Effectiveness of dietary interventions on weight outcomes in childhood: a systematic review meta-analysis of randomized controlled trials.

BACKGROUND: Rapid growth and elevated weight in childhood are significantly associated with obesity in later life, but evidence regarding dietary interventions and weight outcomes is lacking. This study aimed to determine the effectiveness of dietary interventions on body mass index (BMI) and BMI z-score in childhood.
METHODS: PubMed, EmBase, and the Cochrane library were searched from inception till June 2019. Studies that investigated the effectiveness of dietary interventions on BMI and BMI z-score in childhood were considered eligible in our study. The changes in BMI and BMI z-score between dietary interventions and control were calculated by pooled weighted mean differences (WMDs) and 95% CIs were evaluated using random-effects model.
RESULTS: Twenty-eight randomized controlled trials involving a total of 17,488 children were included. The summary WMDs indicated that children who received dietary interventions had greater reduction in BMI (WMD: -0.12; 95% CI: -0.20 to -0.05; P=0.001) and BMI z-score (WMD: -0.04; 95% CI: -0.06 to -0.02; P=0.001) when compared to the usual controls. Subgroup analyses revealed that the sample size, mean age, duration of interventions, and study quality could affect the effectiveness of dietary interventions in children.
CONCLUSIONS: The findings of this meta-analysis suggested that dietary interventions improved BMI and BMI z-score, whereas these results are limited due to substantial heterogeneity and study quality of the included studies. 2021 Translational Pediatrics. All rights reserved.

Introduction

Childhood obesity is mainly associated with wide range of adverse psychosocial and physical health outcomes. Therefore, an effective intervention strategy should be developed for preventing childhood obesity, as it is significantly correlated with public health (1). According to a previous study, the effectiveness of early intervention in the first few years of life for preventing overweight and obesity is clearly evident, and the prevalence of overweight is nearly 6.7% in children below 5 years of age (2). Although overweight during infancy is significantly correlated with the occurrence of obesity during childhood, adolescents, and adulthood, the occurrence of overweight during infancy might not be diagnosed by the providers (3,4). Previous studies have reported that educated mothers focused on both nutritional as well as environmental aspects, which played an important role in preventing non-communicable diseases (5). However, it is still controversial whether dietary interventions through education affects the weight outcomes in childhood.

Nowadays, numerous factors are associated with the progression of obesity in childhood, which included infant feeding practices, children’s eating habits and television watching time, and these are the most modifiable factors (6-10). Moreover, infant feeding practices showed significant correlation with the eating behaviors of children as well as adults in the later life (11). Therefore, obesity-related behaviors were affected by a range of settings, especially the dietary interventions. Recently, several interventional strategies have been employed for preventing and treating childhood obesity, whereas these intervention strategies provided controversial results with slight reduction in the body mass index (BMI) (12,13). Numerous randomized controlled trials (RCTs), some participants were preschool age children (14-17) and others were school-age children (18-41), have already investigated the role of dietary interventions on BMI and BMI z-score in childhood, whereas the effect estimates varied owing to differences in the intervention theories. Therefore, we attempted to comprehensively examine the published RCTs for determining the effectiveness of dietary interventions on weight outcomes in childhood. Moreover, stratified analyses were conducted to explore whether the effectiveness of interventions differed according to the sample size, mean age, duration of interventions, and study quality.

We present the following article in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) Statement reporting checklist (42) (available at http://dx.doi.org/10.21037/tp-20-183).

Methods

Ethical statement

Institutional Review Board approval was not required because this article is a meta-analysis. The data comes from published articles and does not require ethical approval.

Data sources, search strategy, and selection criteria

Studies that are designed as RCTs and those that evaluated the effectiveness of dietary interventions on BMI or BMI z-score were considered eligible for inclusion in this meta-analysis, and there was no restriction to published language and status. PubMed, EmBase, and the Cochrane library were systematically searched for studies published from inception till June 2019. The core search terms used were as follows: (“child*” OR “infant*”) AND (“health education” OR “school health services” OR “child health services” OR “community health planning” OR “primary health care” OR “child nutrition sciences” OR “child nutrition disorders” OR “food habits” OR “nutrition assessment” OR “diet” OR “diet therapy”). Moreover, the reference lists of the obtained studies were manually searched for inclusion of any new study.

