contrary, consumption of sugary drinks in between meals was not associated with obesity risk in one study [41]. Similarly, consumption of fast foods/ takeaway foods was not associated with changes in BMI [34,47,52], body fat % [52], waist circumference [52], odds of developing obesity [52] or change in OV/OB prevalence [47]. However, one study showed that consumption of fast foods less than once a week at baseline was associated with lower BMI-z score, waist circumference, body fat %, trunk fat mass and trunk to peripheral fat mass ratio at follow-up, but only in girls [20].

3.3.3. Meal Context

The aspects of meal context studied were ‘eating with whom’, ‘what doing in parallel’, and ‘place of eating’ and in particular, eating with family, eating while doing another activity, esp. screen use, and eating at school. Studies have mainly focused on breakfast and dinner, with scarce reports on lunch.
Family meals were associated in one study with reduced obesity prevalence at follow-up, with each additional family meal significantly reducing obesity prevalence by 4% [28]. Two studies assessed eating breakfast with family, and both showed beneficial effect [29,51]. Children who consumed family breakfast three to seven times a week at baseline had decreased odds of OV/OB two years later, compared to those who never had breakfast with family [51], while children with overweight at baseline and healthy weight at follow-up were more likely to eat breakfast together with at least one member of the family than children with overweight, both at baseline and follow-up [29]. Four studies examined the role of family dinner in obesity risk [20,29,34,51], two of which found a protective role, but the other two no association. In one study, children who had family dinners three to seven times per week at baseline were more likely to have lower BMI and reduced odds of overweight/obesity two years later [51]. Similarly, daily family dinners, in comparison to non-daily dinners (≤6 times/week), were associated with decreased BMI-z score, waist circumference, trunk fat mass and trunk to peripheral fat mass ratio, in girls but not boys [20]. On the other hand, two studies found no association of frequency of family dinners and BMI over time [29,34].
Four studies examined the effect of watching TV while eating meals [20,34,36,42]. In specific, one study showed lower BMI-z score, waist circumference, body fat%, trunk fat mass and trunk to peripheral fat mass ratio in boys and lower body fat% in girls that ate meals while watching TV less than once a week, compared to those who ate meals in front of TV more frequently [20]. Greater mealtime screen use and less use of a table while eating a meal (described as intermediate and informal meal settings) was associated with increasing overweight weight trajectories (i.e. being at the “normal weight” category at baseline and at the “overweight” category at follow up), compared to lower mealtime screen use and eating a meal on the table (formal meal setting) [42]. One study, however, found that eating while watching TV at age 9 was a significant protective predictor of adiposity in females at age 19 [36], which was highlighted as “surprising” by authors, while no associations were observed in another study [34]. Also, a more positive interpersonal quality of family meals at baseline, characterised by the involvement of conversations without media distractions, was not associated with obesity prevalence at follow up, compared to a less positive interpersonal quality of family meals that involved more media distractions and less conversations [28]. Finally, in one study, eating with friends, eating while doing homework and eating in the bedroom were also identified as protective factors against adiposity in females, findings which have also been highlighted as “surprising” by the authors, who nevertheless commented that the findings may imply a mitigation effect of such eating behaviours when healthy snacks are offered, assumption which might be supported by their finding of “parents fixing a snack” being a protective adiposity predictor [36].
Meals served at school have been also studied in regards to obesity risk [29,37,39]. Consumption of breakfast at school (yes/no) in fifth grade was not associated with obesity risk in eighth grade [37], neither an intervention to increase breakfast consumption at school had an effect in the combined OV/OB incidence or prevalence [39]. However, when incidence and prevalence of obesity were examined separately, the incidence of obesity alone was higher in intervention schools than in control schools (11.6% vs 4.4%; OR 2.43; 95% CI, 1.47-4.00), as well as prevalence of obesity (28.0% vs 21.2%; OR 1.46; 95% CI, 1.11-1.92) after 2.5 years of intervention [39]. Only one study examined consumption of school-provided lunch (yes/no), and found that it was associated with decreased odds of weight gain at follow up 3 years later [29]. In all studies, no information on whether a school-provided meal reduced meal skipping, replaced a meal eaten at home or was consumed in addition to a home-provided meal, was mentioned.

4. Discussion

This systematic review collected evidence on the associations of meal patterns with childhood OV/OB and MUO risk. In the absence of published studies on MUO risk, only evidence on OV/OB was presented. To our knowledge, this is the first systematic review that captures different dimensions of meal patterns (patterning, format, context) and focuses on studies with a longitudinal design (prospective studies and randomised controlled trials), aiming to explore aetiologic associations between meal patterns and OV/OB and MUO risk. Regular consumption of breakfast and family meals, as well as avoiding watching TV while eating, may be protective factors against childhood OV/OB, whereas meal skipping (primarily breakfast) may be a detrimental factor. Mixed effects of meal frequency on OV/OB risk were observed. No effects were observed regarding frequency of lunch consumption or of fast-food consumption and of meals served at school, while there was insufficient evidence to support the role of other meal patterns such as meal timing and eating in social contexts other than family in OV/OB risk.
Despite methodological considerations, frequent/daily consumption of breakfast (or avoiding breakfast skipping) has been, according to current findings, highlighted as a protective factor against childhood OV/OB longitudinally, indicating a potential aetiologic association between breakfast consumption and OV/OB risk. Recently published systematic reviews of both cross-sectional and longitudinal studies are in agreement with these findings, particularly regarding breakfast skipping [16,55]. This constitutes an important public health message for preventive policies and practices, especially taking into consideration that literature shows a discernible decline in breakfast consumption in children while entering adolescence [56]. Promotion of frequent/daily consumption of breakfast throughout childhood and adolescence would be an efficient strategy towards reducing OV/OB risk.
The self-reported data collection methods employed in most included studies might have variously impacted current findings. For example, most of the studies did not provide guidance as to what constitutes a meal and a small number of studies used varied definitions, leading to ambiguities regarding the studies’ comparison and results interpretation. Participant-identified eating occasions are prone to subjectivity, as the interpretation and allocation of each eating event may vary significantly [11]. Future studies should ensure the use of robust definitions for meals, such as the proposed definition for breakfast by O’Neil et al. [57] as “the first meal of the day that breaks the fast after the longest period of sleep and is consumed within 2 to 3 hours of waking; it is comprised of food or beverage from at least one food group, and may be consumed at any location”. Lack of a universally-accepted definition of breakfast may prove challenging in the identification of pathways through which breakfast consumption may play a significant role in the development of childhood OV/OB and MUO. Also, the use of methodologically more reliable assessment tools for the evaluation of meals in future research, such as dietary records and recalls, will not only provide more reliable results on the role of breakfast/meal frequency, but also on the role of meal format/composition in the development of OV/OB.
Results from this review also highlighted a potential protective effect of family-shared meals against OV/OB risk, which is supported by older systematic reviews/meta-analyses of both cross-sectional and longitudinal studies [9,58]. The parental influence in shaping children’s eating behaviours has been well documented, from either the perspective of parenting feeding practices, dietary habits or weight status [59,60,61]. Family shared meals could be considered a factor under the construct of home organisation within the family environment, and associations have been found between family meal routines and childhood and adolescence obesity markers, after correcting for the moderating role of socioeconomic factors [62]. It would be interesting to explore which parameters of the family meal routines and family environment mediate this protective effect towards OV/OB risk. For example, an earlier, cross-sectional study by Skafida et al. [63] showed that parameters such as children eating the same food as their parents, having conversations with parents during meals and having “enjoyable” mealtimes were positively associated with children’s diet quality, while eating with parents at the same time was not a significant predictor [63]. Only one study in this review [28] assessed interpersonal quality of family meals and specifically conversations with family during meals, compared to media distraction, with no significant results in relation to OV/OB risk. Future research should focus on the longitudinal effect of family meal routines and related parameters, which could also include food availability, food quality and consumption of homemade foods on the OV/OB risk. The family environment should be also considered as a potential target for future interventions that will focus on the improvement of the frequency and quality of family meals and explore their effect on weight status.
Included studies that investigated the effect of eating meals while being distracted by media and particularly by watching TV on OV/OB risk indicated a protective effect when abstaining from TV while eating. Findings from a systematic review showed a positive association between TV viewing and consumption of energy-dense foods, such as pizza, fried food, sweets and sugar-sweetened beverages, and a negative association with consumption of fruits and vegetables [64]. TV viewing has been hypothesised to increase OV/OB risk through increased sedentary time [65,66], influence of TV advertisements of energy-dense foods [67], promotion of mindless eating during viewing [68], and increased snacking [66].
Although this review identified several longitudinal studies involving the family environment, considerably fewer assessed meal patterns in the school environment (in the form of school-provided meals); thus, there is insufficient evidence to support a positive or detrimental effect on children’s weight status. Evidence from two large free school meal programs in the U.S.A., the Community Eligibility Provision and the Healthy, Hunger-Free Kids Act have shown mixed results on childhood obesity trends [69,70]. Provision of free breakfast and lunch to students had led to a modest decrease in obesity prevalence in the Community Eligibility Provision program within 5 years of implementation [69], whereas no effect was recorded in the Healthy, Hunger-Free Kids Act study on obesity risk [70]. These results may be due to the fact that it is unknown how school meals affect children’s daily meal and dietary intake, that is whether they replace existing meals or are added to a given meal pattern leading probably to increased dietary intake. Even though it may be unclear whether provision of meals at schools are associated OV/OB risk, such school programs have documented improvements in diet quality [71], food security [72] and academic performance [72]. It would be essential for future research studying the effect of school provided meals on childhood OV/OB risk to assess the quality and quantity of school-provided meals, including snacks, and explore whether they replace or are consumed additionally to the rest of the meals by correcting also for total energy intake as a confounding factor.
Some evidence suggests that sex might have a mediating role in meal patterns and OV/OB risk, and this was mainly observed in the literature exploring the effect of breakfast consumption. According to findings from two studies [20,33], regular breakfast consumption was associated with decreased adiposity markers, while skipping breakfast with increased OV/OB risk in girls, but not boys. One of these studies also found a protective effect of daily family dinners on obesity markers in girls but not in boys. Mahmood et al. [51] suggested that a potential explanation for sex differences might be due to differences in dietary and social behaviours. Girls tend to eat more frequently with family and friends than boys [73], whereas boys eat more takeaway meals at home, compared to girls [74]. Also, girls may be more prone to societal influences of dieting, leading to increased prevalence of meal (particularly breakfast) skipping [73]. It is unknown whether dieting and social behaviours would explain sex difference in the included studies. One of the studies in this review explored the impact of dieting on OV/OB risk and found no associations regardless of sex, but the variable “dieting” was not included as a confounder for the association of breakfast skipping and OV/OB risk. The other study did not explore other dietary or social behaviours that could potentially explain the different results according to sex. Biological dimension of sex mediating the role between frequency of consumption of meals and OV/OB risk is also largely unknown and constitutes a potential area for future research.
It should also be noted that the current review identified only one RCT regarding meal patterns (provision of breakfast) and OV/OB markers and no studies for MUO risk. Given the importance of preventative measures against obesity-related comorbidities, future research should focus on the role of meal patterns and the family environment in the development of comorbidities in children with OV/OB. Also, future interventions could target aspects of meal patterning, format and/or context in preventive interventions.
This review has strengths and limitations. In order to promote reproducibility and transparency, the PRISMA guidelines and a detailed search strategy were implemented. The inclusion of most recent studies with a longitudinal design is considered a strength of the review, so evidence on potential aetiologic associations of the role of meal patterns and OV/OB risk and MUO risk is synthesised. Findings are, however, affected by the scientific quality of the included studies, presenting in most of the cases with high or very high risk of bias due to the large heterogeneity on the assessment of exposures [particularly deriving from self (parent)-reported questionnaires]. Also, search was limited in two, widely available databases (Pubmed/MEDLINE and Scopus) and reference lists of included studies; it was not possible to conduct the search in databases where subscription was required. Finally, included studies were conducted in Westernised countries, limiting the generalisabilty of the findings in non-westernised cultural settings.

