Dietary Patterns and Childhood Obesity Risk: A Systematic Review

Abstract

Background:

Childhood overweight and obesity are recognized as predictors of the risk of obesity in adulthood. The aim of this systematic literature review was to determine the association between dietary pattern and obesity risk among children.

Methods:

Articles were selected from databases (Cochrane Library, Lilacs, Eric, Livivo, and PubMed/Medline), without limitations regarding language or date. Database-specific search terms included the key words “obesity,” “diet,” “dietary pattern,” “childhood,” “children,” “adolescents,” and relevant synonyms. The review included studies that reported the assessment of the dietary pattern in childhood and that correlated eating patterns with the obesity risk through cluster analysis (CA) and/or factor analysis and/or principal component analysis and presented odds ratios (ORs). The methodology of the selected studies was evaluated using the JBI Critical Appraisal Checklist for Analytical Cross-Sectional Studies. Owing to the heterogeneity among the assessments of the association of dietary patterns and obesity, the results are described narratively.

Results:

Using a selection process in two phases, 16 articles were included. Fifteen studies used a cross-sectional design, and one case–control study. The included studies showed variation in sample size (range = 232 to 10,187 children/adolescents) and age (range = 1–19 years old). The studies reported between two and five dietary patterns each. The OR for the relationship between the dietary pattern and the risk of childhood obesity ranged from OR = 1.02; confidence interval (95% CI) (0.91–1.15) to OR = 3.55; 95% CI (1.80–7.03). In this study, the dietary patterns identified by the studies were given different names. The food intake patterns identified could, in most factor or CA studies, be categorized as (1) potentially obesogenic foods that increased risk of becoming overweight (including fatty cheeses, sugary drinks, processed foods, fast food, candies, snacks, cakes, animal products, whole milk, and refined grains) or (2) food classified as healthy with the weakest association with the risk of becoming overweight or obese (including low levels of sugar and fat and high levels of fruits, vegetables, whole grains, fish, nuts, legumes, and yogurt).

Conclusion:

Overall, the results indicated from most studies that a diet with a lower percentage of obesogenic foods should be effective in reducing the risk of developing obesity.

Introduction

Among the most serious public health problems of the 21st century, both in developed and developing countries, being overweight in childhood and childhood obesity are recognized as predictors of the risk of obesity in adulthood.^1,2 Between 1980 and 2013, these problems underwent a large increase in developing countries from 8.1% to 12.9% for boys and 8.4% to 13.4% for girls³; in 2016, the WHO reported 41 million overweight children aged <5 years worldwide.⁴

Childhood obesity has increased over time and has different prevalence rates among different countries. In the United States, for example, the prevalence of childhood obesity increased from 12.4% in 2010⁵ to 30% in 2014,⁸ whereas in South Korea,⁶ it increased from 14.3% in 2010 to 15.3% in 2013. In 2011, in China, the prevalence of obese children was 10.1%⁷ and the prevalence was 7% in Portugal.⁹ Between 2012 and 2016, the increase in the prevalence of obese children was 30%–35% in Mexico,¹⁰ 12% in Indonesia, 11% in Thailand,¹¹ and 5%–15% in Latin American countries,¹² with lower prevalence rates in Asian countries, such as 5% in the Philippines, 3% in Myanmar, 2% in Laos, and 2% in Cambodia.¹¹

As a complex condition, childhood obesity is a risk factor for several noncommunicable diseases, such as hypertension, dyslipidemia, liver disease, diabetes, polycystic ovary syndrome, and obstructive sleep apnea, as well as psychosocial problems, including discrimination, social isolation, and low self-esteem, which can affect health, education, and quality of life.^5,9

It is estimated that 40% to 60% of obese children will become obese adults, with consequences for their health and the world economy,^1,9,13 thus it is necessary to develop effective approaches to prevent childhood obesity as a public health priority.¹⁴ As the lifestyle patterns adopted in childhood are likely to continue in adulthood,⁸ analysis of diet or dietary patterns is among the approaches that can be performed.¹⁵

Because dietary intake, as a rule, follows a pattern of consumption and diet itself is a modifiable risk factor, understanding dietary patterns can provide useful information about the associations between diet quality and its influence on health, especially in relation to obesity.^16,17 The analysis of dietary patterns has been identified as a more realistic representation of dietary habits,¹⁸ since it takes into account the complex interactions between nutrients and other components of a diet, thus making interventions to change eating habits possible.^16,17

However, Shi et al.¹⁹ suggested that the ideal is to interpret an eating pattern as a predictor of the risk of being overweight in childhood or childhood obesity, beyond associating eating patterns with nutrient intake. Thus, this study aimed to address this gap, with the purpose of this article being to review the literature to determine the association between dietary pattern and the risk of obesity among children and adolescents.

Methods

Protocol and Registration

This systematic review followed the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-analysis checklist (PRISMA).²⁰ The protocol underpinning this review is available through the PROSPERO database (see registration number CRD42019132217).

Eligibility Criteria

Inclusion criteria

Cohort studies, surveys, and cross-sectional and longitudinal studies that reported the dietary pattern assessment in childhood and that correlated eating patterns with obesity risk were included. Studies that were published in any language or date were considered.

Population: studies with samples of children aged <11 years, and/or adolescents aged 12–18 years (or a mean within these ranges) as subjects of study at baseline were included. Dietary information was reported by the children/adolescents aged >12 years or their parents, mother–child pairs, participating families, or parents or caregivers of children younger than 12 years.

This systematic review only included studies that reported risk estimates [hazards ratios, odds ratios (ORs), and relative risks] of obesity and measures of variability [effective sizes (ESs) or confidence interval (95% CIs) from which these could be derived] and studies that used validated instruments to measure food and/or dietary patterns (rather than nutrients) through cluster analysis (CA) and/or factor analysis (FA) and/or principal component analysis (PCA).

The studies reported that individuals were given factor scores for each of the patterns. Factor scores were divided into percentiles, tertiles, or quartiles on the basis of their contribution to each pattern. The studies were included when they reported dietary patterns for each food group, first categorized into percentile, tertiles, or quartiles of energy consumption and then classified into three categories: low consumption (first percentile, tertile, or quartile), moderate consumption (second percentile, tertile, or second and third quartile), and high consumption (third percentile, tertile, or fourth quartile).

Studies that determined the correlation between eating patterns and obesity risk, and used a variety of obesity measures were included: weight status was determined using age- and gender-specific BMI percentiles, BMI z-scores, BMI standard deviation scores (SDSs), BMI cutoff points, or waist circumference (WC). BMI was calculated as weight in kilograms divided by the square of height in meters [BMI = weight (kg)/(height (m²))]. Abdominal obesity was considered as waist-to-height ratio (WHR) >0.5. The WC was measured midway between the lowest rib and the superior border of the iliac crest with a measuring tape at the end of normal expiration to the nearest 0.1 cm.

The primary criteria used to define overweight and obesity include overweight: BMI or WC ≥85th percentile to 95th percentile, BMI = 1 standard deviation above the average; obesity: BMI or WC = 95th percentile, WHR >0.5, BMI = 2 standard deviations above the average, and overweight if their BMI z-score was between the 85th and 95th percentile and obese if their BMI z-score was >95th percentile.

Exclusion criteria

The following exclusion criteria were applied: 1—studies that did not correlate eating patterns with obesity through reported risk estimates [hazards ratios, ORs, and relative risks]; 2—studies that did not consider dietary patterns through CA and/or FA and/or PCA; 3—participants who were not children and adolescents; 4—studies that analyzed only one isolated macronutrient or nutrient and not the entire diet; 5—studies with well-defined/characterized samples of nonhealthy subjects (e.g., people with type 2 diabetes, hypertension patients, and eating disorders). Supplementary Appendix Table A1 contains the articles excluded and the reasons for exclusion.

Information Sources and Search Strategy

Detailed individual search strategies were performed in each of the following electronic databases: Cochrane Library, Lilacs, Eric, Livivo, and PubMed/Medline. A gray literature search was performed in Google Scholar. In addition to the electronic search, a hand search was made and the reference lists of the selected articles were screened). Search terms and Boolean operators of the medical subject headings, keywords, and/or other controlled vocabulary terms were used in searches of the PubMed and the others electronic databases. All electronic searches were conducted from their earliest records up to March 29, 2018 and updated to October 10, 2018. Database-specific search terms included the key words “obesity,” “diet,” “dietary pattern,” “childhood,” “children,” “adolescents,” and relevant synonyms. Supplementary Appendix Table A2 contains the complete strategy search terms and databases.

Study Selection

The articles were selected in two phases. In phase 1, two authors (R.L. and E.K.) independently reviewed the titles and abstracts of all the references. During this phase, any articles that did not meet the eligibility criteria already described in the inclusion and exclusion criteria session were excluded. In phase 2, they applied the same selection criteria to the full text of the articles to determine which articles to include. The same two authors independently reviewed the full texts of all the articles.

Data Collection Process and Data Items

One author (R.L.) extracted the data from the selected studies, including study characteristics (author, year of publication, study design, and location), population and sample characteristics, diet assessment method, dietary patterns identified, dietary patterns and their key components, findings [factors adjusted for in the analyses (multivariable), OR, prevalence ratios, and 95% CI], and main conclusion. A second author (E.K.) cross-checked all the information retrieved.