Two independent authors conducted the literature search and study selection processes, and any disagreement was resolved by discussion with each other until a consensus was reached. Studies were included if they met the following inclusion criteria: (I) study design: studies with RCT design; (II) participants: all individuals were less than 18.0 years, irrespective of their weight status; (III) intervention: individuals who received medical health education containing dietary contents; (IV) control: usual health program; and (V) outcomes: studies reporting BMI or BMI z-score.

Data collection and quality assessment

The collected information from the retrieved studies included the first authors’ surname, publication year, country, sample size, mean age, intervention populations, dietary intervention target, control, duration of intervention, and changes in BMI or BMI z-score. The quality assessment was evaluated using the Cochrane Collaboration risk of bias instrument, and each item was answered with yes, no, or unclear (43). Data collection and quality assessment were conducted by 2 authors, and any conflicts were resolved by an additional author by referring to the original article.

Statistical analysis

The effectiveness of dietary interventions on BMI or BMI z-score in childhood was used as continuous variable based on the mean, standard deviation, and sample size in each individual trial. The summary weighted mean differences (WMDs) with its 95% CIs were used to calculate the effect size of dietary interventions by using random-effects model (44,45). I2 and Q statistics were employed to assess the heterogeneity across the included trials, and I2 >50.0% or P<0.10 was regarded as significant heterogeneity (46,47). Sensitivity analyses for BMI and BMI z-score were carried out to assess the impact of each trial from the overall analysis (48). Subgroup analyses for BMI and BMI z-score were conducted based on sample size, mean age, duration of interventions, and study quality to evaluate the effectiveness of dietary interventions according to the study or individuals’ characteristics. After this, the P value between subgroups was compared through interaction test using Student’s t-test as the number of included studies was small (49). Publication biases for BMI and BMI z-score were evaluated using funnel plot, Egger (50), and Begg (51) tests, and P<0.10 was considered as potentially significant publication bias. The inspection level for pooled results was 2-sided, and P<0.05 was regarded as statistically significant. All analyses in this study were conducted using STATA software (Version 10.0; Stata Corporation, College Station, TX, USA).

Results

Literature search

Our initial electronic database search yielded 12,471 records. Of these, 5,953 articles were excluded due to duplications. After that, the title and abstract of the remaining 6,518 studies were reviewed, and 6,379 of these were excluded due to irrelevant topics. A total of 139 articles were selected for further evaluation, and 111 of these were excluded due to the following reasons: the intervention strategy did not contain the dietary content (n=46); other health outcomes were reported (n=42); studies reported on similar population (n=17); and data on BMI or BMI z-score were not available (n=6). The reference lists yielded 87 potential studies, whereas all these studies were obtained from electronic searches, and so are excluded as duplicates. Finally, 28 RCTs were selected for quantitative meta-analysis. The detailed literature search and study selection process are presented in .

Figure 1

Flowchart of study selection process.

Study characteristics

Of the 28 eligible RCTs, 17,488 children were included for final analysis. The baseline characteristics of the included studies and children are summarized in . The duration of dietary interventions in individuals was 2.0–72.0 months, and each study included 106–2,950 children. The age of the children included ranged from 0.75–13.0 years, and most of the included trials were conducted in western countries. Study quality of individual trials was shown in .

Table 1

Baseline characteristic of studies included in the systematic review and meta-analysis