5. Conclusions

Some evidence supports a protective role of regular/daily breakfast consumption, regular shared family meals and avoiding watching TV while dining against OV/OB risk, deriving mainly from longitudinal studies, while no relevant published study reporting on childhood MUO risk was identified. There was insufficient evidence of the role of other meals such as lunch, other meal patterns such as meal timing, and other social environments such as consuming meals at school, in OV/OB risk. Overall, the quality of the findings is poor due to the high bias of the included studies. The use of harmonised definitions for the assessment of different meals, as well as better methodological approaches, are warranted to provide more robust results in future studies. Future interventions should also target the family environment with a view of determining the protective parameters of shared family meals in OV/OB risk and MUO risk.

Supplementary Materials

The following supporting information can be downloaded at the website of this paper posted on Preprints.org, Supplementary Material S1: PRISMA 2020 checklist; Supplementary Material S2: Detailed search strategy in MEDLINE/Pubmed and Scopus

Author Contributions

Conceptualization, M.D.K. and M.G.; methodology, G.S, E.B, MG, V.B. and U. S.; formal analysis, G.S, E.B, A.K. and V. B.; investigation, G.S., E.B. and A.K.; resources, G.S, A.K; writing—original draft preparation, G.S, E.B.; writing—review and editing, M.G., M.Y., V.B., M.D.K. G.D.; supervision, M.D.K. and G.D.; funding acquisition, M.D.K. and G.D. All authors have read and agreed to the published version of the manuscript.