Risk of Bias in Individual Studies

Risk of bias in the individual studies included was conducted using the JBI-MAStARI (JBI Meta Analysis of Statistics Assessment and Review Instrument). The questionnaire consists of eight questions that were answered with yes, no, unclear, or not applicable.^54,55 In the JBI-MAStARI, two criteria listed in the instrument were deemed irrelevant to the nature of the studies. The studies were classified as follows: high methodological quality (>5 “yes” responses), moderate methodological quality (3–4 “yes” responses), or low methodological quality (0–2 “yes” responses).⁵⁶ Supplementary Appendix Table A3 contains the risk of bias of the selected studies by JBI-MAStARI.

Summary Measures

The dietary patterns identified by CA and/or FA and/or PCA and the association with obesity were considered as the outcome. Risk estimates [hazards ratios, ORs, and relative risks] of obesity and measures of variability (SEs or 95% CIs from which these could be derived) for the outcome measurement were considered in this review. The original studies reported the results of dietary patterns in terms of quartiles or tertiles of dietary factor scores and obesity risk.

Data Synthesis and Analysis

We created tables of evidence containing information from the included studies. Owing to the heterogeneity among the assessments of the association of dietary patterns and BMI or obesity, it was not possible to perform a meta-analysis; therefore, the results are described narratively. It was not possible to assess publication bias through funnel plots as no statistical data synthesis could be performed.

Results

Study Selection

From the 1570 references initially identified, a comprehensive evaluation of the abstracts was performed after removal of duplicates; 1411 articles were excluded, resulting in finally 159 articles. After title and abstract reading, 52 potentially useful studies were acquired. Of these 36 articles were excluded for various reasons (Supplementary Appendix Table A1). Thus, 16 studies were retained for the final selection. A flowchart of the process of study identification and selection is presented in Figure 1.

Figure 1.

Flow diagram of search.

Study Characteristics

The 16 studies were published between 2010 and 2018. The studies were all cross-sectional studies, with the exception of one case–control study.⁷⁰ The studies were conducted in nine different countries and the majority of studies were from Asian countries (Iran,⁷⁰ China,^{61,65,66,69,71} and Lebanon^64,67), with four from American countries (Mexico,⁵⁸ Brazil,^60,63 and Colombia⁶²), two from European countries (Norway^57,59), one from Australia,¹⁹ and one from Polynesia (Samoan island of Upolu⁶⁸).

Most of the articles were published in English,^{19,57–59,61,64–71} two articles were published in Portuguese,^60,63 and one article was published in Spanish.⁶²

Sample sizes ranged from 232 to 10,187 children and adolescents. Mean age among the study participants ranged from 1 to 19 years. Six studies were with children aged between 1 and 6 years,^{58,61,62,65,69,70} five studies with children aged between 5 and 18 years,^{19,60,67,68,71} two studies with children aged between 9 and 13 years,^57,59 and three studies with adolescents in the age group between 11 and 19 years.^63,64,66

Seven studies included >1000 participants.^{58,61,62,65,66,70,71} The total sample from the 16 selected studies included 44,793 children and adolescents. The total sample from Asian countries was n = 22,263,^{61,64–67,69–71} the American countries was n = 19,145,^58,60,62,63 the European countries was n = 1351,^57,59 Australia was n = 288,¹⁹ and was n = 305 from the Samoan island of Upolu.⁶⁸

The most popular used methods for deriving dietary patterns are PCA, FA, and CA. Of the 16 selected studies, 1 used CA,⁵⁸ 7 used FA,^{19,61,64–67,71} and 8 used PCA.^{57,59,60,62,63,68,69,70}

Using PCA or FA or CA to identify dietary patterns, the majority of the studies included in this review extracted two,^{64,67–69,71} three,^{19,60,63,65,66,70} or four patterns.^57,59,61 The cumulative proportion of the total variance in dietary intake explained by the patterns of these studies ranged from 16.36%⁶⁹ to 38.18%,⁷⁰ accounting for only a limited portion of the between-person variance in diet. Of note, the identification of three factors in the analysis by Yang et al.⁶⁶ that accounted for 65.6% of the cumulative proportion of the total variance is out of the range of what was reported in the aforementioned studies.

Because age, gender, socioeconomic status, demographic and environmental factors, and physical activity level are associated with both the dietary patterns and the risk of overweight and/or obesity in children and adolescents, these confounders must be included in the adjusted models. In addition, the risk of Ow/Ob in children might be linked to parental weight status and child's behavioral factors such as sedentary activity and duration of sleep.

Regardless of the difference in confounding variables and their characterization among the included studies, statistical adjustments in the multivariate analysis varied across the studies. This systematic review only included studies that reported risk estimates (ORs). For the estimative risk analysis and strategies to deal with confounding factors stated (calculate unadjusted and adjusted ORs), the studies used the following statistics: multiple logistic regression,^{19,57,58–60,63,65–69,71} logistic regression random effects model,⁶¹ binomial regression model,⁶² multiple linear regression,⁶⁴ conditional logistic model.⁷⁰

Possible confounding factors were controlled in the patterns of regression and ORs were calculated by adjusting for potential risk factors of obesity. The studies show many confounders, however, the majority of studies considered the age,^{19,58,61–65,67–71} gender,^{19,58,61–65,67–71} parental characteristics (family education, family income, parental obesity, and unhealthy behavior),^{19,57,59,61,63,64,67,69,71} and physical activity level,^{57,59,61,64–66,70} as important confounders.

The most important confounders associated with the obesity risk are age,^{19,58,61,62,64,65,68–71} gender,^{19,59,61,62,64,65,68–71} physical activity level,^{57,64,65,67,70} parental characteristics (parental BMI,^57,58,64,67 education,^{19,57,59,64,67} family income,^59,65 and family history of chronic diseases⁷⁰), BMI,^64,67,70 living area,^65,68,70 screen time,^64,67,70 smoking status,^64,67 and frequency of eating out per week.^64,67 A summary of the descriptive characteristics of the included studies is provided in Table 1.

Table 1.