Study	Country	Sample size	Mean age (years)	Intervention populations	Dietary intervention target	Control group	Duration of intervention	Study quality
James 2004, (18)	England	644	8.7	School children	Focused educational programme on reducing the consumption of diet carbonated drinks	No changes in relation to interventions schools	12 months	1
Muckelbauer 2009, (19)	Germany	2,950	8.3	School children	Single school-based intervention provided cooled and optionally carbonated water	Did not receive any intervention	12 months	4
Jansen 2011, (20)	Netherlands	2,622	9.2	Normal, overweight and obese school children	Multi-component intervention main lessons on healthy nutrition, active living and healthy lifestyle choices	Control schools continued with their usual curriculum	9 months	4
Manios 2002, (21)	Crete	1,043	6.0	School children and parents	Multi-component school-based Intervention on health and nutrition	Not assign	72 months	2
Siegrist 2013, (22)	Germany	724	8.4	Normal weight and overweight or obese school children	Multi-component JuvenTUM Intervention on healthy eating	Usual activities	12 months	1
Gentile 2009, (23)	USA	1,323	9.6	School children and parents	Multiple interventions to eat five fruits/vegetables or more per day	Did not receive any materials	7 months	1
Graf 2005, (24)	Germany	651	6.9	School children	Multi-component school-based intervention on nutrition	Usual health program	20.8 months	0
Williamson 2012, (25)	USA	2,060	10.5	Normal weight and overweight or obese school children	Multi-component intervention to promote healthy nutrition	The control group received none of the prevention components that are hypothesized to yield weight gain prevention	28 months	2
Sichieri 2009, (26)	Brazil	1,134	10.9	School children	Single school-based intervention focused on the reduction in consumption of sugar-sweetened carbonated beverages by students	Received two one-hour general sessions on health issues and printed general advises regarding healthy diets	7 months	3
Natale 2014, (14)	USA	307	3.9	Preschool-aged children and their parents	Healthy menu changes and family-based education focused on fresh produce intake, decreased intake of simple carbohydrate snacks	Attention control program	12 months	3
Johnston 2013, (27)	USA	835	7.8	School children	Professional-facilitated intervention on health and nutrition educational materials	self-help control	24 months	4
McFarlin 2013, (28)	USA	221	13.0	School children	School-based intervention focused on diet	self-help control	12 months	1
De Coen 2012, (29)	Belgium	1,102	2.5-6.5	Preschool-aged children and school children	Nutrition and Physical Activity Health Targets of the Flemish Community	Usual health program	24 months	3
Kesztyüs 2013, (30)	Germany	719	7.6	School children	The intervention focused on consumption of sweetened beverages	Usual health program	22 months	1
Llargués 2012, (31)	Spain	426	6.0	School children	Intervention program aimed dietary habits	Usual health program	24 months	2
Kain 2014, (16)	Chile	1,474	6.6	School children	Multi-component intervention focused on nutritional status	Usual health program	15 months	4
Puder 2011, (32)	Europe	652	5.1	Preschool-aged children	The multidimensional culturally tailored lifestyle intervention focused on lessons on nutrition	Usual health program	10 months	5
Grydeland 2014, (36)	Norway	1,324	11.2	School children	The multidimensional culturally tailored lifestyle intervention topromote a healthy diet	Usual health program	20 months	3
Magnusson 2012, (15)	Iceland	321	7.4	School children	The intervention primarily focused on promoting healthy dietary habits, both at school and at home	Usual health program	24 months	2
Larsen 2016, (37)	Denmark	106	12.0	School children	A six-week day-camp intervention focused on healthy diet	Standard intervention arm consisting of one weekly exercise session for six weeks	12 months	5
Davis 2016, (33)	USA	980	<4.0	Preschool-aged children	CHILE intervention to improving dietary intake	Usual health program	19 months	3
Amini 2016, (38)	Iran	334	Median: 10–12	School children	The intervention focused on change in food items sold at the schools’ canteens	Usual health program	4.2 months	2
Taveras 2017, (39)	USA	721	8.0	Preschool-aged children and school children	Enhanced primary care plus contextually-tailored focused on decreases in sugar-sweetened beverages and improving diet quality	Enhanced primary care	12 months	4
Ojeda-Rodríguez 2018, (40)	Spain	107	11.3	School children	Lifestyle intervention on nutrient adequacy and diet quality	Usual health program	2 months	2
Ahmad 2018, (34)	Malaysia	134	9.6	School children	The intervention primarily focused on decreases in sugar-sweetened beverages and improving fruits and vegetables	Wait-list control	6 months	3
Gómez 2018, (35)	Spain	2,086	10.1	School children	TCHP focused on eating habits and cooking techniques	Usual health program	15 months	3
Adab 2018, (41)	UK	1,397	6.3	School children	The intervention focused on healthy eating	Usual health program	12 months	5
Enö Persson 2018, (17)	Sweden	1,091	0.75-4.0	Preschool-aged children	The intervention to promote healthy food habit	Usual healthcare	12 months	1