Funding

This systematic review has been conducted within the BIO-STREAMS project, which has received funding from the European Union’s HORIZON 2022 research and innovation program under grant agreement No 101080718.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. NCD Risk Factor Collaboration (NCD-RisC). Worldwide Trends in Underweight and Obesity from 1990 to 2022: A Pooled Analysis of 3663 Population-Representative Studies with 222 Million Children, Adolescents, and Adults. Lancet Lond. Engl. 2024, 403 (10431), 1027–1050. [CrossRef]
  2. World Health Organisation Regional Office for Europe. Report on the Fifth Round of Data Collection, 2018–2020: WHO European Childhood Obesity Surveillance Initiative (COSI); WHO/EURO:2022-6594-46360-67071; World Health Organisation, 2022; p 70. https://iris.who.int/bitstream/handle/10665/363950/WHO-EURO-2022-6594-46360-67071-eng.pdf?sequence=2.
  3. Tsoi, M.-F.; Li, H.-L.; Feng, Q.; Cheung, C.-L.; Cheung, T. T.; Cheung, B. M. Y. Prevalence of Childhood Obesity in the United States in 1999-2018: A 20-Year Analysis. Obes. Facts 2022, 15 (4), 560–569. [CrossRef]
  4. Shaban Mohamed, M. A.; AbouKhatwa, M. M.; Saifullah, A. A.; Hareez Syahmi, M.; Mosaad, M.; Elrggal, M. E.; Dehele, I. S.; Elnaem, M. H. Risk Factors, Clinical Consequences, Prevention, and Treatment of Childhood Obesity. Child. Basel Switz. 2022, 9 (12). [CrossRef]
  5. Lister, N. B.; Baur, L. A.; Felix, J. F.; Hill, A. J.; Marcus, C.; Reinehr, T.; Summerbell, C.; Wabitsch, M. Child and Adolescent Obesity. Nat. Rev. Dis. Primer 2023, 9 (1), 24. [CrossRef]
  6. Liberali, R.; Kupek, E.; Assis, M. A. A. de. Dietary Patterns and Childhood Obesity Risk: A Systematic Review. Child. Obes. Print 2020, 16 (2), 70–85. [CrossRef]
  7. Jakobsen, D. D.; Brader, L.; Bruun, J. M. Association between Food, Beverages and Overweight/Obesity in Children and Adolescents-A Systematic Review and Meta-Analysis of Observational Studies. Nutrients 2023, 15 (3). [CrossRef]
  8. Bradlee, M. L.; Singer, M. R.; Qureshi, M. M.; Moore, L. L. Food Group Intake and Central Obesity among Children and Adolescents in the Third National Health and Nutrition Examination Survey (NHANES III). Public Health Nutr. 2010, 13 (6), 797–805. [CrossRef]
  9. Dallacker, M.; Hertwig, R.; Mata, J. The Frequency of Family Meals and Nutritional Health in Children: A Meta-Analysis. Obes. Rev. Off. J. Int. Assoc. Study Obes. 2018, 19 (5), 638–653. [CrossRef]
  10. Leech, R. M.; Worsley, A.; Timperio, A.; McNaughton, S. A. Characterizing Eating Patterns: A Comparison of Eating Occasion Definitions. Am. J. Clin. Nutr. 2015, 102 (5), 1229–1237. [CrossRef]
  11. Leech, R. M.; Worsley, A.; Timperio, A.; McNaughton, S. A. Understanding Meal Patterns: Definitions, Methodology and Impact on Nutrient Intake and Diet Quality. Nutr. Res. Rev. 2015, 28 (1), 1–21. [CrossRef]
  12. Guan, V. X.; Probst, Y. C.; Neale, E. P.; Batterham, M. J.; Tapsell, L. C. Identifying Usual Food Choices at Meals in Overweight and Obese Study Volunteers: Implications for Dietary Advice. Br. J. Nutr. 2018, 120 (4), 472–480. [CrossRef]
  13. Murakami, K.; Livingstone, M. B. E. Meal and Snack Frequency in Relation to Diet Quality in US Children and Adolescents: The National Health and Nutrition Examination Survey 2003-2012. Public Health Nutr. 2016, 19 (9), 1635–1644. [CrossRef]
  14. Murakami, K. Nutritional Quality of Meals and Snacks Assessed by the Food Standards Agency Nutrient Profiling System in Relation to Overall Diet Quality, Body Mass Index, and Waist Circumference in British Adults. Nutr. J. 2017, 16 (1), 57. [CrossRef]
  15. Kaisari, P.; Yannakoulia, M.; Panagiotakos, D. B. Eating Frequency and Overweight and Obesity in Children and Adolescents: A Meta-Analysis. Pediatrics 2013, 131 (5), 958–967. [CrossRef]
  16. Monzani, A.; Ricotti, R.; Caputo, M.; Solito, A.; Archero, F.; Bellone, S.; Prodam, F. A Systematic Review of the Association of Skipping Breakfast with Weight and Cardiometabolic Risk Factors in Children and Adolescents. What Should We Better Investigate in the Future? Nutrients 2019, 11 (2). [CrossRef]
  17. Zou, M.; Northstone, K.; Perry, R.; Johnson, L.; Leary, S. The Association between Later Eating Rhythm and Adiposity in Children and Adolescents: A Systematic Review and Meta-Analysis. Nutr. Rev. 2022, 80 (6), 1459–1479. [CrossRef]
  18. Sharma, V.; Coleman, S.; Nixon, J.; Sharples, L.; Hamilton-Shield, J.; Rutter, H.; Bryant, M. A Systematic Review and Meta-Analysis Estimating the Population Prevalence of Comorbidities in Children and Adolescents Aged 5 to 18 Years. Obes. Rev. Off. J. Int. Assoc. Study Obes. 2019, 20 (10), 1341–1349. [CrossRef]
  19. Reinehr, T.; Tittel, S. R.; Holle, R.; Wiegand, S.; Gellhaus, I.; Hebebrand, J.; Greber-Platzer, S.; Denzer, C.; Linke, S.; Kiess, W.; Holl, R. W. Comparison of Cardiovascular Risk Factors between Children and Adolescents with Classes III and IV Obesity: Findings from the APV Cohort. Int. J. Obes. 2005 2021, 45 (5), 1061–1073. [CrossRef]
  20. Gingras, V.; Rifas-Shiman, S. L.; Taveras, E. M.; Oken, E.; Hivert, M.-F. Dietary Behaviors throughout Childhood Are Associated with Adiposity and Estimated Insulin Resistance in Early Adolescence: A Longitudinal Study. 2018, 15 (1). [CrossRef]
  21. López-Contreras, I. N.; Vilchis-Gil, J.; Klünder-Klünder, M.; Villalpando-Carrión, S.; Flores-Huerta, S. Dietary Habits and Metabolic Response Improve in Obese Children Whose Mothers Received an Intervention to Promote Healthy Eating: Randomized Clinical Trial. BMC Public Health 2020, 20 (1), 1240. [CrossRef]
  22. Page, M. J.; McKenzie, J. E.; Bossuyt, P. M.; Boutron, I.; Hoffmann, T. C.; Mulrow, C. D.; Shamseer, L.; Tetzlaff, J. M.; Akl, E. A.; Brennan, S. E.; Chou, R.; Glanville, J.; Grimshaw, J. M.; Hróbjartsson, A.; Lalu, M. M.; Li, T.; Loder, E. W.; Mayo-Wilson, E.; McDonald, S.; McGuinness, L. A.; Stewart, L. A.; Thomas, J.; Tricco, A. C.; Welch, V. A.; Whiting, P.; Moher, D. The PRISMA 2020 Statement: An Updated Guideline for Reporting Systematic Reviews. BMJ 2021, 372, n71. [CrossRef]
  23. Page, M. J.; Moher, D.; Bossuyt, P. M.; Boutron, I.; Hoffmann, T. C.; Mulrow, C. D.; Shamseer, L.; Tetzlaff, J. M.; Akl, E. A.; Brennan, S. E.; Chou, R.; Glanville, J.; Grimshaw, J. M.; Hróbjartsson, A.; Lalu, M. M.; Li, T.; Loder, E. W.; Mayo-Wilson, E.; McDonald, S.; McGuinness, L. A.; Stewart, L. A.; Thomas, J.; Tricco, A. C.; Welch, V. A.; Whiting, P.; McKenzie, J. E. PRISMA 2020 Explanation and Elaboration: Updated Guidance and Exemplars for Reporting Systematic Reviews. BMJ 2021, 372, n160. [CrossRef]
  24. Thomas, J.; Kneale, D.; McKenzie, J. E.; Brennan, S. E.; Bhaumik, S. Determining the Scope of the Review and the Questions It Will Address. In Cochrane Handbook for Systematic Reviews of Interventions; 2019; pp 13–31. [CrossRef]
  25. Independent Accountability Panel for Every Woman, Every Child, Every Adolescent. Report 2017: Transformative Accountability for Adolescents: Accountability for the Health and Human Rights of Women, Children and Adolescents in the 2030 Agenda; World Health Organisation: Geneva, 2017.
  26. Higgins, J. P. T.; Morgan, R. L.; Rooney, A. A.; Taylor, K. W.; Thayer, K. A.; Silva, R. A.; Lemeris, C.; Akl, E. A.; Bateson, T. F.; Berkman, N. D.; Glenn, B. S.; Hróbjartsson, A.; LaKind, J. S.; McAleenan, A.; Meerpohl, J. J.; Nachman, R. M.; Obbagy, J. E.; O’Connor, A.; Radke, E. G.; Savović, J.; Schünemann, H. J.; Shea, B.; Tilling, K.; Verbeek, J.; Viswanathan, M.; Sterne, J. A. C. A Tool to Assess Risk of Bias in Non-Randomized Follow-up Studies of Exposure Effects (ROBINS-E). Environ. Int. 2024, 186, 108602. [CrossRef]
  27. Sterne, J. A. C.; Savović, J.; Page, M. J.; Elbers, R. G.; Blencowe, N. S.; Boutron, I.; Cates, C. J.; Cheng, H.-Y.; Corbett, M. S.; Eldridge, S. M.; Emberson, J. R.; Hernán, M. A.; Hopewell, S.; Hróbjartsson, A.; Junqueira, D. R.; Jüni, P.; Kirkham, J. J.; Lasserson, T.; Li, T.; McAleenan, A.; Reeves, B. C.; Shepperd, S.; Shrier, I.; Stewart, L. A.; Tilling, K.; White, I. R.; Whiting, P. F.; Higgins, J. P. T. RoB 2: A Revised Tool for Assessing Risk of Bias in Randomised Trials. BMJ 2019, 366, l4898. [CrossRef]
  28. Berge, J. M.; Hazzard, V. M.; Trofholz, A.; Noser, A. E.; Hochgraf, A.; Neumark-Sztainer, D. Longitudinal Associations between Family Meal Quality and Quantity: Does One Matter More for Child, Parent, and Family Health and Well-Being or Are They Synergistic? Appetite 2023, 191. [CrossRef]
  29. Chang, Y.; Gable, S. Predicting Weight Status Stability and Change from Fifth Grade to Eighth Grade: The Significant Role of Adolescents’ Social-Emotional Well-Being. J. Adolesc. Health Off. Publ. Soc. Adolesc. Med. 2013, 52 (4), 448–455. [CrossRef]
  30. Chang, Y.; Halgunseth, L. C. The Association between Family Meals and Early-Adolescents’ Weight Status Change in the Context of Parental Discipline Practices: The Moderating Roles of Ethnicity and Acculturation. J. Immigr. Minor. Health 2015, 17 (2), 450–458. [CrossRef]
  31. Balvin Frantzen, L.; Treviño, R. P.; Echon, R. M.; Garcia-Dominic, O.; DiMarco, N. Association between Frequency of Ready-to-Eat Cereal Consumption, Nutrient Intakes, and Body Mass Index in Fourth- to Sixth-Grade Low-Income Minority Children. J. Acad. Nutr. Diet. 2013, 113 (4), 511–519. [CrossRef]
  32. Goetz, A. R.; Jindal, I.; Moreno, J. P.; Puyau, M. R.; Adolph, A. L.; Musaad, S.; Butte, N. F.; Bacha, F. The Roles of Sleep and Eating Patterns in Adiposity Gain among Preschool-Aged Children. Am. J. Clin. Nutr. 2022, 116 (5), 1334–1342. [CrossRef]
  33. Liechty, J. M.; Lee, M.-J. Body Size Estimation and Other Psychosocial Risk Factors for Obesity Onset among US Adolescents: Findings from a Longitudinal Population Level Study. Int. J. Obes. 2005 2015, 39 (4), 601–607. [CrossRef]
  34. Lipsky, L. M.; Haynie, D. L.; Liu, D.; Chaurasia, A.; Gee, B.; Li, K.; Iannotti, R. J.; Simons-Morton, B. Trajectories of Eating Behaviors in a Nationally Representative Cohort of U.S. Adolescents during the Transition to Young Adulthood. Int. J. Behav. Nutr. Phys. Act. 2015, 12, 138. [CrossRef]
  35. Loren, D. M.; Rea, E. M.; Harber, K. A.; Bohnert, A. M. An Ecological Model of Weight Gain among School-Age Children. Health Psychol. Off. J. Div. Health Psychol. Am. Psychol. Assoc. 2022, 41 (3), 193–203. [CrossRef]
  36. Narla, A.; Rehkopf, D. H. Novel Ranking of Protective and Risk Factors for Adolescent Adiposity in US Females. 2019, 5 (2), 177–186. [CrossRef]
  37. Sudharsanan, N.; Romano, S.; Cunningham, S. A. School Breakfast Receipt and Obesity among American Fifth- and Eighth-Graders. J. Acad. Nutr. Diet. 2016, 116 (4), 599-607.e3. [CrossRef]
  38. Wang, S.; Schwartz, M. B.; Shebl, F. M.; Read, M.; Henderson, K. E.; Ickovics, J. R. School Breakfast and Body Mass Index: A Longitudinal Observational Study of Middle School Students. Pediatr. Obes. 2017, 12 (3), 213–220. [CrossRef]
  39. Polonsky, H. M.; Bauer, K. W.; Fisher, J. O.; Davey, A.; Sherman, S.; Abel, M. L.; Hanlon, A.; Ruth, K. J.; Dale, L. C.; Foster, G. D. Effect of a Breakfast in the Classroom Initiative on Obesity in Urban School-Aged Children: A Cluster Randomized Clinical Trial. JAMA Pediatr. 2019, 173 (4), 326–333. [CrossRef]
  40. Anderson, S. E.; Sacker, A.; Whitaker, R. C.; Kelly, Y. Self-Regulation and Household Routines at Age Three and Obesity at Age Eleven: Longitudinal Analysis of the UK Millennium Cohort Study. Int. J. Obes. 2005 2017, 41 (10), 1459–1466. [CrossRef]
  41. Kelly, Y.; Patalay, P.; Montgomery, S.; Sacker, A. BMI Development and Early Adolescent Psychosocial Well-Being: UK Millennium Cohort Study. 2016, 138 (6). [CrossRef]
  42. Parkes, A.; Green, M.; Pearce, A. Do Bedroom Screens and the Mealtime Environment Shape Different Trajectories of Child Overweight and Obesity? Research Using the Growing Up in Scotland Study. Int. J. Obes. 2005 2020, 44 (4), 790–802. [CrossRef]
  43. Wijtzes, A. I.; Jansen, W.; Bouthoorn, S. H.; van Lenthe, F. J.; Franco, O. H.; Hofman, A.; Jaddoe, V. W. V.; Raat, H. Meal-Skipping Behaviors and Body Fat in 6-Year-Old Children. J. Pediatr. 2016, 168, 118-125.e2. [CrossRef]
  44. de Winter, A. F.; Visser, L.; Verhulst, F. C.; Vollebergh, W. A. M.; Reijneveld, S. A. Longitudinal Patterns and Predictors of Multiple Health Risk Behaviors among Adolescents: The TRAILS Study. Prev. Med. 2016, 84, 76–82. [CrossRef]
  45. Traub, M.; Lauer, R.; Kesztyüs, T.; Wartha, O.; Steinacker, J. M.; Kesztyüs, D. Skipping Breakfast, Overconsumption of Soft Drinks and Screen Media: Longitudinal Analysis of the Combined Influence on Weight Development in Primary Schoolchildren. BMC Public Health 2018, 18 (1), 363. [CrossRef]