Summary of the Characteristics of the Included Studies

Author (year) LocationStudy design	Population and sample	Diet assessment method	Dietary patterns identified and their key components	Factors adjusted for in analyses (multivariable) model (M) confounders	Risk obesity—lowest and highest by dietary Odds ratios (95% confidence interval), p value	Conclusion
Oellingrath et al. (2010)⁵⁷ Norway Cross-sectional	924 out of 1477 9–10 years Primary School Dietary information reported by their parents	Short FFQ 39 food items, 11 types of dinks, 13 snack items, and 5 main meals during the past 6 months BMI cutoff points	Snacking Sugar-sweetened drinks consumed between meals, low breakfast and dinner frequency and low intake of water, vegetables, and brown bread Junk/convenient High-fat, high-sugar processed, fast foods, french fries, processed pizza, processed meat products, sweets, ice cream, and soft drinks Varied Norwegian Food items typical of a traditional Norwegian diet, fish, meat for dinner, brown bread, regular white or brown cheese, lean meat, fish spread, fruit, and vegetables Dieting Artificially sweetened soft drinks, fat cheese, and fat- and sugar-reduced yoghurt	M1 (unadjusted) M2 (parental BMI, maternal education and physical activity level of the child)	Tertile 1 (Ref eating patterns) Snacking M1—1.3 (0.8–1.9) M2—1.0 (0.7–1.6) Junk/convenient M1—0.6 (0.4–0.9) M2—0.6 (0.4–0.9) Varied Norwegian M1—1.7 (1.1–2.6) M2—2.1 (1.3–3.2) Dieting M1—2.6 (1.7–4.0) M2—2.2 (1.4–3.4)	Confounders associated with the obesity risk: Parental characteristics Physical activity Lower likelihood of being obese: “junk/convenient” Obesity was most prevalent among children with the dietary pattern: “Varied Norwegian” or a “dieting”
Rodríguez-Ramírez et al. (2011)⁵⁸ Mexican Cross-sectional	8252 out of 9383 5–11 years Mexican school Dietary information reported by their mother or guardian	7-day FFQ 102 food items BMI cutoff points	Rural Corn tortilla, legume, low intake of sweets and some cereals such as white bread, whole-wheat bread, rice, and noodles Sweet cereal and corn dishes Sweet cereals, corn dishes, low-fat milk, sweets, cereals in general, dairy products Diverse Sweetened beverages, meat and poultry, dairy products, vegetables and fruits, corn tortillas Western Soft drinks, cakes, dishes with fat, and salty snacks. Whole milk and sweet Whole milk, sweet groups, lower proportion of corn tortilla, cereals, sweetened beverages, and low-fiber cereals	M1 (unadjusted) M2 (age) M3 (age, área)	Rural (percentiles) (Ref eating patterns) Sweet cereal and corn dishes M1—1.34 (1.16–2.06) p < 0.05 M2—1.43 (1.30–2.27) p < 0.05 M3—1.29 (1.09–1.94) p < 0.05 Diverse M1—1.28 (1.06–1.92) p < 0.05 M2—1.34 (1.13–2.05) p < 0.05 M3—1.21 (0.97–1.76) Western M1—1.45 (1.33–2.47) p < 0.05 M2—1.53 (1.49–2.72) p < 0.05 M3—1.35 (1.17–2.19) p < 0.05 Whole milk and sweet M1—1.23 (0.99–1.80) M2—1.35 (1.14–2.04) M3—1.17 (0.91–1.67)	Lower likelihood of being obese: “rural” Obesity was most prevalent among children with the dietary pattern: “Western” or a “sweet cereal and corn dishes”
Oellingrath et al. (2011)⁵⁹ Norway Cross-sectional	1.789 out of 2980 (924–4th grade) and 865–7th grade) 9–13 years Primary School Dietary information reported by their parents	FFQ 40 food items, 11 types of dinks, 13 snack items, and 5 main meals during the past 6 months BMI cutoff points	Snacking Snack, sugar-sweetened drinks, low intake of water, vegetables and brown bread. Junk/convenient High-fat and high-sugar processed fast foods, french fries, hamburger or kebab, french fries, biscuits, cakes, sausages, hot dog, processed pizza and meat, waffles, sweets, salty snacks, white bread, ice cream. varied Norwegian Fish and meat, brown bread, regular white or brown cheese, lean meat, fish spread, fruit and vegetables Dieting Cereals without sugar, juice between meals, artificially sweetened soft drinks, carbonated and noncarbonated, fat- and sugar-reduced yoghurt	M1 (unadjusted) M2 (maternal and paternal overweight, and education, family income, child physical activity, sedentary and gender)	Results from 7th grade—(691 children) Tertile 1 (Ref eating patterns) Snacking M1—0.9 (0.5–1.5) M2—1.0 (0.6–1.6) Junk/convenient M1—0.8 (0.5–1.3) M2—0.8 (0.5–1.4) varied Norwegian M1—0.5 (0.3–0.8) M2—0.6 (0.4–0.9) Dieting M1—2.1 (1.3–3.4) M2—2.2 (1.3–3.8)	Confounders associated with the obesity risk: Parental characteristics Physical activity Gender Lower likelihood of being obese: “Varied Norwegian” Obesity was most prevalent among children with the dietary pattern: “dieting”
Nobre et al. (2012)⁶⁰ Brazil Cross-sectional	232 out of 310 1–4 years Preschool children Dietary information reported by their mother–child pairs	FFQA 24 food items BMI z-score	Mixed diet Beef and pork, cooked beans, milk and dairy products, rice and roots, farinaceous, sweet and savory cookies, cakes, leafy vegetables, fruits, natural juices Snack Milk and dairy products, sweet and savory biscuits, fruits, natural juices, breads, margarine, chocolate milk Unhealthy Sweets, artificial juices, soft drinks, sweets and desserts, stuffed biscuit, fried or boiled eggs	M1 (unadjusted) Normal weight (Ref eating patterns)	Mixed diet M1—0.57 (0.24–1.37) ns Snack M1—1.29 (0.59–2.83) ns Unhealthy M1—0.97 (0.44–2.14) ns	Lower likelihood of being obese: “Mixed diet” Obesity was most prevalent among children with the dietary pattern: “snack”
Shang et al. (2012)⁶¹ China Cross-sectional	5.267 out of 7500 6–13 years Primary schools Dietary information reported by their parents	24 hours-recall 3 consecutive days Foods to 28 groups BMI z-score	Healthy Milk, yogurt, eggs, fruit, and vegetables and low intake of meat such as pork, poultry, organ meat, and beef/lamb/other red meat Transitive organ meat, pork, seafood, processed meat, edible fungi and algae, light vegetables Western rice, refined grains, deep color vegetables, pork, sugar, fish, and shrimp, beef, lamb, and red meat, wheat, starch tubers, light color vegetables	M1 (gender, age) M2 (M1+feeding types, birth weight, parents' weight, parents' educational level, average family income per month per capita and school in center) M3 (M2+total energy intake (kcal/day) and physical activity energy expenditure)	Healthy (percentiles) (Ref eating patterns) Transitive M1—1.21 (0.98–1.48) ns M2—1.17 (0.96–1.44) ns M3—1.11 (0.89–1.38) ns Western M1—2.04 (1.38–3.02) p < 0.001 M2—1.79 (1.20–2.67) p < 0.004 M3—1.80 (1.15–2.81) p < 0.009	Confounders associated with the obesity risk: Gender Age Lower likelihood of being obese: “Healthy” Obesity was most prevalent among children with the dietary pattern: “Western”
Ocampo et al. (2014)⁶² Colombian Cross-sectional	10.187 out of 17.897 5–17 years Instituto Colombiano de Bienestar Familiar (ICBF) Dietary information reported by adolescents and their parentes (either mother or father)	FFQ 28 food items BMI z-score	Proteico/Fibra Fruits, cheese, yogurt, cream cheese, vegetables, cooked vegetables, whole fruits, bread, whole foods, chicken, beef viscera, butter, tuna or sardinas, fish or seafood Bocadillo/snack Packaged foods, candy or sweets, soft drinks, fast food, sausages Tradicional/Almidón Sugar, honey, rice, or pasta, fried foods dry grains tubers or bananas, eggs, beef, veal, pork, rabbit, goat, coffee or tea, chicken noodles	M1 (unadjusted) M2 (age, gender) M3 (age, gender, level of wealth)	Quartile 1 (Ref eating patterns) Proteico/Fibra M1—1.64 (1.32–2.00) p < 0.001 M2—1.64 (1.23–2.04) p < 0.001 M3—1.26 (1.00–1.58) p < 0.001 Bocadillo/snack M1—1.18 (0.91–1.52) ns M2—1.13 (0.87–1.46) ns M3—0.84 (0.64–1.09) ns Tradicional/Almidón M1—0.56 (0.46–0.69) p < 0.001 M2—0.58 (0.47–0.71) p < 0.001 M3—0.65 (0.53–0.80) p < 0.001	Confounders associated with obesity risk: Age Gender Level of wealth Lower likelihood of being obese: “Tradicional/Almidón” Obesity was most prevalent among children with the dietary pattern: “Proteico/Fibra”
Pinho et al. (2014)⁶³ Brazil Cross-sectional	474 out of 535 11–17 years Public schools Dietary information reported by their parents	FFQA 26 food items BMI z-score	Junk food pasta, whole dairy products, margarine, breads, chocolate milk powder, sweets, and desserts, sugar, filled biscuits, fatty foods, and mayonnaise Healthy skim dairy products, cereals, leafy vegetables, fruits, and natural juices Traditional rice, tuberous roots, beans, red and white meat, eggs, processed meats, and artificial juices	M1 (unadjusted) M2 (gender) M3 (age, family income)	Normal weight (percentiles) (Ref eating patterns) Junk food M1—1.05 (0.68–1.63) Healthy M2—0.56 (0.35–0.91) Traditional M1—0.76 (0.50–1.20)	Confounders associated with the obesity risk: gender Lower likelihood of being obese: “Healthy” Obesity was most prevalent among children with the dietary pattern: “junk food”