Table 2

Risk of bias for individual study

Study	Random sequence generation (selection bias)	Allocation concealment (selection bias)	Blinding of participants and personnel (performance bias)	Blinding of outcome assessment (detection bias)	Incomplete outcome data (attrition bias)	Selective reporting (reporting bias)	Other bias
James 2004, (18)	Yes	No	No	No	Unclear	Unclear	No
Muckelbauer 2009, (19)	Yes	Yes	No	Yes	Yes	Yes	Unclear
Jansen 2011, (20)	Yes	Yes	No	Yes	Yes	Yes	Unclear
Manios 2002, (21)	Yes	No	No	Unclear	No	Unclear	No
Siegrist 2013, (22)	Yes	No	No	No	Unclear	Unclear	No
Gentile 2009, (23)	Yes	No	No	No	No	Unclear	Unclear
Graf 2005, (24)	Yes	No	No	No	Unclear	Unclear	No
Williamson 2012, (25)	Yes	No	No	Unclear	No	Unclear	No
Sichieri 2009, (26)	Yes	No	No	Yes	Yes	Unclear	No
Natale 2014, (14)	Yes	No	No	Yes	Yes	Unclear	No
Johnston 2013, (27)	Yes	Yes	No	Yes	Yes	Yes	Unclear
McFarlin 2013, (28)	Yes	No	No	Unclear	No	Unclear	Unclear
De Coen 2012, (29)	Yes	No	No	Yes	Yes	Unclear	No
Kesztyus 2013, (30)	Yes	No	No	No	Unclear	Unclear	No
Llargués 2012, (31)	Yes	No	No	Yes	Unclear	Unclear	No
Kain 2014, (16)	Yes	Yes	No	Yes	Yes	Yes	Unclear
Puder 2011, (32)	Yes	Yes	No	Yes	Yes	Yes	Yes
Grydeland 2014, (36)	Yes	No	No	Yes	Unclear	Unclear	No
Magnusson 2012, (15)	Yes	No	No	Yes	Unclear	Unclear	No
Larsen 2016, (37)	Yes	Yes	No	Yes	Yes	Yes	Yes
Davis 2016, (33)	Yes	No	No	Yes	Unclear	Unclear	No
Amini 2016, (38)	Yes	No	No	Yes	No	Unclear	No
Taveras 2017, (39)	Yes	Yes	No	Yes	Yes	Yes	Unclear
Ojeda-Rodríguez 2018, (40)	Yes	No	No	Yes	No	Unclear	No
Ahmad 2018, (34)	Yes	No	No	Yes	Unclear	Unclear	No
Gómez 2018, (35)	Yes	No	No	Yes	Unclear	Unclear	No
Adab 2018, (41)	Yes	Yes	No	Yes	Yes	Yes	Yes
Enö Persson 2018, (17)	Yes	No	No	No	Unclear	Unclear	No

BMI

Data regarding the effectiveness of dietary intervention on BMI in childhood were available in 20 trials (22 cohorts). The results revealed that children who received dietary interventions had greater reduction in BMI when compared with usual healthcare (WMD: –0.12; 95% CI: –0.20 to –0.05; P=0.001; ), and showed a significant heterogeneity among the included trials (I2=99.9%; P<0.001). The results of sensitivity analysis indicated that the conclusion was stable and unaltered by excluding any particular trial (). Subgroup analyses indicated that dietary intervention significantly reduced BMI when compared with healthcare when the sample size <1,000 (WMD: –0.18; 95% CI: –0.37 to –0.00; P=0.047), mean age of individuals ≤6.0 (WMD: –0.25; 95% CI: –0.43 to –0.09; P=0.002) or >10.0 years (WMD: –0.21; 95% CI: –0.30 to –0.11; P<0.001), the duration of intervention ≥12.0 months (WMD: –0.18; 95% CI: –0.27 to –0.09; P<0.001), and studies with high quality (WMD: –0.25; 95% CI: –0.37 to –0.13; P<0.001). Moreover, significant effectiveness of dietary interventions on BMI could be affected by sample size, mean age, duration of intervention, and study quality ().

Figure 2

Effect of dietary intervention on body mass index.

Figure 3

Sensitivity analyses for body mass index (A) and body mass index z-score (B).