  46. Kesztyüs, D.; Traub, M.; Lauer, R.; Kesztyüs, T.; Steinacker, J. M. Skipping Breakfast Is Detrimental for Primary School Children: Cross-Sectional Analysis of Determinants for Targeted Prevention. BMC Public Health 2017, 17 (1), 258. [CrossRef]
  47. Bel-Serrat, S.; Heinen, M. M.; Mehegan, J.; O’Brien, S.; Eldin, N.; Murrin, C. M.; Kelleher, C. C. Predictors of Weight Status in School-Aged Children: A Prospective Cohort Study. Eur. J. Clin. Nutr. 2019, 73 (9), 1299–1306. [CrossRef]
  48. Juton, C.; Berruezo, P.; Torres, S.; Castañer, O.; Según, G.; Fitó, M.; Homs, C.; Gómez, S. F.; Schröder, H. Association between Meal Frequency and Weight Status in Spanish Children: A Prospective Cohort Study. Nutrients 2023, 15 (4). [CrossRef]
  49. Stea, T. H.; Vik, F. N.; Bere, E.; Svendsen, M. V.; Oellingrath, I. M. Meal Pattern among Norwegian Primary-School Children and Longitudinal Associations between Meal Skipping and Weight Status. Public Health Nutr. 2015, 18 (2), 286–291. [CrossRef]
  50. Jaeger, V.; Koletzko, B.; Luque, V.; Gispert-Llauradó, M.; Gruszfeld, D.; Socha, P.; Verduci, E.; Zuccotti, G. V.; Etienne, L.; Grote, V. Time of Dietary Energy and Nutrient Intake and Body Mass Index in Children: Compositional Data Analysis from the Childhood Obesity Project (CHOP) Trial. 2022, 14 (20). [CrossRef]
  51. Mahmood, L.; Gonzalez-Gil, E. M.; Makrilakis, K.; Liatis, S.; Schwarz, P.; Herrmann, S.; Willems, R.; Cardon, G.; Latomme, J.; Rurik, I.; Radó, S.; Iotova, V.; Usheva, N.; Tankova, T.; Karaglani, E.; Manios, Y.; Moreno, L. A. Cross-Sectional and Longitudinal Associations between Family Meals Frequency and Children’s Overweight/Obesity in Families at High Risk of Type 2 Diabetes: The Feel4Diabetes-Study. 2023, 18 (4). [CrossRef]
  52. Gopinath, B.; Flood, V. M.; Burlutsky, G.; Louie, J. C. Y.; Baur, L. A.; Mitchell, P. Frequency of Takeaway Food Consumption and Its Association with Major Food Group Consumption, Anthropometric Measures and Blood Pressure during Adolescence. 2016, 115 (11), 2025–2030. [CrossRef]
  53. Taylor, R. W.; Iosua, E.; Heath, A.-L. M.; Gray, A. R.; Taylor, B. J.; Lawrence, J. A.; Hanna, M.; Cameron, S. L.; Sayers, R.; Galland, B. Eating Frequency in Relation to BMI in Very Young Children: A Longitudinal Analysis. Public Health Nutr. 2017, 20 (8), 1372–1379. [CrossRef]
  54. de la Rie, S.; Washbrook, E.; Perinetti Casoni, V.; Waldfogel, J.; Kwon, S. J.; Dräger, J.; Schneider, T.; Olczyk, M.; Boinet, C.; Keizer, R. The Role of Energy Balance Related Behaviors in Socioeconomic Inequalities in Childhood Body Mass Index: A Comparative Analysis of Germany, the Netherlands, the United Kingdom, and the United States. Soc. Sci. Med. 1982 2023, 317, 115575. [CrossRef]
  55. Wang, K.; Niu, Y.; Lu, Z.; Duo, B.; Effah, C. Y.; Guan, L. The Effect of Breakfast on Childhood Obesity: A Systematic Review and Meta-Analysis. Front. Nutr. 2023, 10, 1222536. [CrossRef]
  56. Alexy, U.; Wicher, M.; Kersting, M. Breakfast Trends in Children and Adolescents: Frequency and Quality. Public Health Nutr. 2010, 13 (11), 1795–1802. [CrossRef]
  57. O’Neil, C. E.; Byrd-Bredbenner, C.; Hayes, D.; Jana, L.; Klinger, S. E.; Stephenson-Martin, S. The Role of Breakfast in Health: Definition and Criteria for a Quality Breakfast. J. Acad. Nutr. Diet. 2014, 114 (12 Suppl), S8–S26. [CrossRef]
  58. Hammons, A. J.; Fiese, B. H. Is Frequency of Shared Family Meals Related to the Nutritional Health of Children and Adolescents? Pediatrics 2011, 127 (6), e1565-1574. [CrossRef]
  59. Hampl, S. E.; Hassink, S. G.; Skinner, A. C.; Armstrong, S. C.; Barlow, S. E.; Bolling, C. F.; Avila Edwards, K. C.; Eneli, I.; Hamre, R.; Joseph, M. M.; Lunsford, D.; Mendonca, E.; Michalsky, M. P.; Mirza, N.; Ochoa, E. R., Jr; Sharifi, M.; Staiano, A. E.; Weedn, A. E.; Flinn, S. K.; Lindros, J.; Okechukwu, K. Clinical Practice Guideline for the Evaluation and Treatment of Children and Adolescents With Obesity. Pediatrics 2023, 151 (2), e2022060640. [CrossRef]
  60. Savage, J. S.; Fisher, J. O.; Birch, L. L. Parental Influence on Eating Behavior: Conception to Adolescence. J. Law Med. Ethics J. Am. Soc. Law Med. Ethics 2007, 35 (1), 22–34. [CrossRef]
  61. Mahmood, L.; Flores-Barrantes, P.; Moreno, L. A.; Manios, Y.; Gonzalez-Gil, E. M. The Influence of Parental Dietary Behaviors and Practices on Children’s Eating Habits. Nutrients 2021, 13 (4). [CrossRef]
  62. Bates, C. R.; Buscemi, J.; Nicholson, L. M.; Cory, M.; Jagpal, A.; Bohnert, A. M. Links between the Organization of the Family Home Environment and Child Obesity: A Systematic Review. Obes. Rev. Off. J. Int. Assoc. Study Obes. 2018, 19 (5), 716–727. [CrossRef]
  63. Skafida, V. The Family Meal Panacea: Exploring How Different Aspects of Family Meal Occurrence, Meal Habits and Meal Enjoyment Relate to Young Children’s Diets. Sociol. Health Illn. 2013, 35 (6), 906–923. [CrossRef]
  64. Avery, A.; Anderson, C.; McCullough, F. Associations between Children’s Diet Quality and Watching Television during Meal or Snack Consumption: A Systematic Review. Matern. Child. Nutr. 2017, 13 (4). [CrossRef]
  65. Ekelund, U.; Brage, S.; Froberg, K.; Harro, M.; Anderssen, S. A.; Sardinha, L. B.; Riddoch, C.; Andersen, L. B. TV Viewing and Physical Activity Are Independently Associated with Metabolic Risk in Children: The European Youth Heart Study. PLoS Med. 2006, 3 (12), e488. [CrossRef]
  66. Collings, P. J.; Kelly, B.; West, J.; Wright, J. Associations of TV Viewing Duration, Meals and Snacks Eaten When Watching TV, and a TV in the Bedroom with Child Adiposity. Obes. Silver Spring Md 2018, 26 (10), 1619–1628. [CrossRef]
  67. Potvin Kent, M.; Soares Guimaraes, J.; Pritchard, M.; Remedios, L.; Pauzé, E.; L’Abbé, M.; Mulligan, C.; Vergeer, L.; Weippert, M. Differences in Child and Adolescent Exposure to Unhealthy Food and Beverage Advertising on Television in a Self-Regulatory Environment. BMC Public Health 2023, 23 (1), 555. [CrossRef]
  68. Ogden, J.; Coop, N.; Cousins, C.; Crump, R.; Field, L.; Hughes, S.; Woodger, N. Distraction, the Desire to Eat and Food Intake. Towards an Expanded Model of Mindless Eating. Appetite 2013, 62, 119–126. [CrossRef]
  69. Localio, A. M.; Knox, M. A.; Basu, A.; Lindman, T.; Walkinshaw, L. P.; Jones-Smith, J. C. Universal Free School Meals Policy and Childhood Obesity. Pediatrics 2024, 153 (4), e2023063749. [CrossRef]
  70. Kenney, E. L.; Barrett, J. L.; Bleich, S. N.; Ward, Z. J.; Cradock, A. L.; Gortmaker, S. L. Impact Of The Healthy, Hunger-Free Kids Act On Obesity Trends. Health Aff. Proj. Hope 2020, 39 (7), 1122–1129. [CrossRef]
  71. Kinderknecht, K.; Harris, C.; Jones-Smith, J. Association of the Healthy, Hunger-Free Kids Act With Dietary Quality Among Children in the US National School Lunch Program. JAMA 2020, 324 (4), 359–368. [CrossRef]
  72. Cohen, J. F. W.; Hecht, A. A.; McLoughlin, G. M.; Turner, L.; Schwartz, M. B. Universal School Meals and Associations with Student Participation, Attendance, Academic Performance, Diet Quality, Food Security, and Body Mass Index: A Systematic Review. Nutrients 2021, 13 (3). [CrossRef]
  73. Keller, K. L.; Kling, S. M. R.; Fuchs, B.; Pearce, A. L.; Reigh, N. A.; Masterson, T.; Hickok, K. A Biopsychosocial Model of Sex Differences in Children’s Eating Behaviors. Nutrients 2019, 11 (3). [CrossRef]
  74. Adams, J.; Goffe, L.; Brown, T.; Lake, A. A.; Summerbell, C.; White, M.; Wrieden, W.; Adamson, A. J. Frequency and Socio-Demographic Correlates of Eating Meals out and Take-Away Meals at Home: Cross-Sectional Analysis of the UK National Diet and Nutrition Survey, Waves 1-4 (2008-12). Int. J. Behav. Nutr. Phys. Act. 2015, 12, 51. [CrossRef]
Preprints 113910 g001
Figure 1. PRISMA Flow diagram for meal patterns, overweight/obesity risk and metabolically unhealthy obesity risk. OV/OB: overweight/obesity; MUO: metabolically unhealthy obesity.
Figure 1. PRISMA Flow diagram for meal patterns, overweight/obesity risk and metabolically unhealthy obesity risk. OV/OB: overweight/obesity; MUO: metabolically unhealthy obesity.
Preprints 113910 g001
Table 1. Population, Intervention/Exposure, Control, Outcome criteria.
Table 1. Population, Intervention/Exposure, Control, Outcome criteria.
Variable Definition
Population Children/ adolescents: 2-19 years old
Exposure/Intervention High / Low eating / meal and snack frequency; early / late timing of Breakfast, lunch or dinner; low / high levels of omitting of a meal; types of food eaten in different social contexts (alone v. with others); types of food consumed while watching TV v. other activities; meals consumed at home v. out of the home; high / low meal composition/quality
Interventions promoting increased meal frequency; early timing; low levels of omitting meals; eating food with others; food consumed without activities; meals consumed at home; eating high quality meals
Comparator Low / High eating / meal and snack frequency; late / early timing of breakfast, lunch or dinner; high / low levels of omitting of a meal; types of food eaten in different social contexts (alone v. with others); types of food consumed while watching TV v. other activities; meals consumed at home v. out of the home; low / high meal composition/quality
No intervention; intervention of a different meal pattern; standard care
Outcome
  • OV/OB risk
  • MUO risk
Table 2. Risk of bias assessment for the longitudinal studies reviewed regarding the meal patterns factors associated with overweight/obesity risk.
Table 2. Risk of bias assessment for the longitudinal studies reviewed regarding the meal patterns factors associated with overweight/obesity risk.
Study (Author, year) Risk of bias for Longitudinal studies
D1 D2 D3 D4 D5 D6 D7 Overall
Anderson, 2017 [40] - ++ - - - - + ++
Bel-Serrat, 2018 [47] + ++ - - ++ - - ++
Berge 2023 [28] - +++ - - + + - +++
Chang and Gable, 2013 [29] + + - - - - - +
Chang & Halgunseth, 2015 [30] + +++ - - + - - +++
de la Rie, 2023 [54] - + - - ++ - + ++
Balvin Frantzen, 2013 [31] ++ - - - ++ - - ++
Gingras, 2018 [20] + + - - - - - +
Goetz, 2022 [32] + - - - - - - +
Gopinath, 2016 [52] - + - - - - - +
Jaeger, 2022 [50] ++ ++ - - + - - ++
Juton, 2023 [48] + - - + ++ + - ++
Kelly, 2016 [41] ++ + - - + + + ++
Kesztyüs, 2016 [46] + ++ - + ++ ++ - ++
Liechty & Lee 2015 [33] + - - - + ++ - ++
Lipsky, 2015 [34] - - - - + ++ - ++
Loren, 2022 [35] + +++ - - - - - +++
Mahmood, 2023 [51] + + - + ++ ++ - ++
Narla & Rehkopf, 2018 [36] + ++ - - - + ++ ++
Parkes, 2020 [42] - + - - + ++ - ++
Stea, 2014 [49] - - - - +++ + - +++
Sudharsanan, 2016 [37] - ++ - - ++ ++ - ++
Taylor, 2017 [53] + ++ - - + - - ++
Traub, 2018 [45] + ++ - + + ++ - ++
Wang, 2017 [38] + ++ - - + - - ++
Wijtzes, 2016 [43] - ++ - - + + ++ ++
Winter, 2016 [44] ++ +++ - - + + - +++
Di=domain of risk of bias, i=1-7 as follows: D1: due to confounding; D2: arising from measurement of the exposure; D3: in selection of participants into the study (or into the analysis); D4: due to post-exposure interventions; D5: due to missing data; D6: arising from measurement of the outcome; D7: in selection of the reported result. Colour/symbol coding of risk of bias: - low risk of bias; + some concerns; ++ high risk; +++ very high risk.
Table 3. Risk of bias assessment for the randomised controlled trial reviewed regarding the meal patterns factors associated with overweight/obesity risk.
Table 3. Risk of bias assessment for the randomised controlled trial reviewed regarding the meal patterns factors associated with overweight/obesity risk.
Study (Author, year) Risk of bias for randomized controlled trial
D1 D2 D3 D4 D5 Overall
Polonsky, 2019 [39] - + - - - +
Di=domain of risk of bias, i=1-5 as follows: D1: arising from the randomization process; D2: due to deviations from intended interventions; D3: due to missing data; D4: in measurement of the outcome; D5: in selection of the reported result; Colour/symbol coding of risk of bias: - low risk of bias; + some concerns .
Table 4. Characteristics of prospective/longitudinal epidemiological studies exploring the association between meal patterns and the risk of overweight/obesity.
Table 4. Characteristics of prospective/longitudinal epidemiological studies exploring the association between meal patterns and the risk of overweight/obesity.
Study Setting Population FU Exposures Outcomes Statistical analysis Results
Author, year, country Title (acronym) N Age a Sex b
Anderson, 2017, UK [40] The UK Millennium Cohort Study (MCS) Child Benefit register maintained by the Department of Social Security 1099515382 at BL; 11592 at FU; 10995 analysed 36.8 mo (36.3-37.7) M: 5557 (50.3%)F: 5438 (49.7%) 8 y Regular timing of meals (“Always” was coded as having a regular mealtime routine; “Never or almost never”, “Sometimes” and “Usually” were coded as inconsistent mealtime routine)

