Society Paper

Role of Dietary Factors, Food Habits, and Lifestyle in

Childhood Obesity Development: A Position Paper From
the European Society for Paediatric Gastroenterology,
Hepatology and Nutrition Committee on Nutrition
*†
Elvira Verduci, ‡Jiri Bronsky, , §Nicholas Embleton, ||Konstantinos Gerasimidis, ¶Flavia Indrio,
#
Jutta Köglmeier, **Barbara de Koning, ††‡‡Alexandre Lapillonne, §§Sissel Jennifer Moltu,
||||
Lorenzo Norsa, and ¶¶Magnus Domellöf ESPGHAN Committee on Nutrition

ABSTRACT

Childhood obesity has high societal and economic impact but current treat- What Is Known
ment approaches are sub-optimal. In the last decade, important studies
have been conducted aiming to identify strategies to prevent obesity dur- • Childhood obesity is the most prevalent food-based
ing critical periods of life. Updated recommendations for childhood obe- disorder among children and adolescents worldwide.
sity prevention are needed. We present data from systematic reviews and • Although national and international surveys report
meta- analysis, randomised controlled trials (RCTs) and large observational a levelling-off of the prevalence of obesity in some
studies, published from 2011 onwards that consider the possible role of countries, the burden of pediatric obesity is still
the following factors in obesity development: breast-feeding; macronutri- high, not least in Mediterranean area countries.
ent composition and method of complementary feeding; parenting style; • The aetiology of obesity is complex, with several risk
dietary patterns; sugar-sweetened beverage consumption; eating behaviour factors and mechanisms that are interconnected.
(eg, skipping breakfast, family dinners. etc); meal frequency and composi-
tion (fast foods, snacking), portion size; dietary modulators of gut micro-
What Is New
biota (including pre-, pro-, and synbiotics); physical activity and sedentary
behaviour. We used the Medline database and the Cochrane Library to
search for relevant publications. Important research gaps were also identi-
• The European Society for Paediatric Gastroenterol-
fied. This position paper provides recommendations on dietary factors, food
ogy, Hepatology, and Nutrition Committee on Nutri-
habits, and lifestyle to prevent childhood obesity development, based on the
tion provides an update of the previous position
available literature and expert opinion. Clinical research and high-quality
paper (2011) that also evaluates the role of dietary
trials are urgently needed to resolve numerous areas of uncertainty.
patterns and food habits in the prevention of obesity
in children and adolescents ages 2 to 18years. Fac-
Keywords: childhood obesity, diet macronutrient intakes, early nutrition, tors in infants and children <2years of life, as well as
parenting style, prevention the role of physical activity and sedentary behaviour
will also be considered in this updated version.
(JPGN 2021;72: 769–783) • A logic model was developed to demonstrate aspects
of the complex system underlying the development
of child and adolescent obesity.
INTRODUCTION • Updated recommendations are provided to prevent
Obesity is a global public health problem associated with a childhood obesity.
wide range of metabolic abnormalities and a negative impact on the
mental health of individuals. By 2050, obesity is predicted to affect
60% of adult men, 50% of adult women, and 25% of children if
present trends continue (1). The World Health Organization (WHO) challenges for the 21st century (2,3) with a significant burden to soci-
regards childhood obesity as one of the most serious global health ety. Obese children have increased metabolic and cardiovascular risks

Received October 16, 2020; accepted January 17, 2021.

From the *Department of Health Sciences, †Department of Pediatrics, pital for Children NHS Foundation Trust, London, United Kingdom,
Vittore Buzzi Children's’ Hospital-University of Milan, Italy, ‡Depart- **Paediatric Gastroenterology, Erasmus MC–Sophia Children's Hospi-
ment of Paediatrics, University Hospital Motol, Prague, Czech Repub- tal, Rotterdam, the Netherlands, ††Paris Descartes University, APHP
lic, §Newcastle Neonatal Service, Newcastle Hospitals NHS Trust and Necker-Enfants Malades hospital, Paris, France, ‡‡CNRC, Baylor Col-
Newcastle University, Newcastle upon Tyne, ||Human Nutrition, School lege of Medicine, Houston, TX, §§Department of Neonatal Intensive
of Medicine, Dentistry and Nursing, University of Glasgow, New Lister Care, Oslo University Hospital, Norway, ||||Pediatric Hepatology Gas-
Building, Glasgow Royal Infirmary, Glasgow, United Kingdom, ¶Dipar- troenterology and Transplantation, ASST Papa Giovanni XXIIII, Ber-
timento di Scienze Mediche e Chirurgiche, University of Foggia, Italy, gamo, Italy, and ¶¶Department of Clinical Sciences, Pediatrics, Umeå
#Department of paediatric Gastroenterology, Great Ormond Street Hos- University, Umeå, Sweden.