Naja et al. (2015)⁶⁴ Lebanese Cross-sectional	446 out of 3636 13–19 years Lebanese adolescents Dietary information reported by adolescents and their parents (either mother or father)	Single multiple-pass 24-hours recall Semiquantitative FFQ BMI z-score	Traditional Lebanese vegetables, legumes, bread, rice, pasta, and cereals, bulgur, fruits, fish, and vegetable oils Western poultry and eggs, red meat, mayonnaise, fast-food sandwiches, pizza, and pies	M1 (gender, age) M2 (m1+mother's education, working status, parental obesity, smoking status, physical activity level, frequency of eating while watching television per week, frequency of eating out per week, BMI and total energy intake)	Tertile 1 (Ref eating patterns) Traditional Lebanese M1—1.60 (0.92–2.77) M2—1.41 (0.64–3.13) Western M1—2.12 (1.24–3.63) M2—2.31 (1.12–4.73)	Confounders associated with the obesity risk: Physical activity Frequency of breakfast consumption Lower likelihood of being obese: “Traditional Lebanese” Obesity was most prevalent among children with the dietary pattern: “Western”
Zhang et al. (2015)⁶⁵ China Cross-sectional	1282 7–17 years from the 2011 China Health and Nutrition Survey (CHNS) Dietary information reported by children ≥12 years and their mother for children younger than 12 years.	Three consecutive 24 hours recall 20 food items BMI percentile	Modern Milk, fast foods, eggs, other livestock meat, poultry, and cake Traditional north wheat flour products and other cereals, combined with low consumption of rice, pork, and poultry Traditional south vegetables, rice, pork, and legumes	M1 (age, gender, living area, annual household income, physical activity) M2 (M1+energy intake)	Quartile 1 (Ref eating patterns) Modern M1—3.55 (1.80–7.03) p < 0.000 M2—3.10 (1.52–6.32) p < 0.000 Traditional north M1—2.45 (1.35–4.44) p < 0.000 M2—2.42 (1.34–4.39) p < 0.000 Traditional south M1—1.44 (0.81–2.57) ns M2—1.15 (0.62–2.11) ns	Confounders associated with the obesity risk: age, gender, living area, physical activity, annual household income per family member, and energy intake Lower likelihood of being obese: “Traditional south” Obesity was most prevalent among children with the dietary pattern: “Modern” or a “Traditional north”
Yang et al. (2016)⁶⁶ China Cross-sectional	1590 out of 2214 11–17 years From urban and rural primary and middle schools Dietary information reported by adolescents	FFQ 25 food items 24 hours recall BMI WC	Chinese and Western Rice, flour, whole grains, fresh vegetables, fresh fruit, poultry, eggs, freshwater fishes, and shrimps, deep-sea fish, milk and dairy products, beans and bean products, nuts, snacks, sugar, and barbecue Western rice and rice products, flour and flour products, fresh fruit, fat meat, hamburgers and fried foods, processed products, snacks, coke, sprite, coffee, ice cream, instant noodles, and barbecue Meat diet fat meat, beef and mutton, poultry, animal liver, freshwater fishes and shrimps and deep-sea fish.	M1 (unadjusted) M2 (gender, passive smoking, drinking, calcium supplements, body mass index) M3 (physical activity)	Tertile 1 (Ref eating patterns) Chinese and Western M1—0.89 (0.62–1.29) p < 0.005 M2—1.07(0.69–1.65) ns M3—1.08(0.70–1.68) ns Westernization M1—1.76(1.15–2.73) p < 0.000 M2—1.76 (1.14–2.73) p < 0.000 M3—1.92 (1.29–2.85) p < 0.000 Meat diet M1—1.45 (1.00–2.14) p < 0.000 M2—1.23 (1.02–1.52) p < 0.000 M3—1.11 (0.81–1.51) ns	Lower likelihood of being obese: “Chinese and Western” Obesity was most prevalent among children with the dietary pattern: “Westernization”
Nasreddine et al. (2017)⁶⁷ Lebanese Cross-sectional	525 out of 531 2–5 years From preschoolers Dietary information reported by their mother	5-Step Multiple pass FFQ 15 foods/food groups BMI z-score	Traditional Lebanese Vegetables, legumes, bread, rice, pasta, and cereals, bulgur, fruits, fish, and vegetable oils Western poultry and eggs, red meat, mayonnaise, fast-food sandwiches, and pizza and pies	M1 (age, gender) M2 (mother's education, working status of the mother, parental obesity, smoking status, physical activity level, frequency of eating while watching television per week, frequency of eating out per week, BMI and total energy intake)	Tertile 1 (Ref eating patterns) Traditional Lebanese M1—1.60 (0.92–2.77) p < 0.000 M2—1.41 (0.64–3.13) ns Western M1—2.12 (1.24–3.63) p < 0.000 M2—2.31 (1.12–4.73) p < 0.000	Confounders associated with the obesity risk: Mother's education Weekly frequency of Meals Age Gender Lower likelihood of being obese: “Traditional Lebanese” Obesity was most prevalent among children with the dietary pattern: “Western”
Choy et al. (2018)⁶⁸ Samoan island of Upolu Cross-sectional	305 2–5 years From 10 villages on the Samoan island of Upolu Dietary information reported by their mother–child pairs	117-item FFQ with a 30 day reference period. BMI cutoff points	Modern Unprocessed red meats, whole grains (which are not traditional to Samoa), corn, French fries, butter, mayonnaise, nuts, and snack foods such as potato chips Neotraditional Vegetables, local starchy crops such as taro and breadfruit, coconut, fish, poultry and a low intake of desserts, pizza, dairy products (milk, cheese ice cream) and infant food items	M1 (unadjusted) M2 (age, gender, total energy intake, census region and material lifestyle score (quartiles)	Quartile 1 (Ref eating patterns) Modern M1—1.00 (0.38–2.61) ns M2—0.71 (0.74–2.08) ns Neo Traditional M1—1.80 (0.61–5.30) ns M2—3.43 (1.03–11.39) ns	Confounders associated with the obesity risk age, gender, total energy intake, census region and lifestyle Lower likelihood of being obese: “Modern” Obesity was most prevalent among children with the dietary pattern: “Neotraditional”
Shi et al. (2018)¹⁹ Australia Cross-sectional	288 1–4 years From the FINS (Food Intake and Nutritional Status of preschool children) Dietary information reported by their parents	3-day weighed food diary 25 food items BMI z-score	Traditional Grains, fruit, vegetables, and red meat Processed snack foods, processed meats, beverages, and confessionary Health conscious Polyunsaturated margarines, low-fat dairy products, fish, eggs, and composite foods that contain vegetables as a key ingredient	M1 (unadjusted) M2 (age, gender, duration of breastfeeding, mother's education and other dietary patterns) M3 (M2+energy and protein intake)	Tertile 1 (Ref eating patterns) T3—traditional M1—2.74 (1.08–6.94) p.03 M2—3.12 (1.11–8.74) p.03 M3—1.78 (0.59–5.43) ns T3—Processed M1—0.81(0.33–1.98) ns M2—1.05(0.37–2.99) ns M3—1.06 (0.35–3.24) ns T3—Health conscious M1—1.55 (0.68–3.53) ns M2—1.80 (0.75–4.31) ns M3—1.84 (0.75–4.52) ns	Confounders associated with the obesity risk age, gender, duration of breastfeeding, mother's education, and other dietary patterns Lower likelihood of being obese: “Processed” or a “Health conscious” Obesity was most prevalent among children with the dietary pattern: “Traditional” was positively associated with protein intake
Zhen et al. (2018)⁶⁹ China Cross-sectional	489 out of 736 6–14 years China Health and Nutrition Survey (CHNS) Dietary information reported by children ≥12 years and their parents for children younger than 12 years	24 hours recall 28 food groups BMI cutoff points	Traditional Chinese Rice, red meat, pork, poultry, vegetables (leafy) and fish Modern Wheat buns, cakes, legume products, nuts, pickled and salted vegetables, fruit, red meat, processed meats, poultry, eggs, fish, milk, and fast food	M1 (gender, age) M2 (M1+residency, highest level of parental education, region, physical activity and energy intake) M3 (M2+geographical regions)	Quartile 1 (Ref eating patterns) Traditional Chinese M1—0.28 (0.11–0.70) M2—0.19 (0.07–0.49) M3—0.19 (0.09–0.40) Modern M1—2.17 (0.99–4.77) M2—2.02 (0.88–4.65) M3—2.02 (1.17–3.48)	Confounders associated with the obesity risk Age, gender Lower likelihood of being obese: “traditional Chinese” Obesity was most prevalent among children with the dietary pattern: “Modern”
Kelishadi et al. (2018)⁷⁰ Iran Case–control	3843 out of 4200 7–18 years, Provinces in Iran Dietary information reported by their parents	Questionnaire eating habits 16 food groups Abdominal obesity	Healthy Diet Fresh fruit, dried fruit, fresh fruit juice, vegetables, milk, yogurt, and cheese Western Salty snacks (cheese puffs, chips and pretzels), beer, fast foods (sausages, burgers and pizza) and coffee Sweet Sweet snacks (cakes, cookies, pastries, biscuits, chocolate), soft drinks, packaged fruit juice, tea, cubed sugar, and sugar	M1 (age, gender, living area, SES, parent perception about student body image, weigh satisfaction, physical activity, sleep duration, family history of chronic diseases, parents' BMI, screen time).	Healthy Diet M1—1.22 (0.95–1.56) Western M1—0.84 (0.65–1.08) Sweet M1—1.29 (1.01–1.66)	Lower likelihood of being obese: “Western” Obesity was most prevalent among children with the dietary pattern: “Sweet”
Rutayisire et al. (2018)⁷¹ China Cross-sectional	8900 out of 9100 3–6 years Cities in eastern China Dietary information reported by their parents or caretaker	FFQ 36 food groups BMI percentiles	Traditional Chinese Wheat and other cereals, tubers, legumes, fruits, vegetables, fresh juices, eggs, low-fat dairy products, poultry, and fish Modern Western fast food, Chinese fast food, preserved food, fried vegetables, fried meats, sweet course, sugary foods, chocolates/ice cream, carbonated beverages, flavored milk drinks and synthesized fruit/vegetable juice	M1 (gender, age, sibling) M2 (maternal education, paternal smoking status, parental BMI).	Tertile 1 (Ref eating patterns) Traditional Chinese M1—1.02 (0.91–1.15) ns M2—1.02 (0.90–1.15) ns Modern M1—0.91 (0.81–1.03) ns M2—0.92 (0.82–1.04) ns	Lower likelihood of being obese: “Modern” Obesity was most prevalent among children with the dietary pattern: “Traditional Chinese”