Assessment: computer-assisted personal interview (caregivers)		 FU BMI z-score (IOTF)FU obesity (BMI z-score at or above 98.9th centile) Statistical model: Logistic regression [OR (95%CI)]Covariates: birth weight, household income, household size, parental age at the time of child’s birth, ethnicity, parental academic and vocational qualifications, country (England, Wales, Scotland, Northern Ireland), bedtime routine, TV/video time routine, self-regulation Decreased odds for developing obesity at age 11 when usually having meals at regular times compared to always having regular meals [(0.77 (0.62-0.97)]; when sometimes, almost never or never have regular meals vs. always having regular meals 0.62 (0.41-0.94)
Bel-Serrat, 2018, Republic of Ireland [47] The WHO European ChildhoodObesity Surveillance Initiative (COSI) school setting 2755163 schools at BL (2008); 81% in 2010; 81% in 2012; 72% in 2015 7.9±1.1 y
(6.0-10.0)		 F: 53.7% 3 y Consumption of breakfast, fast foods and savoury snacks (‘never/< once a week’, ‘some days (1–3 days)’, ‘most days (4–6 days)’, ‘every day’)

Assessment: self-reported qualitative food frequency questionnaire (parents)		 ΔBMI z-score (IOTF)OV/OB incidence (new cases) and prevalence (total number) Statistical model: Multivariate mixed-effects logistic regression models [OR (95%CI)]Covariates: measurement round, time to follow-up, age at baseline, sex, z-BMI at baseline, abdominal obesity at baseline, school SES, and school urbanisation level Frequency of eating breakfast and fast foods was not associated with the OV/OB incidence or prevalence or ΔBMIz at FU. Low frequency of eating savoury snacks at BL was associated with decreased OV/OB prevalence at FU [some days Vs every day 0.48 (0.23-0.99); never Vs every day 0.27 (0.10-0.72)]; decreased OV/OB risk at FU [some days Vs every day 0.49 (0.24-1.00); never Vs every day 0.22 (0.07-0.69); decreased ΔBMI z-score [mean change (SD) never Vs every day -0.18 (0.51), p-trend<0.001]
Berge, 2023, USA [28] The Family Matters study Primary care clinics 1259

1307 at BL		 5-9 y NS 18 mo Quantity: Consumption of family meals (never, 1-2 times, 3-4 times, 5-6 times, 7 times, >7 times). A continuous variable was created using the midpoint of each response range (0, 1.5, 3.5, 5.5, 7) and 8 for the highest category
Family meal interpersonal quality (whether during a family meal people a. watch TV; b. have conversations; c. play video games; d. use tablets/computers; e. read a book; f. listen to headphones. A positive interpersonal quality represented conversations without media distraction; a negative interpersonal quality represented no conversations and media distractions

Assessment: self-reported questionnaire (parents)		 BMI% (CDC) Linear regression and modified Poisson regression (prevalence ration/PR 95% CI)
Covariates: household race/ethnicity, parent
age, parent gender, and parent educational attainment; models for child and family-level outcomes were additionally adjusted for child age and gender.		 Greater weekly family meal quantity at baseline was associated with reduced obesity prevalence at follow-up, with each additional family meal significantly reducing obesity prevalence by 4% [0.96 (0.93-0.99)]

Interpersonal quality was not associated with obesity prevalence [0.99 (0.84-1.17)]
Chang and Gable, 2013, USA [29] The Early Childhood Longitudinal Study-Kindergarten(ECLS-K) school setting 6,220 11.2 y ± 4.3 mo
(10.3-12.8 y)		 M: 49%
F: 51%		 3 y (5th to 8th grade) Breakfast at home (times/wk)School-provided foods for Lunch (Yes/No)Family dinner (times/wk)

Assessment: self-reported questionnaire (parents)		 FU BMI (CDC)Weight trajectory groups: (1) stable obese (Obe-Sta); (2) obese to overweight (ObePos1); (3) obese to healthy (ObePos2); (4) stable overweight (OverSta); (5) overweight to healthy (OverPos); (6) overweight to obese (OverNeg); (7) stable healthy (HelSta); (8) healthy to overweight (HelNeg1); and (9) healthy to obese (HelNeg2). Statistical model: Multivariate logistic regression [OR (95%CI)]Covariates: parental health status, child's health status, child gender and race, age in months at fifth grade, highest level of parent education in the household, family structure, and household poverty level Children who at the 5th grade ate school-provided lunches less frequently were more likely to be in ObePos when compared with ObeSta in the 8th grade [1.10 (1.01-1.19)]. Also, children who ate breakfast more frequently at home were more likely to be in OverPos, when compared with OverSta in 8th grade [1.02 (1.00-1.03)].No other associations between meals and weight trajectories were observed.
Chang and Halgunseth, 2015, USA [30] The Early Childhood Longitudinal Study-Kindergarten Cohort(ECLS-K) school setting 6,860 11 y M: 49%
F: 51%		 3 y (5th to 8th grade) Family meal frequency (sum of frequency of breakfast and evening meals) (times/wk)School-provided foods for Lunch ("A full meal including salad, soup, a sandwich or a hot meal that is offered each day at a fixed price: Yes/No)