Table 3

Subgroup analyses for BMI and BMI z-score

Outcomes	Factors	Subgroup	Number of cohorts	WMD and 95% CI	P value	Heterogeneity	P value between subgroups
BMI	Sample size	≥1,000	10	–0.06 (–0.12 to 0.00)	0.059	99.8 (<0.001)	<0.001
		<1,000	12	–0.18 (–0.37 to –0.00)	0.047	99.9 (<0.001)
	Mean age (years)	≤6.0	5	–0.25 (–0.43 to –0.09)	0.002	99.9 (<0.001)	<0.001
		6.0–10.0	12	–0.02 (–0.15 to 0.10)	0.702	99.9 (<0.001)
		>10.0	5	–0.21 (–0.30 to –0.11)	<0.001	99.6 (<0.001)
	Duration of intervention	≥12.0 months	16	–0.18 (–0.27 to –0.09)	<0.001	99.9 (<0.001)	<0.001
		<12.0 months	6	0.02 (–0.05 to 0.09)	0.629	99.3 (<0.001)
	Study quality	High	8	–0.25 (–0.37 to –0.13)	<0.001	99.8 (<0.001)	0.028
		Low	14	–0.05 (–0.15 to 0.05)	0.315	99.9 (<0.001)
BMI z-score	Sample size	≥1,000	6	–0.05 (–0.10 to –0.01)	0.020	99.9 (<0.001)	<0.001
		<1,000	12	–0.03 (–0.05 to –0.00)	0.022	99.7 (<0.001)
	Mean age	≤6.0	4	–0.01 (–0.07 to 0.05)	0.739	99.8 (<0.001)	<0.001
		6.0–10.0	7	–0.04 (–0.09 to 0.02)	0.220	99.9 (<0.001)
		>10.0	7	–0.05 (–0.06 to –0.03)	<0.001	99.4 (<0.001)
	Duration of intervention	≥12.0 months	16	–0.04 (–0.06 to –0.02)	<0.001	99.9 (<0.001)	0.223
		<12.0 months	2	0.02 (–0.09 to 0.13)	0.714	85.2 (0.009)
	Study quality	High	5	–0.10 (–0.17 to –0.04)	0.003	99.9 (<0.001)	<0.001
		Low	13	–0.01 (–0.03 to 0.01)	0.209	99.6 (<0.001)

BMI, body mass index; WMD, weighted mean difference.

BMI z-score

Data regarding the effectiveness of dietary interventions on BMI z-score in childhood were available in 17 trials (18 cohorts). The summary WMD indicated that children who received dietary interventions had greater reduction in BMI z-score (WMD: –0.04; 95% CI: –0.06 to –0.02; P=0.001; ), showing a significant heterogeneity across the included trials (I2=99.8%; P<0.001). This conclusion was unaffected by sequential exclusion of included trials (). Subgroup analyses indicated that dietary intervention was associated with greater reduction in BMI z-score irrespective of sample size (≥1,000: WMD, –0.05, 95% CI, –0.10 to –0.01, P=0.020; <1,000: WMD, –0.03, 95% CI, –0.05 to –0.00, P=0.022), mean age >10.0 years (WMD: –0.05; 95% CI: –0.06 to –0.03; P<0.001), the duration of intervention ≥12.0 months (WMD: –0.04; 95% CI: –0.06 to –0.02; P<0.001), and study with high quality (WMD: –0.10; 95% CI: –0.17 to –0.04; P=0.003). These results suggested that the effectiveness of dietary interventions on BMI z-score was affected by sample size, mean age, and study quality ().

Figure 4

Effect of dietary intervention on body mass index z-score.

Publication bias

Review of funnel plots for BMI and BMI z-score did not yield any potential publication bias (). The results of Egger and Begg tests showed no significant publication bias for BMI (P value for Egger: 0.477; P value for Begg: 0.176) and BMI z-score (P value for Egger: 0.774; P value for Begg: 0.880).

Figure 5

Funnel plots for body mass index (A) and body mass index z-score (B).

Discussion

To our knowledge, this is the first meta-analysis study that focused on dietary intervention on weight outcomes in childhood. This quantitative meta-analysis recruited 17,488 children from 28 RCTs with varied individual characteristics. The findings of this study provided stable evidence and suggested that dietary intervention was associated with greater reduction in BMI and BMI z-score in childhood. The significant effectiveness of dietary intervention on BMI mainly focused on sample size <1,000, mean age of individuals ≤6.0 or >10.0 years, duration of intervention ≥12.0 months, and high quality studies, whereas significant effectiveness on BMI z-score was mainly detected irrespective of sample size, mean age >10.0 years, duration of intervention ≥12.0 months, and high quality studies.

Several systematic reviews and meta-analyses that focused on diet, physical activity and behavioral interventions in children as well as adolescents with obesity have already been conducted. van Hoek et al. have pooled 27 studies and revealed that obese young children who received both dietary and physical activity education and behavioral therapy had the largest pooled change in BMI z-score (52). However, these results were obtained by conducting subgroup analysis of 2 studies. Brown et al. have conducted a meta-analysis of 29 studies including both South Asian children and adults. They pointed out that individuals who received diet or physical activity interventions showed significant improvement in weight, whereas no significant differences were observed in BMI and waist circumference (53). A meta-analysis conducted by Oosterhoff et al. have included 85 RCTs and found that school-based lifestyle interventions resulted in beneficial changes in BMI and blood pressure (54). A meta-analysis of 70 RCTs conducted by Mead et al. have evaluated the effectiveness of diet, physical activity and behavioral interventions in treating overweight or obese children aged 6–11 years. The results revealed that the above interventions yielded small and short-term benefits in BMI, BMI z-score and weight in children aged 6–11 years. However, whether the effectiveness of diet, physical activity and behavioral interventions on weight outcomes differed according to the individual characteristics were not elucidated (13). Hens et al. in a meta-analysis of 12 studies reported that children and adolescents who received diet or exercise interventions had better improvement in hepatic adiposity (55). However, the above meta-analyses evaluated by combining diet, physical activity and behavioral interventions on weight outcomes in children, and the effectiveness of dietary interventions on BMI and BMI z-score stratified by study or individuals’ characteristics were not illustrated. Therefore, the current meta-analysis was conducted to evaluate the effectiveness of dietary intervention on weight outcomes in childhood.