Abdominal obesity was considered as WHR >0.5. WHR waist-to-height ratio.

BMI was calculated as weight in kilograms divided by the square of height in meters [BMI = weight (kg)/(height (m²))]. We included studies that determine the correlated eating patterns with obesity risk, and used variety of obesity measures: body weight status was determined using age- and gender-specific BMI percentiles, BMI z-scores, BMI SDSs, BMI cutoff points or WC. BMI was calculated as weight in kilograms divided by the square of height in meters [BMI = weight (kg)/(height (m²))]. Abdominal obesity was considered as WHR >0.5. WHR waist-to-height ratio. And WC measured midway between the lowest rib and the superior border of the iliac crest with an inelastic measuring tape at the end of normal expiration to the nearest 0.1 cm. The primary criteria used to define overweight and obesity include overweight: BMI or WC ≥85th percentile to 95th percentile, BMI = 1 standard deviation above the average; obesity: BMI or WC = 95th percentile, BMI = 2 standard deviations above the average. Children were classified as overweight if their BMI z-score was between the 85th and 95th percentile and obese if their BMI z-score was above the 95th percentile.

24 hour recall, 24 hour dietary recall method; DP, dietary patterns; FFQ, food frequency questionnaire; FFQA, Food Frequency Questionnaire for Adolescents; ns, not significant; SDSs, standard deviation scores; WC, waist circumference; WHR, waist-to-height ratio.

Dietary Pattern

Table 1 presents the 16 studies that examined the relationship between the dietary pattern derived from CA and/or FA and/or PCA and the risk of childhood obesity. Information on dietary exposures was collected by various methods.

The food frequency questionnaire (FFQ) was one of the most commonly used methods to assess individual long-term dietary intakes of foods and nutrients in the studies, and is appropriate for investigating dietary patterns on the basis of frequencies. The development of food lists (ranging from 5 to 350 items) is crucial for the success of an FFQ, with the median number being 79. The Food Frequency Questionnaire for Adolescents (FFQA) adapted from the FFQ was used to investigate dietary patterns on the basis of frequencies from a list of foods and portion sizes according to age (adolescents) and coherent with dietary patterns and eating habits of this population group. Dietary intake data were most commonly collected using an FFQ,^{57,59,62,64,66–68,71} followed by multiple 24-hour recalls,^{61,64–66,69} with two studies using FFQA,^60,63 only one study used an eating habits questionnaire,⁷⁰ and one study using a 3-day weighed food diary.¹⁹

In all the studies, children and adolescents were classified as overweight or obese according to the BMI cutoff points,^{57–59,68,69} BMI z-score,^{19,60–64,67} or BMI percentiles,^65,71 with only one study using WC (95th percentile)⁶⁶ and one study using WHR (>0.5).⁷⁰

The studies reported between two and five dietary patterns each. The food items with high and low loadings in each pattern are presented in Table 1.

In Norway, Oellingrath et al.^57,59 reported four dietary patterns: “snacking,” “junk/convenience,” “varied Norwegian,” and “dieting.”

Rodríguez-Ramírez et al.⁵⁸ reported five dietary patterns in the Mexican study: “rural,” “sweet cereal and corn dishes,” “diverse,” “Western,” and “whole milk and sweet,” and Ocampo et al.⁶² reported three dietary patterns in the Colombian study: “Proteico/fibras,” “Bocadillo/snack,” and “traditional/almidon.” In Brazil, Nobre et al.⁶⁰ and Pinho et al.⁶³ reported three dietary patterns each: “mixed diet;” “snack” and “unhealthy;”⁶⁰ and “junk food,” “healthy,” and “traditional.”⁶³

In Australia,¹⁹ three dietary patterns were reported: “traditional,” “processed,” and “health conscious.” In Lebanon, two studies^64,67 reported two dietary patterns: “traditional Lebanese” and “Western.”^64,67 Choy et al.⁶⁸ reported two dietary patterns from the Samoan island of Upolu: “modern” and “neotraditional.”

Kelishadi et al.⁷⁰ reported three patterns from Iran: “healthy,” “Western,” and “sweet.” In China, three studies^61,65,66 reported three dietary patterns: “healthy,” “transitive,” and “Western”⁶¹; “modern,” “traditional north,” and “traditional south”⁶⁵; “Chinese wester,” “Westernization,” and “meat diet”⁶⁶; and two studies^69,71 reported two dietary patterns: “traditional Chinese” and “modern.”^69,71

Dietary Patterns Associated with Childhood Obesity Risk

The studies that reported the association between dietary patterns and a significantly higher likelihood of being overweight or obese are given in Table 1. The OR for the relationship between dietary pattern and childhood obesity risk ranged from OR = 1.02; 95% CI (0.91–1.15) to OR = 3.55; 95% CI (1.80–7.03).

From the European countries, in Norway, Oellingrath et al.^57,59 reported higher rates of obesity risk with a “dieting” diet, OR = 2.2; 95% CI (1.4–3.4) in children aged 9–10 years⁵⁷ and OR = 2.2; 95% CI (1.3–3.8) in children aged 12–13 years.⁵⁹

From the American countries, Mexican children adhering to a “Western”⁵⁸ eating pattern had a significantly higher likelihood of being overweight, OR = 1.53; 95% CI (1.4–2.7). In Brazil, children adhering to a “snack” diet,⁶⁰ OR = 1.29; 95% CI (0.5–2.8) and/or “junk food” diet,⁶⁴ OR = 1.05; 95% CI (0.6–1.6) had a significantly higher likelihood of being overweight. And in Colombia, Ocampo et al.⁶² reported higher rates of obesity risk with a “Proteico/fibra” diet, OR = 1.64; 95% CI (1.3–2.0).

From the countries of Oceania, in Australia,¹⁹ children adhering to a “traditional” eating pattern had a significantly higher likelihood of being overweight, OR = 3.12; 95% CI (1.1–1.7). On the Samoan island of Upolu,⁶⁸ those that followed the “neo-traditional” eating pattern presented a significantly higher likelihood of being overweight, OR = 3.43; 95% CI (1.03–11.4).

From the Asian countries, in Iran children adhering to a “sweet”⁷⁰ eating pattern had a significantly higher likelihood of being overweight, OR = 1.29; 95% CI (1.01–1.66). Lebanese children who presented a “Western”^64,67 eating pattern had a significantly higher likelihood of being obese, OR = 2.31; 95% CI (1.12–4.73). In China, children adhering to a “Western” eating pattern had a significantly higher likelihood of being overweight, OR = 2.04; 95% CI (1.38–3.02),⁶¹ as well as those who presented a “modern,” OR = 3.55; 95% CI (1.8–7.03),⁶⁵ and OR = 2.17; 95% CI (0.99–4.77),⁶⁹ “Westernization,” OR = 1.92; 95% CI (1.29–2.85),⁶⁶ and “traditional Chinese” eating pattern, OR = 1.02; 95% CI (0.9–1.1).⁷¹

Mapping of High Positive Loadings Dietary Patterns and Their Key Components Associated with Childhood Obesity Risk

The studies that reported the association between dietary patterns and obesity risk are given in Table 1. A higher and lower likelihood of being overweight or obese and the relationship between the dietary patterns and their key components and the risk of childhood obesity were seen in the following dietary patterns.

Higher likelihood of being overweight or obese

From European countries: “dieting,” in Norway^57,59 (cereals without sugar, juice between meals, artificially sweetened, carbonated, and noncarbonated soft drinks, and fat- and sugar-reduced yoghurt).

From the Asia countries: in China, “modern,”⁶⁹ “Western,”⁶¹ “Westernization,”⁶⁶ and “traditional north”⁶⁵ (milk, sugar, coke, sprite, coffee, ice cream, fast foods, snacks, eggs, poultry, red meat, fatty meat, processed meats, pork, fish and shrimp, lamb, hamburgers and fried foods, wheat buns, processed products wheat, cakes, legume products, tubers, nuts, pickled and salted vegetables, starch tubers and light colored vegetables, fruit and fresh fruit, rice and rice products, refined grains, flour and flour products, and barbecued foods) and “traditional Chinese”⁷¹ (wheat and other cereals, tubers, legumes, fruits, vegetables, fresh juices, eggs, low-fat dairy products, poultry, and fish). “Sweet” in Iran⁷⁰ (sweet snacks, cakes, cookies, pastries, biscuits, chocolate, soft drinks, packaged fruit juice, tea, cubed sugar, and granulated sugar); “Western” in Lebanon^64,67 (poultry and eggs, red meat, mayonnaise, fast-food sandwiches, pizza, and pies).

From the America countries: in Brazil, “snack,”⁶⁰ “junk food”⁶⁴ (milk and dairy products, chocolate milk powder, sweet, savory biscuits, desserts, sugar, and filled biscuits, fruits, natural juices, breads, pasta, margarine, mayonnaise, whole dairy products, sweets, and fatty foods); “Proteico/fibra” in Colombia⁶² (fruits, cheese, yogurt, cream cheese, vegetables, cooked vegetables, whole fruits, bread, whole foods, chicken, beef viscera, butter, tuna or sardines, fish, or seafood); “Western” in Mexico⁵⁸ (soft drinks, sugar, cakes, wheat, starch tubers, dishes with fat, salty snacks, fish and shrimp, beef, lamb and red meat, pork, poultry and eggs, mayonnaise, fast-food sandwiches, pizza and pies, rice, refined grains, and dark and light colored vegetables).

From the Oceania countries: “neotraditional,” on the Samoan island of Upolu⁶⁸ (vegetables, local starchy crops such as taro and breadfruit, coconut, fish, poultry and a low intake of desserts, pizza, and dairy products (milk, cheese, and ice cream) and infant food items (infant formula, breast milk, infant packaged food purées, and prepared meals)); and “traditional” in Australia¹⁹ (grains, fruit, vegetables, and red meat).

Children aged from 1 to 6 years, adhering to a “snack,”⁶⁰ “Western,”⁶⁷ “neotraditional,”⁶⁸ “traditional,”¹⁹ and “traditional Chinese”⁷¹ diet; children aged from 5 to 18 years, adhering to a “Western,”^58,61 “Proteico/Fibra,”⁶² “modern,” or a “traditional north,”⁶⁵ “modern,”⁶⁹ and “sweet”⁷⁰ diet; children from 9 to 13 years, adhering to a “varied Norwegian” or a “dieting”^57,59 diet and adolescents from 11 to 18 years, adhering to a “junk food,”⁶³ “Western,”⁶⁴ and “Westernization”⁶⁶ diet had a significantly higher likelihood of being overweight.