Assessment: self-reported questionnaire (parents)		 FU BMI (CDC)Weight status trajectories: (1) stable healthy, (2) stable overweight, (3) healthy change, and (4) unhealthy change. Statistical model: Multivariate logistic regression [OR (95%CI)]Covariates: Parental health status, child's health status, child gender, ethnicity and acculturation, age in months at fifth grade, and highest level of parent education in the household, family structure, and household poverty level There was no association of the frequency of family meals or purchase of school lunches in 5th grade with weight status trajectories in eighth grade (p>0.05 for all tests).
de la Rie, 2023, Germany, the Netherlands, UK, USA [54] Development of Inequalities in Child EducationalAchievement (DICE) (overall)
Data from 4 independent studies		 Day-care facilities / school setting (Germany)Community / general population (the Netherlands*)NS (UK)School setting (USA) 1,275 (Germany)4,007 (the Netherlands)11,285 (UK)6,740 (USA)Sample at BL: 2,349 (Germany)9,749 (the Netherlands)18,552 (UK)18,170 (USA) 5.2±0.4 y (Germany)6.1±0.4 y (the Netherlands)7.2±0.2 y (UK)7.1±0.4 y (USA) F: 51.5% (Germany)F: 49.2% (the Netherlands)F: 48.9% (UK)F: 48.6% (USA) 3-4 y Breakfast consumption (less than 7 d/wk (5 d for US) and 7 d/wk (5 d for US)

Assessment: self-reported questionnaire (parents)		 ΔBMI (WHO 2007 growth standards) Statistical model: regression analysis [Regression coefficient (SE)]Covariates: parental education, child sex, child age in months (at baseline and follow-up), mother foreign-born, maternal age at the birth of the child, single parent household indicator, BMI at baseline, physical activity, screen time Breakfast consumption significantly predicted BMI in the Netherlands [0.26 (0.14)] and the UK 0.40 [(0.13)], but not the USA [0.06 (0.16)], indicating that children in the Netherlands and the UK who ate breakfast daily had lower BMI than children who did not eat breakfast every day. No data for Germany.
Balvin Frantzen, 2013, USA [31] Bienestar: A School-Based Type 2 DiabetesPrevention Program school setting 625

1024 assigned; 706 provided consent; 625 analysed		 9.1±0.5 y M: 309 (49%)
F: 316 (51%)		 2 years (2001-2022 to 2003-2004) Breakfast consumption (Ready to eat cereal (RTEC)) (0=no RTEC breakfast, 1=1 d of RTEC breakfast, 2=2 d of RTEC breakfast and 3=3 days of RTEC breakfast.)Definition: breakfast was considered the first meal of the morning consisting of any solid food, beverages, or both and named by the respondent as “breakfast.”


Assessment: Dietary recall (children)		 ΔBMI percentile (CDC) Statistical model: multivariate linear model analysisCovariates: sex, ethnicity, age, energy, total carbohydrates, and total fat Frequency of RTEC consumption significantly (P=0.001) affected a child’s BMI (R2 change 0.031) with a decrease of 2 percentiles [mean (SD) -1.977 (0.209) for every day of RTEC consumption.No information regarding other types of breakfast or no breakfast consumption.
Gingras, 2018, USA [20] Project Viva Clinics from Atrius Harvard Vanguard Medical Associates (mothers) 9951244 (BL); 1038 (FU); 995 analysed 3.2 y M:504
F: 491		 7 years (up to the age of 11; age 10 for breakfast consumption) Frequency of eating breakfast, eating dinner together with family (“always/daily” versus “<= six times per week”; eating fast food, eating meals while watching television “less than once per week (between zero and three times per month)” versus “>= once per week”.

Assessment: self-reported questionnaire (parents from ages four to eight years and children from ages nine to eleven years)		 FU BMI z-score (U.S. national reference data)FU WC
FU Whole-BF%,
FU trunk fat mass,
FU trunk to peripheral fat mass ratio (BIA, DXA)		 Statistical model: multivariate linear mixed effect models [β (95%CI)]Covariates: mothers' age, education level, parity, marital status, household income, height and pre-pregnancy BMI (kg/m2); child’s sex, race/ethnicity Eating breakfast daily was associated in both boys and girls with lower BMI-z [boys -0.13 (-0.24, -0.02); girls -0.13 (-0.23, -0.02)] and DXA BF% [boys -1.43 (-2.42, -0.45); girls -1.47 (-2.25, -0.68)], and in girls only with lower WC [-1.59 (-2.67, -0.51)], BI BF% [-1.47 (-2.39, -0.54)], DXA trunk fat mass [-0.92 (-1.33, -0.51)] and trunk to peripheral fat ratio [-0.05 (-0.06, -0.03)].Daily family dinner was associated in girls only with lower BMI-z [-0.17 (-0.24. -0.11)], WC [-1.14 (-1.80, -0.48)], BI BF% [girls -1.34 (-1.91, -0.77)], DXA trunk fat mass [-0.32 (-0.57, -0.06) and trunk to peripheral fat ratio [-0.02 (-0.03, -0.01)].Eating meals while watching television less than once per week was associated in boys with lower BMI-z [-0.13 (-0.20, -0.05] , WC [-1.55 (-2.39, -0.71)], BI BF% [ -1.33 (-1.98, -0.69)], DXA BF% [-1.10 (-1.77, -0.44)], DXA trunk fat mass [-0.56 (-0.88, -0.23)] and trunk to peripheral fat ratio [-0.02 (-0.03, -0.01)]. In girls, eating meals while watching television less than once per week throughout childhood was associated with lower BI BF% [-0.74 (-1.35, -0.14)].Eating fast foods less than once a week was associated in girls with lower BMI-z [-0.09 (-0.17, -0.02)], WC [-1.23 (-1.99, -0.48)], DXA BF% [-0.89 (-1.45, -0.33)], DXA trunk fat mass [-0.60 (-0.90, -0.31)] and trunk to peripheral fat ratio [-0.03 (-0.04, -0.02)]
Goetz, 2022, USA [32] Preschool Study Local clinics and preschool centres 116118 BL; 116 FU 4.6±0.9 y M: 50% 1 y First and last eating eventsDefinition: First meal in the morning (06:00 to <10:00) and at night (19:00 to <06:00)

Assessment: Dietary recall (parents)		 FM, FFM, %BF (DXA) Statistical model: Multiple regression model [effect estimate (95%CI)]Covariates: age, sex, race and ethnicity, childcare attendance and income-to-needs ratio, BMIz, BL FM, BL %BF Time of first time eating at baseline was not associated with fat mass at 1 year [-0.01 (-0.14, 0.11)] or %BF [-0.1 (-0.50, 0.27)]A later time of last eating event at baseline was associated with increased FM at 1 year [0.17 (0.02, 0.33)] and % BF [0.83 (0.24, 1.42)]
Gopinath, 2016, Australia [52] The Sydney Childhood Eye Study school setting 6992353 BL; 1216 FU; 699 analysed 12.7 y M: n=319
F: n=380		 5 y (2004-2005 to 2009-2011) Frequency of takeaway food (Chinese, fish and chips, hamburger, and chips/fries, pizza) consumption ("less than once per week", "once per week or more")

Assessment: self-reported semi-quantitative FFQ (children/adolescents)		 FU BMI, OV/OB categories (IOTF)FU BF% (BIA)FU WC Statistical model: Logistic regression [OR (95%CI)]Covariates: ethnicity of the child, country of birth, education, occupation and parental age of both parents, physical activity, screen time 12-year-olds who ate takeaway foods once per week or more compared with those who ate takeaway foods infrequently did not have significantly higher BMI, WC or BF% at 17 years (p>0.05) and did not have significantly higher odds of OV/OB at 17 years, [0.99 (0.59, 1.66) and 1.59 (0.86, 2.94), respectively].
Jaeger, 2022, Belgium, Germany, Italy, Poland, and Spain [50] Childhood Obesity Project (CHOP) Trial NS 729 3 y F: 53% 5 y Eating occasion (breakfast, lunch, and supper for meals; morning, afternoon, and evening snacks)Definition: An EO is defined as any occasion where food or beverages are consumed

Assessment: Dietary records (parents)		 ΔBMI z-score (WHO reference guidelines, IOTF) Statistical model: Regression analysis [β (SE)], compositional data analysis (for meal timing)Covariates: parental BMI, country, TEI, misreporting and an interaction term between TEI and country The redistribution of energy intake with an increase in energy at breakfast, lunch, supper, or snacks as compared to the other EOs was not significantly associated with zBMI [p > 0.05].
Juton, 2023 Spain [48] The POIBC Study (Spanish acronym for the Prevention of Childhood Obesity: A Community-Based Model) School setting 1400

2249 recruited; 1400 analysed		 10.1±0.6 y M: 692 (49.4%)
F: 708 (50.6%)		 15 mo Meal frequency (3 categories: 5 meals/d, 4 meals/d and <4 meals/d; meals assessed: breakfast, mid-morning snack, lunch, afternoon snack, dinner)

Assessment: questionnaire (researcher/children)		 FU BMI z-scoreFU odds of OV/OB (IOTF)
FU WHtRFU odds of AO (WHtR ≥0.50)		 Statistical model: general linear models (mean and 95%CI for FU zBMI and FU WHtR), logistic regression analysis (OR and 95%CI for incidence of OV/OB or AO)Covariates: sex, age, school, intervention group, maternal education, physical activity, adherence to the Mediterranean diet, baseline zBMI/WHtR Higher BL meal frequency was associated with lower FU zBMI increase [0.78 (0.61-0.95) for <4 meals/d; 0.67 (0.61-0.73) for 4 meals/d; 0.62 (0.58-0.66) for 5 meals/d] and lower FU WHtR [0.471 (0.467-0.475) for <4 meals/d; 0.465 (0.463-0.468) for 4 meals/d; 0.463 (0.461-0.465) for 5 meals/d].The FU odds of OV/OB or AO decreased with increase in meal frequency (P for linear trend = 0.035 and 0.028, respectively).
Kelly, 2016, UK [41] UK Millennium Cohort Study NS 16936 (with data in at least one sweep)

9523 (56.2%) with data in all 4 sweeps, 3810 (22.5%) in any 3, 2024 (12.0%) in any 2, and 1579 (9.3%) in 1 sweep		 3 y F:8,259 (48.8%) 8 y (from ages 3 to 11)
Sweeps at the ages of 3, 5, 7, 11		 Consumption of sugary drinks (cola, milkshakes, fruit juice) between mealsSkipping breakfast