The summary results indicated that children who received dietary interventions had greater BMI reduction when compared to those with usual healthcare, while several studies have reported inconsistent results. Jansen et al. have recruited 2,622 children and found that multi-component intervention showed association with greater reduction in BMI in children of grades 3–5, whereas this effect was not detected in children of grades 6–8 (20). Siegrist et al. have included 724 children and suggested that multi-component JuvenTUM intervention did not yield any benefit on BMI when compared with usual activities (22). Graf et al. have indicated that preventive strategies used in primary schools could significantly improve motor skills, whereas energy intake and weight showed significant increase (24). Sichieri et al. have suggested that the intake of decreased sugar-sweetened beverages was associated with greater reduction in BMI, especially in girls, whereas this effect was balanced by limited effectiveness in boys (26). Puder et al. have recruited 652 pre-school children and found that a multidimensional intervention has significantly increased aerobic fitness and reduced body fat, whereas no significant effect was observed on BMI (32). Magnusson et al. have indicated that children who received intervention showed inconsistent results with regard to fitness. Moreover, the intervention did not yield any statistically significant effect on body composition (15). There are several reasons for these inconsistencies, which were as follows: (I) the intensive of intervention outside the school was not feasible, which further required extensive involvement of parents, community, and policies (56); (II) the prevalence of overweight at baseline differed, showing association with varied requirements to lower BMI; (III) additional specific measurements of the body fat are necessary to evaluate the effectiveness of dietary intervention as the effect of other components of intervention could affect BMI in generally non-obese children; and (IV) BMI is considered as a measure of general adiposity, providing no more information with regard to fat distribution, and these characteristics could be affected by other components of intervention.

In our study, individuals who received dietary intervention had greater reduction in BMI z-score. However, 5 of the included studies showed contrast conclusions (22,31,33-35). These studies indicated that children who received dietary interventions are associated with small reduction in BMI z-score. Moreover, they have pointed out that the prevalence of excess weight in school children with high socioeconomic status who received intervention showed a significant decrease. Furthermore, the skinfolds are more sensitive to changes in fat mass in children who received dietary interventions.

The results of subgroup analyses indicated the effectiveness of dietary interventions on BMI or BMI z-score, and are affected by sample size, mean age, duration of interventions, and study quality. The potential reasons for this could be due to that the (I) sample size of retrieved trials might affect the weight of the overall analysis, showing association with more stable effect size and smaller standard deviation; (II) the mean age of children is significantly correlated with the behavior of individuals and learning ability, affecting the effectiveness of dietary interventions; (III) the duration of interventions could improve the knowledge and implementation ability, and long duration of interventions is always associated with greater effect size of dietary interventions; and (IV) study quality is significantly correlated with the reliability of results in individual trial, affecting the effectiveness of dietary intervention due to uncontrolled biases.

However, there are several limitations in this study that should be highlighted. Firstly, the components of educational interventions differed among the included trials, showing significant correlation with the effectiveness of dietary interventions on weight outcomes in childhood. Secondly, substantial heterogeneity across the included trials could not be fully interpreted through sensitivity and subgroup analyses, restricting the recommendation of conclusions in this study. Thirdly, publication bias was inevitable because of the analysis published RCTs. Fourthly, language bias was inevitable as non-English databases were not searched. Finally, the analysis was based on pooled data, restricting us to conduct a more detailed analysis.

In conclusion, the results of this meta-analysis indicated that dietary interventions showed significant improvement in BMI and BMI z-score in childhood. Moreover, the effectiveness of dietary interventions was affected by sample size, mean age, duration of interventions, and study quality. Further large-scale RCTs should be conducted to evaluate the differing effectiveness of dietary interventions between boys and girls.

The article’s supplementary files as