The studies show that, regardless of age, children aged 11 years, and/or adolescents aged 12–19 years, if the subjects had a diet based on obesogenic products (e.g., poultry and eggs, red meat, pasta, whole dairy products, margarine, breads, chocolate milk powder, sweets and desserts, sugar, filled biscuits, fatty foods, pizza and pies, fatty meat, hamburgers and fried foods, processed products, mayonnaise snacks, coke, sprite, coffee, ice cream, instant noodles, and barbecued food), they were at risk of developing obesity.

Lower likelihood of being overweight or obese

From European countries: in Norway “junk/convenience”⁵⁷ (high-fat, high-sugar processed, fast foods such as french fries, processed pizza, processed meat products, sweets, ice cream, and soft drinks) and “varied Norwegian”⁵⁹ (fish and meat, brown bread, regular white or brown cheese, lean meat, fish spread, fruit, and vegetables).

From the Asia countries: in China, “healthy”⁶¹ (milk, yogurt, eggs, fruit, and vegetables), “traditional south,”⁶⁵ and “traditional Chinese”⁶⁹ [vegetables, rice, pork, and legumes rice, red meat, pork, poultry, vegetables (leafy), and fish], “Chinese and Western”⁶⁶ and modern⁷¹ (rice, flour, whole grains, fresh vegetables, fresh fruit, poultry, eggs, freshwater fish and shrimps, sea fish, milk and dairy products, beans and bean products, nuts, snacks, sugar, and barbecued foods). “Western” in Iran⁷⁰ [salty snacks (cheese puffs, chips, and pretzels), beer, fast foods (sausages, burgers, and pizza), and coffee]; “traditional Lebanese” in Lebanon^64,67 (vegetables, legumes, bread, rice, pasta and cereals, bulgur, fruits, fish, and vegetable oils).

From the America countries: “rural” in Mexico,⁵⁸ (corn tortilla, legumes, low intake of sweets and some cereals such as white bread, whole-wheat bread, rice, and noodles); in Brazil, “mixed diet”⁶⁰ (beef and pork, cooked beans, milk and dairy products, rice and tubers, farinaceous foods, sweet and savory cookies, cakes, leafy vegetables, fruits, and natural juices) and “healthy”⁶³ (skimmed dairy products, cereals, leafy vegetables, fruits, and natural juices); “traditional/Almidón” in Colombia⁶² (sugar, honey, rice or pasta, fried foods dry grains, tubers or bananas, eggs, beef, veal, pork, rabbit, goat, coffee or tea, and chicken noodles).

From the Oceania countries: “modern” on the Samoan island of Upolu⁶⁸ (unprocessed red meats, whole grains, which are not traditional to Samoa, corn, french fries, butter, mayonnaise, nuts, and snack foods such as potato chips); “processed” in Australia¹⁹ (snack foods, processed meats, beverages, and confectionary).

Risk of Bias in Individual Studies

The quality assessment of the included studies is presented in Supplementary Appendix Table A3. Of the 16 selected studies, 15 used a cross-sectional design, and 1 was a case–control study.⁷⁰ The selected studies were homogeneous, and all studies presented high quality according to the JBI-MAStARI.

Discussion

The results of this review indicate a positive association between certain eating patterns and the development of being overweight and obesity in children and adolescents in various parts of the world. These findings show that the analysis of dietary patterns and standards classifies foods into groups, taking into account the entire diet, describing eating behaviors⁷² and the interactive effects of foods and nutrients consumed together,²⁷ thus monitoring these dietary patterns among children as a predictor of adult diseases.⁷³

It was decided to synthesize studies that analyzed dietary patterns through PCA, CA, or FA, since these allow the risk of developing diseases to be predicted,⁷⁴ groups of children with consistently healthy or unhealthy diets to be identified, and to provide a perception of the dietary impact on changes in body composition.⁷³

Different studies may give the same name to a dietary pattern, although this does not mean that the food items observed in these patterns are the same. They may vary due to the heterogeneity and specificity of the population (such as geographic, economic, and cultural characteristics), although several similarities can be observed.⁵⁷

For example, the specificity of the ins-and-outs of school dietary patterns analyzed was observed in the study by Vieira et al.,⁴⁵ with 2979 Brazilian children, aged 1 to 6 years, in the southeast region. In this study, the “traditional” pattern within the school was associated with lower income and better nutritional quality (essentially composed of typical Brazilian foods such as rice, beans, high-fiber foods, vitamins, minerals, vegetables, cereals, processed meat, and sausages) and the “traditional” pattern outside the school (similar to the food patterns identified within the school), with a higher intake of saturated and trans fats, sodium, and fiber).

Morris and Northstone⁸² also found differences in the analysis of dietary patterns of 11,868 children 10 years of age in the United Kingdom, where the traditional diet of children living in rural areas was healthier and had less “packed lunch/snack” consumption, whereas the traditional diet of children residing in urban areas presented high consumption of processed foods and “packed lunches/snacks.”

Also in the present review, the dietary patterns identified by the studies (Table 4) were given different names. Some, however, presented similarities in the food groups and, independent of geographical region, indicate that children adhering to eating patterns characterized by frequent ingestion of unhealthy potentially obesogenic foods (including fatty cheeses, sugary drinks, processed foods, fast food, candies, snacks, cakes, animal products, whole milk, and refined grains) are at increased risk of becoming overweight.^{19,57,59–62,64–66,68–70}

Most studies in this systematic review presented results in keeping with the worldwide trend highlighted by the WHO,⁸¹ which reported that currently the greatest frequency is in the consumption of foods that are rich in lipids, refined carbohydrates, bakery products, food of animal origin, and sugar-rich foods, with increasing obesity mainly due to the dependence on soft drinks and sweets, combined with the reduced consumption of fruits and vegetables in many countries.⁷⁵ A study by Kelly et al.⁷⁶ conducted in China, Indonesia, Malaysia, and South Korea attempted to determine the reach of television advertising and marketing of food for children. They found that the products most frequently advertised in all these countries were sugary drinks, with low rates of advertising of vegetables and fruits observed.⁷⁷

Some studies with children in different parts of the world also show this trend of consuming potentially obesogenic foods (baked goods, rich in sugar, such as breads, cakes, and cookies), such as that of Aranceta et al.⁷⁸ with 3534 children and adolescents in Spain, Ritchie et al.⁷⁹ in the United States with 2371 children aged 9 to 10 years, Fernández-Alvira⁷³ with 8341 European children (regions in Italy, Estonia, Cyprus, Belgium, Sweden, Hungary, Germany, and Spain), and Santos et al.⁸³ with 1247 children aged 6 to 12 years in Brazil. Romero-Polvo et al.²⁶ reported an association between the Western diet (high carbohydrates intake, soft drinks, snacks, and corn tortillas, and lower intake of fresh fruits and orange juice), with the risk of insulin resistance in 916 Mexican children and adolescents.

In this study, the dietary patterns that presented the weakest association with the risk of becoming overweight or obese were those in which children consumed more nonobesogenic foods (fruits, vegetables, whole grains, fish, nuts, legumes, and yogurt), that is, food classified as healthy with low levels of sugar and fat.^{59,61,63,64,66,67}

Dietary patterns may also vary depending on the cultural and economic context,⁵² as well as other factors, such as lifestyle and level of physical activity. Alshammari et al.⁴⁴ highlighted that, currently, living standards are on the rise, leading to complex changes in diet, lifestyle, and health.

In this review, some studies showed dietary patterns that are part of the local culture, with a low risk of obesity, such as the “mix-diet,” with foods from Brazilian culture⁶⁰; a rural diet; Mexican food⁵⁸; the diets of traditional Southeast Asia⁶⁵ and Chinese,⁶⁶ with traditional Chinese foods; and various Norwegian traditional foods.⁵⁹

Similar to the present review, several studies have associated dietary patterns with some other patterns. For example, in France, Lioret et al.⁸⁰ reported physical inactivity combined with the consumption of fried snacks as being a predictor of childhood obesity. Magriplis et al.¹⁵ demonstrated that obese children in Greece are those who consume obesogenic foods, sleep less, and spend more hours per day studying. Hebestreit et al.⁴² evaluated the availability of soft drinks during meals in Sweden, Germany, Hungary, Italy, Cyprus, Spain, Belgium, and Estonia and highlighted that parents are important predictors of sweet and fatty food intake in children. Conversely, some studies analyzed in Table 4 show that children who adhere to an obesogenic diet do not present a risk of developing obesity.^{19,57,58,60,62,70}

One of the main strengths of this review was the analysis of studies that analyzed, through PCA, FA, and/or CA, the grouping of dietary patterns in children and adolescents and their associations with overweight and obesity, showing that these statistical methods provide an alternative for analysis of the effect of the diet as a whole instead of isolated nutrients.

Another strength of this review is that it allows the combinations of nutrients and/or foods to be demonstrated through these multivariate methods and the similarities and differences between countries to be highlighted, while simultaneously understanding that obesogenic behaviors are not unique. This premise provides support for future dietary interventions to be tailored to these specific groups, taking into account behavioral and cultural differences.