Assessment: self -reported questionnaire (parents)		 BMI trajectories (stable; decreasing; moderate increasing; high increasing) Statistical model: multivariate multinomial logistic regression [OR (95%CI)]Covariates: sociodemographic characteristics Sugary drink consumption was not a predictor of BMI trajectory.Skipping breakfast in early childhood was associated with higher odds of increasing BMI trajectory [1.66 (1.37-2.02) of moderately increasing and 1.76 (1.26-2.56) of highly increasing compared to stable] but also higher odds of decreasing trajectory [2.01 (1.03-3.92)] compared to stable.
Kesztyüs, 2016, Germany [46] Join the Healthy Boat school setting 1733 (1212 for the result of interest) 7.1±0.6 y M:881 (50.8%)
F:852 (49.2%)		 1 y Breakfast frequency before school (never/rarely; often/always)

Assessment: self-reported questionnaire (parents)		 WHtR Statistical model: linear regression model [B(SE)]Covariates: school clustering Skipping breakfast was not a significant predictor of changes in WHtR [0.36 (0.19)]
Liechty and Lee 2015, USA [33] National Longitudinal Study of Adolescent to Adult Health (Add Health) school setting 1356820,745 (BL); 13568 analysed 15.8±1.6 y M: 6,605 (48.7%)
F: 6,963 (51.3%)		 1 y Breakfast skipping (yes/no)

Assessment: self-reported questionnaire (children)		 OV/OB onset (change from UW or HW to OV between BL and FU was coded as OV onset. Change from UW, HW or OV to OB was coded as OB onset) [BMI z-score (CDC)] Statistical model: Multinomial regression analysis [RR (95%CI)]Covariates: age, race/ethnicity, parent education and family structure, BMIz score Skipping breakfast increased OV onset risk among female adolescents [1.44 (1.00-2.07)] but not male adolescents [0.78 (0.49-1.24)]. Skipping breakfast was not associated with OB onset.
Lipsky, 2015, USA [34] NEXT Generation Health Study school setting 2,785 (78% retention rate at wave 4) 16.3±0.03 y M: 45.5%
F: 54.5%		 3 years (wave 1 BL; a yearly wave until wave 4) Frequency of breakfast consumption (d/wk); family meals (evening) (d/wk); watching TV during meals (d/wk); eating fast foods (d/wk); sweet and salty snacks (times/d)

Assessment: self-reported questionnaire (children)		 Prospective 1-year BMI change (next year BMI – current BMI) for waves1 through 3 Retrospective 1-year BMI change (current BMI – previous year BMI) for waves 2 through 4] Statistical model: linear GEE (generalised estimating equations) models [βest (SE)]Covariates: sex, race/ethnicity, Family Affluence Scale (car and computer ownership, family vacations, bedroom sharing), parental educational status, physical activity, time-varying height There was an inverse association of consumption of sweet and salty snacks with time-varying BMI [-0.33 (0.12), p=0.02]. None of the meal practices (breakfast, family meals, watching TV during meals and fast food) was associated with BMI longitudinally. None of the meal practices were associated with prospective or retrospective 1-year BMI change.
Loren, 2022, USA [35] The Early Childhood Longitudinal Study-Kindergarten Cohort(ECLS-K:2011) school setting 8,225 Age ‘kindergarden’ M: 51% 3 y (second grade y) Family meal frequency (morning and/or evening meal; d/wk)

Assessment: Interview (parents)		 FU BMI z-score (CDC) Statistical model: structural equation modeling (SEM); path models (standardised coefficient γ)
Covariates: race/ethnicity, income-to-needs, sex		 Number of family meals per week at BL was not associated with BMI z-score at FU (0.02, p>0.05)
Mahmood, 2023, Greece, Spain, Bulgaria, Hungary, Belgium, Finland [51] European Feel4Diabetes study school setting 9892748 at BL; 989 analytic sample M: 7.3 ± 0.99 y
F: 7.4 ± 1.02 y		 F: 52% 1 year (end of intervention; T1)2 years (end of FU; T2) Frequency of family meals (breakfast, lunch, dinner)(1) three categories (never, 1-2 times/wk, 3-7 times/wk)(2) five categories (never, remained low, decreased, increased, remained high)

Assessment: self-reported questionnaires (parents)		 ΔBMI = BMI T2 - BMI at BL
BMI categories change (normal weight at BL and T2; OV/OB at BL but normal weight at T2, normal weight at BL but OV/OB at T2, OV/OB at BL and T2 [BMI z-score (IOTF)]		 Statistical model: multivariable regression model (β); multilevel logistic regression [OR (95%CI)]Covariates: country, group (intervention-control), age and DQ of children, and parental characteristics (age, marital status, educational level, employment, sex, DQ, BMI), family meals frequency and BMI of children at baseline Increase in family breakfasts frequency over time was negatively associated with ΔBMI in girls (β=- 0.078, p = 0.035) but not boys (β=-0.051, p = 0.066). Increase in family dinners frequency was inversely associated with ΔBMI of boys (β=-0.102, p = 0.019) and girls (β=-0.198, p < 0.001). Boys and girls whose family breakfasts frequency increased were more likely to have lower BMI (boys: 0.68; 0.49–0.91; girls: 0.69; 0.34–0.92) than those with a decreased frequency of family breakfasts. A similar association was found between a change in family dinner frequency (boys: 0.57; 0.39–0.83); girls (0.69; 0.42–0.91). The odds of FU OV/OB were decreased for boys (0.76; 0.52, 1.04) and girls (0.72, 0.58, 0.93) who consumed family breakfasts 3-7 times a week at BL, compared to those who never had breakfasts with family. Having ≥3 family-shared dinners/wk at BL was associated with reduced odds of OV/OB at T2 in boys (0.65; 0.41, 0.96) and girls (0.53; 0.31-0.87) compared with those who never shared family dinners during childhood. Increased family breakfasts frequency over time was associated with lower odds of OV/OB in boys (0.78; 0.52–1.11) and girls (0.78; 0.55–1.01) compared to never having breakfast. Improved family-shared dinners over time showed lower odds of OV/OB at T2 (boys: 0.54; 0.33–0.83; girls: 0.61; 0.40-0.97).No associations were observed for family lunch meals with any outcome.
Narla and Rehkopf, 2019, USA [36] The NHLBI Growth and Health Study (NGHS) general population 2024

2379 at BL; 2024 analysed		 10.0 y F: 2024 (100%) 10 y Exposures assessed at y 1 (BL), 2 and 3

Eating breakfast, morning snack, afternoon snack, evening snack, eating fast food (times/wk)
Eating while watching TV, eating with family, eating with homework, eating school lunch, eating alone, skipping lunch, eating with friends, eating in bedroom

Assessment: Food diary		 BF% (skinfolds) Statistical models: multiple linear regression (coefficient of association), random forest analysis and propensity score matching (PSM) [difference score (95%CI)]

Covariates (multiple linear regression): participant’s age in months, household income, race, highest level of parental education		 Significant adiposity predictors: eating alone (y3: 3.94), skipping lunch (y1: 6.58), (y2: 6.08), (y3: 6.84)
Significant protective predictors of adiposity: eating breakfast (y1: -2.14), eating afternoon snack (y1: -4.64), eating evening snack (y1: -2.23), eating while watching TV (y1: -3.41), (y2: -4.24), (y3: -4.03), eating with family (y1: -4.44), (y3: -4.62), eating while doing homework (y2: -5.34), (y3: -5.08), eating with friends (y2: -4.72), (y3: -3.40), eating in bedroom (y2: -3.41)

PSM showed one detrimental for adiposity risk factor [skipping lunch at y2: 4.0 (1.1, 6.7); y3: 4.3 (1.6, 7.2) and five protective factors against adiposity [eating evening snack at y1: -3.1 (-5.9, -0.3), eating with friends at y2: -4.4 (-7.6,-1.4); y3: -6.0 (-10.0,-2.2), eating while watching TV at y2 -5.8 (-10.1, -1.5); y3 -5.3 (-8.5, -2.2), eating while doing homework at y1 -5.7 (-11.1, -0.1); y2 -6.2 (-9.1, -3.5), eating in bedroom at y2 -5.8 (-10.2, -1.2)
Parkes, 2020, UK [42] Growing Up inScotland study general population 28105217 at BL; 2810 analysed 46 mo M: 1432
F: 1378		 76 mo (FU at 58, 70, 94 and 122 mo) Mealtime setting (Factor score of three items: main meal eaten in a “dining” area (= kitchen, dining room, combined living/dining room) or non-dining” area (living room, bedroom, other) (58 and 122 months); mealtime screen use (TV only at 58 months, TV and other screens at 122 months); how often the child sat at a table while eating a main meal (122 months);
Categories: “formal”, “intermediate” or “informal”, where “informal” indicates greater screen use and less use of dining area.

Assessment: self-reported questionnaire (parents)		 BMI trajectories: Low Risk, Decreasing Overweight, Increasing Overweight, High/stable Overweight, High/Increasing obesity (IOTF) Statistical model: GrOVth mixture models (GMM); multinomial regression models [RRR (95%CI)]Covariates: early life factors (child sex, ethnic group, family socio-economic disadvantage, maternal BMI, child birth order, maternal smoking in pregnancy, maternal mental health, infant feeding), early diet patterns (healthy diet, picky diet), household organisation and routines (home organisation, irregular bedtimes, skipping breakfast), child behaviours at school-age (overall screen time, physical activity and sleep) Informal settings were associated with the FU High/Increasing Obesity and Increasing Overweight trajectories [3.67 (1.99-6.77), p<0.001); 1.75 (1.17-2.62, p=0.007 respectively).Intermediate settings were associated with the High/Increasing Obesity and Increasing Overweight trajectories [1.89 (1.09-3.28), p=0.023); 1.50 (1.03-2.19, p=0.036 respectively).
Stea, 2014, Norway [49] NS school setting 4281045 (BL; 4th grade); 1095 (FU; more schools included); 428 with complete data at BL and FU 9-10 y (4th grade) M: 207
F: 221		 3 y (grades 4 to 7) Meal frequency: CONTINUED skippers (skipping meals at both time points); (ii) START all meals (meals skippers in 4th grade, eat all meals in 7th grade); (iii) STOP all meals (eat all meals in 4th grade, meal skippers in 7th grade); and (iv) ALL meals (eat all meals at both time points).
"All meals" if eating all breakfast/lunch/dinner/evening meals daily; "skipping meals" if eating <7d/week

Assessment: self-reported FFQ (parents) 		ΔBMI categories (normal weight and OV) (IOTF) Statistical model: Logistic regression [OR (95%CI)]Covariates: maternal education, gender, physical activity, overweight status at 4th grade Meal skipping was not associated with odds of OV.
STOP all meals 2.8 (0.7, 11.9);
CONTINUED skippers 1.8 (0.5, 6.2);
START all meals 1.4; (0.3, 7.2)
Sudharsanan, 2016, USA [37] ECLS-K school setting 6,4957,304 at BL and FU; 6,495 analysed NS (5th grade) M: 50.3% 3 y (5th-8th grade) Eating school breakfast (yes/no)

Assessment: interview (parents)		 FU obesity status (CDC) (obesity / non-obesity) Statistical model: multivariate models, propensity score matching and fixed-effects logistic regression [OR (95%CI)]Covariates: sex; race/ethnicity, age, physical activity, family socioeconomic status, family marital status, mother's employment, number of breakfasts and dinners the family ate together in a typical week, school type, urbanicity Eating breakfast at school was not associated with obesity at FU (1.31; 0.82-1.97, P=0.129). For children from families below the federal poverty line, eating school breakfast increased the odds of obesity at FU (2.31; 1.25-4.28, P=0.010) compared with children of similar SES who did not receive school breakfast (propensity score matching).A change in school breakfast (from yes to no) between the 5th-8th grade was not statistically associated with a change in weight status longitudinally (0.72; 0.39-1.31, P=0.285) (logistic fixed-effects regression)
Taylor, 2017, New Zealand [53] Prevention of Overweight in Infancy (POI) study maternity hospital (mothers) 371802 at BL; 695 at FU; 371 analysed 2 y M: 196 (52.8%)
F: 175 (47.2%)		 1.5 y Meal frequency / eating occasionDefinition: a separate eating occasion, the start of the next meal or snack had to be more than 15 min after the end of the previous meal or snack

Assessment: feeding diary (parents)		 BMI z score (WHO) Statistical model: Linear regression analysisCovariates: household deprivation and income, maternal parity, mother's intervention group, infant sex, birth weight, maternal education and pre-pregnancy BMI, smoking during pregnancy, exclusive breast-feeding Eating frequency at 2 years of age did not predict change in BMI Z-score at FU (difference in BMI Z-score per additional eating occasion 0.02; 95% CI −0.03, 0.06)
Traub, 2018, Germany [45] Join the Healthy Boat school setting 1,733 7.1
± 0.6 y		 M:881 (50.8%)
F:852 (49.2%)		 1 y Breakfast consumption before school (“Never and rarely” / “often and always”)

Assessment: self-reported questionnaire (parents)		 Δweight (kg)
ΔBMI percentile (German reference cut-off points)
WtHRAO (WHtR ≥ 0.5)		 Statistical model: linear mixed regression for changes in WHtR, Wt, BMI% [B (SE)] and OV/OB and AO [OR (95%CI)]Covariates: school, migration background, family education level, household income, age, gender, participation in the intervention Skipping breakfast at BL was positively associated with increases in WHtR [0.50 (0.19)], weight [0.39 (0.12)] and BMI percentile [2.01 (0.90)].Skipping breakfast was associated with increased odds of FU AO (2.06, 1.23-3.47) and OV (1.71, 1.04-2.80) but not OB (0.90, 0.39-2.07)
Wang, 2017, USA [38] N/A school setting 513 (BL; 5th grade); 553 (6th grade); 468 (7th grade)584 (BL), 602 (6th grade), 539 (7th grade); 513 (BL), 553 (6th grade), and 468 7th grade) analysed 5th grade M:236 (46%)
F: 277 (54%)		 2 y (2011-2012 to 2013-2014) Breakfast location patterns [average number of d/wk (0–7) and location where breakfast was eaten]Categories: frequent skippers, inconsistent school eaters, inconsistent home eaters, regular home eaters, regular school eaters, double breakfast eaters


Assessment: self-reported questionnaire (children)		 FU BMI percentile (CDC) (Two categories: OV/OB and normal/ underweight) Statistical model: Generalized estimating equations (GEE) models [AOR (95%CI)]Covariates: sex, race/ethnicity, school and study year Odds of FU OV/OB was higher for students in the skippers group (2.66; 1.67, 4.24), inconsistent school eaters (2.11; 1.29, 3.46), inconsistent home eaters (2.02; 1.27, 3.21) and regular home eaters (1.70; 1.13, 2.56) compared with double breakfast eaters.There was no evidence of greater weight gain over time among students who consume a double breakfast when compared with all other students.
Wijtzes, 2016, the Netherlands [43] Generation R Study maternity hospital (mothers) 59138305 at BL; 6,690 at FU; 5,913 analysed 4 y M: 2,939 (49.7)
F: 2,974 (50.3)		 2 y Meal skipping (breakfast, lunch, dinner; consumption <7 d/wk)Tracking patterns: stable consumption (consumption at both time points), stable meal skipping (skipping at both time points), decrease in meal skipping (skipping at 4 y and consumption at 6 y); increase in meal skipping (consumption at 4 y and skipping at 6 y).

Assessment: self-reported questionnaire (parents)		 FU BMI SDS (IOTF)FU FM% (DXA scan) Statistical model: linear [β (95%CI] and logistic regression modelsCovariates: child’s sex, age, ethnic background, family socioeconomic position (ie, maternal educational level, maternal employment status, household income, maternal and paternal BMI, and children’s physical activity, sedentary behaviours and dietary behaviours, BMI at age 4 Breakfast skipping at 4 y was associated with increased FM% at 6 y (1.38; 0.36-2.40). Continuously measured breakfast skipping at age 4 years was associated with a higher FM% (0.64; 0.41-0.88) and a higher BMI (0.08; 0.02-0.13) at 6 y. Compared with stable breakfast consumers, children in all 3 breakfast-skipping categories had a significantly increased FM% at 6 y [1.80 (0.75-2.85); 1.24 (0.56-1.92); 0.92 (0.11-1.74)]Lunch and dinner skipping at 4 years were not associated with FM at 6 y.

No meal skipping at age 4 was associated with BMI SDS at age 6.
Winter, 2016, the Netherlands [44] Tracking Adolescents' Individual Lives Survey (TRAILS) school setting Wave 1 (BL) 2,230; Wave 2 (T2) 2,149; Wave 3 (T3) 1,816 11.09 ± 0.56 NS T2 29.5 mo after BL; T3 32.0 mo after T2 Breakfast consumption [no regular breakfast (consumed <5 times/wk; regular breakfast)]

Assessment: self-reported questionnaire (adolescents)		 ΔBMI categories (BMI criteria NS) Statistical model: Repeated multivariate logistic regression analysis [OR (95%CI)]
Covariates: gender		 Not having regular breakfast at T2 was not associated with OV/OB at T3 (1.41; 0.97-2.06)
a Presented as mean ± standard deviation or median (1st, 3rd quartile) and/or range, unless otherwise stated. b Presented as absolute (relative) frequency. Abbreviations: AO: abdominal obesity, BF: body fat, BIA: bioelectrical impedance analysis, BL: baseline, BMI: body mass index, CDC: Centers for Disease Control and Prevention, CI: confidence interval, d: day(s), DQ: diet quality, DXA: dual-energy X-ray absorptiometry, F: females, FFM: fat-free mass, FFQ: food frequency questionnaire, FM: fat mass, FMI: fat mass index, FU: follow-up, HW: healthy weight, IOTF: International Obesity Task Force, M: males, mo: months, NS: not stated, OB: obesity, OR: odds ratio, OV: overweight, PA: physical activity, RR: risk ratio, RRR: relative risk ratio, SDS: standard deviation score, SE: standard error, SES: socioeconomic status, TEI: total energy intake, UK: United Kingdom, USA: United States of America, UW: underweight, WC: waist circumference, WHO: World Health Organisation, WHtR: waist-to-height ratio, wk: week, y: year(s), Δ: change.
Table 5. Characteristics of randomised controlled trials exploring the association between meal patterns and the risk of overweight/obesity.
Table 5. Characteristics of randomised controlled trials exploring the association between meal patterns and the risk of overweight/obesity.
Study Country Setting Population FU Interventions Outcomes Statistical analysis Results
Author, year Title (acronym) N Age a Sex b
Polonsky, 2019 [39] One Healthy Breakfast USA school setting

793

1,463 provided written consent; 1362 at BL; 793 at FU		 10.8 ± 1.0 y M:662 (48.6%)F:700 (51.4%) 2.5 y (mid-point data collected at 1.5 y) Design: parallel cluster RCT
Groups: IG vs. CG

IG (350): breakfast in the classroom, 18 45–min nutrition education lessons (importance of breakfast), social marketing materials and corner stores with healthy choices; monthly newsletters to parents (8 schools)CG (443): breakfast offered in the cafeteria before the beginning of the school day and standard education material (8 schools)		 OV/OB incidence and prevalence
ΔBMI-z (CDC)		 Statistical model: weighted generalized estimating equation (GEE) models [OR (95%CI)

Covariates: paired stratification of randomisation		 There was no difference in combined OV/OB incidence between IG (11.7%) and CG (9.3%) at FU (1.31; 0.85-2.02). The OB incidence alone was higher in IG (11.6%) than in CG (4.4%) between BL and FU (2.43; 1.47-4.00). There was no difference between IG and CG in the combined OV/OB prevalence or BMI zscore across the study period. The OB prevalence alone at the end point was higher in IG (28.0%) than in CG (21.2%) at FU (1.46; 1.11-1.92).
a Presented as mean ± standard deviation or median (1st, 3rd quartile) and/or range, unless otherwise stated. b Presented as absolute (relative) frequency. Abbreviations: BL: baseline, BMI: body mass index, CG: control group, CI: confidence interval, F: females, FU: follow-up, IG: intervention group, M: males, OB: obesity, OV: overweight, RCT: randomized clinical trial, y: year(s), z: z-score, Δ: change.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
Prerpints.org logo

Preprints.org is a free preprint server supported by MDPI in Basel, Switzerland.

facebook logotwitter logolinkedin logo
MDPI Initiatives
Important Links
Subscribe

Disclaimer

Terms of Use

Privacy Policy

© 2024 MDPI (Basel, Switzerland) unless otherwise stated