This review has several limitations. First, it is well known that the use of a cross-sectional design does not reach the confidence level of meta-analyses of randomized clinical trials. Another limitation in the included studies is that the dietary exposures were assessed using self-reports through food frequency questionnaires or 24-hour recall, both relying on memory and are, therefore, susceptible to inherent measurement errors and misclassifications. Finally, the studies included in this review were heterogeneous.

Conclusions

This systematic review suggests that children and adolescents who adhere to dietary patterns composed of obesogenic foods are more likely to develop obesity. Additional confounders, such as parental eating patterns and child physical activity, should be considered.

Footnotes

Acknowledgments

We are grateful to the Post-Doctoral National Program-CAPES (PNPD/CAPES) (Programa Nacional de Pós Doutorado, PNPD-CAPES) that granted Rafaela Liberali a postdoctoral scholarship.

Authors' Contributions

All authors were involved in analyzing the studies, reviewing and interpreting the results, and writing the article; no divergencies exist between any two authors.

Funding Information

This research did not receive any specific grant from funding agencies from the public, commercial, or not-for-profit sectors.

Author Disclosure Statement

No competing financial interests exist.

Supplementary Material

References

Katzmarzyk

, Barlow

, Bouchard

, et al. An evolving scientific basis for the prevention and treatment of pediatric obesity. Int J Obes, 2014; 38:887–905.

Umer

, Kelley

, Cottrell

, et al. Childhood obesity and adult cardiovascular disease risk factors: A systematic review with meta-analysis. BMC Public Health, 2017; 17:683.

The GBD 2013 Obesity

Collaboration

, Ng

, Fleming

, et al. Global, regional and national prevalence of overweight and obesity in children and adults 1980–2013: A systematic analysis. Lancet, 2013; 84:766–781.

WHO. Childhood overweight and obesity on the rise. 2016. Available at www.who.int/dietphysicalactivity/childhood/en Last accessed March 20, 2018.

Estrada

, Eneli

, Hampl

, et al. Children's Hospital Association consensus statements for comorbidities of childhood obesity. Child Obes, 2014; 10:304–317.

Kim

, Sung

, Choi

, et al. Delphi survey for designing a intervention research study on childhood obesity prevention. Korean J Fam Med, 2017; 38:284–290.

Jia

, Li

, Xue

, et al. School environment and policies, child eating behavior and overweight/obesity in urban China: The childhood obesity study in China megacities. Int J Obes, 2017; 41:813.

Goñi

, Arenaza

, Medrano

, et al. Associations between the adherence to the Mediterranean diet and cardiorespiratory fitness with total and central obesity in preschool children: The PREFIT project. Eur J Nutr, 2017; 29:1–9.

Rito

, Sousa

RCD

, Mendes

, Graça

Childhood Obesity Surveillance Initiative: COSI Portugal 2016, 2017. 74p.

10.

Perng

, Fernandez

, Peterson

, et al. Dietary patterns exhibit sex-specific associations with adiposity and metabolic risk in a cross-sectional study in Urban Mexican adolescents. J Nutr, 2017; 147:1977–1985.

11.

Lindsay

, Sitthisongkram

, Greaney

, et al. Non-responsive feeding practices, unhealthy eating behaviors, and risk of child overweight and obesity in Southeast Asia: A Systematic Review. Int J Environ Res Public Health, 2017; 14:436.

12.

Rossi

, Bassett

, Sammán

. Dietary nutritional profile and phenolic compounds consumption in school children of highlands of Argentine Northwest. Food Chem, 2018; 238:111–116.

13.

Yanovski

JA.

Pediatric obesity. An introduction. Appetite, 2015; 93:3–12.

14.

Waters

, Gibbs

, Tadic

, et al. Cluster randomised trial of a school-community child health promotion and obesity prevention intervention: Findings from the evaluation of fun ‘n healthy in Moreland!. BMC Public Health, 2018; 18:92.

15.

Magriplis

, Farajian

, Panagiotakos

, et al. The relationship between behavioural factors, weight status and a dietary pattern in primary school aged children: The GRECO study. Clin Nutr, 2018; 38:1–7.

16.

, Kim

, Joo

, et al. Association of dietary patterns with overweight risk and all-cause mortality in children with cancer. Nutr Res Pract, 2017; 11:492–499.

17.

Johnson

, Toumpakari

, Papadaki

. Social gradients and physical activity trends in an obesogenic dietary pattern: Cross-sectional analysis of the UK National Diet and Nutrition Survey 2008–2014. Nutrients, 2018; 10:388.

18.

Cunha

, Bezerra

, Pereira

, Sichieri

. At-home and away-from-home dietary patterns and BMI z-scores in Brazilian adolescents. Appetite, 2018; 120:374–380.

19.

Shi

, Makrides

, Zhou

. Dietary patterns and obesity in preschool children in Australia: A cross-sectional study. Asia Pac J Clin Nutr, 2018; 27:406–412.

20.

Moher

, Shamseer

, Clarke

, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev, 2015; 4:1.

21.

McNaughton

, Ball

, Mishra

, Crawford

. Dietary patterns of adolescents and risk of obesity and hypertension. J Nutr, 2008; 138:364–370.

22.

Balthazar

, Oliveira

. Differences in dietary pattern between obese and eutrophic children. BMC Res Notes, 2011; 4:567.

23.

del Mar Bibiloni

, Martınez

, et al. Western and Mediterranean dietary patterns among Balearic Islands' adolescents: Socio-economic and lifestyle determinants. Public Health Nutr, 2011; 15:683–692.

24.

Dishchekenian

VRM

, Escrivão

, Meil

, et al. Padrões alimentares de adolescentes obesos e diferentes repercussões metabólicas. [Dietary patterns of obese adolescents and different metabolic effects.] Rev Nutr, 2011; 24:17–29.

25.

Ambrosini

, Emmett

, Northstone

, et al. Identification of a dietary pattern prospectively associated with increased adiposity during childhood and adolescence. Int J Obes, 2012; 36:1299.

26.

Romero-Polvo

, Denova-Gutiérrez

, Rivera-Paredez

, et al. Association between dietary patterns and insulin resistance in Mexican children and adolescents. Ann Nutr Metab, 2012; 61:142–150.

27.

Appannah

, Pot

, O'Sullivan

, et al. The reliability of an adolescent dietary pattern identified using reduced-rank regression: Comparison of a FFQ and 3 d food record. Br J Nutr, 2014; 112:609–615.

28.

Tielemans

, Erler

, Leermakers

ETM

, et al. A priori and a posteriori dietary patterns during pregnancy and gestational weight gain: The Generation R Study. Nutrients, 2015; 7:9383–9399.

29.

Mikkilä

, Vepsäläinen

, Saloheimo

, et al. An international comparison of dietary patterns in 9-11-year-old children. Int J Obes, 2015; 5:S17–S21.

30.

Meguid

, Anwar

, Zaki

, et al. Dietary patterns of children with autism spectrum disorder: A study based in Egypt. Open Access Maced J Med Sci, 2015; 3:262–267.

31.

Gazan

, Béchaux

, Crépet

, et al. Dietary patterns in the French adult population: A study from the second French national cross-sectional dietary survey (INCA2) (2006–2007). Br J Nutr, 2016; 116:300–315.

32.

Asghari

, Yuzbashian

, Mirmiran

, et al. Dietary approaches to stop hypertension (DASH) dietary pattern is associated with reduced incidence of metabolic syndrome in children and adolescents. J Pediatr, 2016; 1:174–184.

33.

Flynn

, Seed

, Patel

, et al. Dietary patterns in obese pregnant women; influence of a behavioral intervention of diet and physical activity in the UPBEAT randomized controlled trial. Int J Behav Nutr Phys Act, 2016; 13:124.

34.

Mohseni-Takalloo

, Hosseini-Esfahani

, Mirmiran

, Azizi

. Associations of pre-defined dietary patterns with obesity associated phenotypes in Tehranian adolescents. Nutrients, 2016; 8:505.

35.

Voortman

, Leermakers

ETM

, Franco

, et al. A priori and a posteriori dietary patterns at the age of 1 year and body composition at the age of 6 years: The Generation R Study. Eur J Epidemiol, 2016; 31:775–783.

36.

Kupek

, Lobo

, Leal

, et al. Dietary patterns associated with overweight and obesity among Brazilian schoolchildren: an approach based on the time-of-day of eating events. Br J Nutr, 2016; 116:1954–1965.

37.

Pérez-Rodrigo

, Gil Á, González-Gross

, et al. Clustering of dietary patterns, lifestyles, and overweight among Spanish children and adolescents in the ANIBES Study. Nutrients, 2016; 8:11.

38.

Chen

, Aris

, Bernard

, et al. Associations of maternal dietary patterns during pregnancy with offspring adiposity from birth until 54 months of age. Nutrients, 2017; 9:2.

39.

Davison

, Saeedi

, Black

, et al. The association between parent diet quality and child dietary patterns in nine- to eleven-year-old children from Dunedin, New Zealand. Nutrients, 2017; 9:483.

40.

Leermakers

, van den Hooven

, Franco

, et al. A priori and a posteriori derived dietary patterns in infancy and cardiometabolic health in childhood: The role of body composition. Clin Nutr, 2018; 37:1589–1595.

41.

Campos

MLB

, Machado

MMT

, Azevedo

, Silva

. Dietary patterns of obese children: Maternal perceptions and experiences. Rev Nutr, 2017; 30:197–207.

42.

Hebestreit

, Intemann

, Siani

, et al. Dietary patterns of European children and their parents in association with family food environment: Results from the I. family study. Nutrients, 2017; 9:126.

43.

Gasga

VMZ

, González

, Piña

GFL

, et al. Dietary patterns, nutritional profile, and body mass index in Mexican schoolchildren: A cross sectional study. Arch Latinoam Nutr, 2017; 67:06–14.

44.

Alshammari

, Suneetha

, Adnan

, et al. Growth profile and its association with nutrient intake and dietary patterns among children and adolescents in Hail region of Saudi Arabia. Biomed Res Int, 2017; 2017:1–9.

45.

Vieira

DAS

, Castro

, Fisberg

. Nutritional quality of dietary patterns of children: Are there differences inside and outside school?. J Pediatr, 2017; 93:47–57.

46.

García-Chávez

, Rodríguez-Ramírez

, Rivera

, et al. Sociodemographic factors are associated with dietary patterns in Mexican schoolchildren. Public Health Nutr, 2018; 21:702–710.

47.

Santos

, Vieira

DAS

, Miranda

AAM

, et al. The traditional lunch pattern is inversely correlated with body mass index in a population-based study in Brazil. BMC Public Health, 2018; 18:33.

48.

Teixeira

, Castro

, Grant

, et al. Dietary patterns are influenced by socio-demographic conditions of women in childbearing age: A cohort study of pregnant women. BMC Public Health, 2018; 18:301.

49.

Pimpin

, Jebb

, Johnson

, et al. Sources and pattern of protein intake and risk of overweight or obesity in young UK twins. Br J Nutr, 2018; 120:820–829.

50.

Muñoz-Hernandez

, Arenaza

, Gracia-Marco

, et al. Influence of physical activity on bone mineral content and density in overweight and obese children with low adherence to the Mediterranean Dietary Pattern. Nutrients, 2018; 10:1075.

51.

Galvan-Portillo

, Sánchez

, Cárdenas-Cárdenas

, et al. Dietary patterns in Mexican children and adolescents: Characterization and relation with socioeconomic and home environment factors. Appetite, 2018; 121:275–284.

52.

Martín-Cañavate

, Sonego

, Sagrado

, et al. Dietary patterns and nutritional status of HIV-infected children and adolescents in El Salvador: A cross-sectional study. PLoS One, 2018; 13:e0196380.

53.

Rashid

, Engberink

, van Eijsden

, et al. Ethnicity and socioeconomic status are related to dietary patterns at age 5 in the Amsterdam born children and their development (ABCD) cohort. BMC Public Health, 2018; 18:115.

54.

National Collaborating Centre for Methods and Tools (2008). Quality Assessment Tool for Quantitative Studies. McMaster University: Hamilton, ON (Updated February 15, 2019). Available at www.nccmt.ca/resources/search/14

55.

Joanna Briggs Institute Critical Appraisal tools for use in JBI Systematic Reviews. Checklist for Analytical Cross Sectional Studies. Available at http://joannabriggs.org/research/critical-appraisal-tools.html Last accessed November 6, 2017.

56.

Joanna Briggs Institute. Joanna Briggs Institute reviewers' manual:. 2014. edition The Systematic Review of Prevalence and Incidence Data. The University of Adelaide: The Joanna Briggs Institute, 2014. Available at https://joannabriggs.org/assets/docs/sumari/ReviewersManual-2014.pdf Last accessed November 6, 2017.

57.

Oellingrath

, Svendsen

, Brantsæter

. Eating patterns and overweight in 9-to 10-year-old children in Telemark County, Norway: A cross-sectional study. Eur J Clin Nutr, 2010; 64:1272.

58.

Rodríguez-Ramírez

, Mundo-Rosas

, García-Guerra

, Shamah-Levy

. Dietary patterns are associated with overweight and obesity in Mexican school-age children. Arch Latinoam Nutr, 2011; 61:270–278.

59.

Oellingrath

, Svendsen

, Brantsæter

. Tracking of eating patterns and overweight-a follow-up study of Norwegian schoolchildren from middle childhood to early adolescence. Nutr J, 2011; 10:106.

60.

Nobre

, Lamounier

, Franceschini

. Padrão alimentar de pré-escolares e fatores associados. [Preschool children dietary patterns and associated factors.]. J Pediatr, 2012; 88:129–136.

61.

Shang

, Li

, Liu

, et al. Dietary pattern and its association with the prevalence of obesity and related cardiometabolic risk factors among Chinese children. PLoS One, 2012; 7:e43183.

62.

Paul Rene

, Gloria Esperanza

, Oscar Fernando

. Patrones de consumo alimentario y exceso de peso infantil; encuesta de la situación nutricional en Colombia, 2010. [Food consumption patterns and excess childhood weight; a survey of the nutritional situation in Colombia, 2010.]. Rev Chil Nutr, 2014; 41:351–359.

63.

Pinho

, Silveira

, Botelho

ACDC

, Caldeira

. Identification of dietary patterns of adolescents attending public schools. J Pediatr, 2014; 90:267–272.

64.

Naja

, Hwalla

, Itani

, et al. A Western dietary pattern is associated with overweight and obesity in a national sample of Lebanese adolescents (13–19 years): A cross-sectional study. Br J Nutr, 2015; 114:1909–1919.

65.

Zhang

, Wang

, et al. Dietary patterns and their associations with childhood obesity in China. Br J Nutr, 2015; 113:1978–1984.

66.

Yang

, Hu

X-M

, Chen

T-J

, Bai

M-J

. Rural-urban differences of dietary patterns, overweight, and bone mineral status in Chinese students. Nutrients, 2016; 8:537.

67.

Nasreddine

, Shatila

, Itani

, et al. A traditional dietary pattern is associated with lower odds of overweight and obesity among preschool children in Lebanon: A cross-sectional study. Eur J Nutr, 2017; 58:1–12.

68.

Choy

, Wang

, Baylin

, et al. Dietary patterns are associated with child, maternal and household-level characteristics and overweight/obesity among young Samoan children. Public Health Nutr, 2018; 21:1–12.

69.

Zhen

, Ma

, Zhao

, et al. Dietary pattern is associated with obesity in Chinese children and adolescents: Data from China Health and Nutrition Survey (CHNS). Nutr J, 2018; 17: 68.

70.

Kelishadi

, Heshmat

, Mansourian

, et al. Association of dietary patterns with continuous metabolic syndrome in children and adolescents; a nationwide propensity score-matched analysis: The CASPIAN-V study. Diabetol Metab Syndr, 2018; 10:52.

71.

Rutayisire

, Wu

, Huang

, et al. Dietary patterns are not associated with overweight and obesity in a sample of 8900 Chinese preschool children from four cities. J Nutr Sci, 2018; 7:e24.

72.

Cespedes

, Hu

. Dietary patterns: From nutritional epidemiologic analysis to national guidelines. Am J Clin Nutr, 2015; 101:899–900.

73.

Fernández-Alvira

, Bammann

, Eiben

, et al. Prospective associations between dietary patterns and body composition changes in European children: The IDEFICS study. Public Health Nutr, 2017; 20:3257–3265.

74.

FB.

Dietary pattern analysis: A new direction in nutritional epidemiology. Curr Opin Lipidol, 2002; 13:3–9.

75.

French

, Story

, Jeffery

. Environmental influences on eating and physical activity. Ann Rev Public Health, 2001; 22:309–335.

76.

Kelly

, Hebden

, King

, et al. Children's exposure to food advertising on free-to-air television: An Asia-Pacific perspective. Health Promot Int, 2014; 31:144–152.

77.

Cairns

, Angus

, Hastings

, Caraher

. Systematic reviews of the evidence on the nature, extent and effects of food marketing to children. A retrospective summary. Appetite, 2013; 62:209–215.

78.

Aranceta

, Perez-Rodrigo

, Ribas

, Serra-Majem

. Sociodemographic and lifestyle determinants of food patterns in Spanish children and adolescents: The enKid study. Eur J Clin Nutr, 2003; 57(S1):S40.

79.

Ritchie

, Spector

, Stevens

, et al. Dietary patterns in adolescence are related to adiposity in young adulthood in black and white females. J Nutr, 2007; 137:399–406.

80.

Lioret

, Touvier

, Lafay

, et al. Dietary and physical activity patterns in French children are related to overweight and socioeconomic status. J Nutr, 2008; 138:101–107.

81.

World Health Organization (WHO). Diet, nutrition and the prevention of chronic diseases. World Health Organ Tech Rep Ser, 2003; 916:i–viii, 1–149.

82.

Morris

, Northstone

. Rurality and dietary patterns: Associations in a UK cohort study of 10-year-old children. Public Health Nutr, 2015; 18:1436–1443.

83.

Santos

NHAD

, Fiaccone

, Barreto

, et al. Association between eating patterns and body mass index in a sample of children and adolescents in Northeastern Brazil. Cad Saude Publica, 2014; 30:2235–2245.

Supplementary Material

Please find the following supplemental material available below.

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

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

0.00 MB

0.09 MB