JPGN • Volume 72, Number 5, May 2021 769

Copyright © ESPGHAN and NASPGHAN. All rights reserved.
Verduci et al JPGN • Volume 72, Number 5, May 2021

both in childhood and adulthood (4), and may show early signs of (Appendix 1, http://links.lww.com/MPG/C237). Only manuscripts
the metabolic syndrome, such as dyslipidaemia, hypertension, and published in English were evaluated as full papers.
disorders of glucose metabolism (5,6). Current treatment approaches
are sub-optimal both in adults and children (4). In the last decade,
important studies have been conducted aiming to identify strategies to FIRST 2 YEARS OF LIFE
prevent obesity during critical periods of life. The underlying aetiol-
ogy of obesity is complex, with several risk factors and mechanisms Breast-feeding
that are interconnected. The weight excess can manifest across the life Breastfed (BF) and formula-fed (FF) infants have different
cycle: mothers who are overweight or obese at the time of conception body composition, measured using a variety of methods (11), with
can transmit effects to their offspring creating intergenerational cycles FF infants showing significantly lower fat mass (FM) compared with
of obesity (7). The topic of genetic and environmental influences on BF infants in the first 6 months of life (−0.09 kg [−0.18; −0.01 kg] at
obesity, and how they interact, is a unique topic for which conceptual 3 to 4 months and −0.18 kg [−0.01; −0.34 kg] at 6 months), although
frameworks are scarce. Also, the different economical settings have to the trend toward higher FM in FF infants, observed in the second
be considered on the major risk of pediatric obesity (8,9). 6 months of life, was not significant at 12 months. A similar effect
The focus of this article is to update and expand on the previ- for FM% was also documented (11). It is important to note that
ous European Society for Paediatric Gastroenterology, Hepatology studies included in this meta-analysis showed marked heterogene-
and Nutrition Committee on Nutrition (ESPGHAN CoN) position ity with respect to study design, techniques used to measure body
paper published in 2011 (10) that evaluated the role of dietary fac- composition, and measurement timepoints.
tors and food habits in the prevention of obesity in children and Two meta-analyses (12,13) and a review of systematic
adolescents. Factors in infants and children <2 years of life, as well reviews (14) published data on the association of breast-feeding
as the role of physical activity and sedentary behaviour will also be and childhood obesity prevention. The most recent meta-analysis
considered in this updated version. The maternal factors and expo- by Horta et al (12), including studies from high-income and low-/
sures during pregnancy and the dietary management of obesity is middle-income countries, showed a preventive effect of BF on later
beyond the scope of this article. overweight/obesity with tight 95% confidence interval (CI) (pooled
odds ratio [OR]: 0.74; 95% CI: 0.70--0.78). The meta-analysis by
Yan et al (13) also reported a possible preventive effect of BF. In the
METHODS meta-analysis by Yan et al (13), however, the majority (22/25) of
In the present article, we have developed a logic model to studies were from high-income countries, the definition of obesity
highlight some key aspects of the complex system of child and ado- was very heterogeneous, and moreover some studies also included
lescent obesity development (Fig. 1). individuals who were overweight but not obese.
We present results from systematic reviews and meta-anal- The review by Stanley did not show a clear effect, perhaps
ysis, randomised controlled trials (RCTs) and large observational indicating a risk of residual confounding (14).
studies, published from 2011 onwards, on the possible role of the A more recent systematic review evaluated the association
following factors in obesity development: breast-feeding; macronu- of exclusive and partial breast-feeding duration with growth during
trient composition and method of complementary feeding; parental infancy (15) and demonstrated that shorter duration of breast-feed-
style; dietary patterns; sugar-sweetened beverages (SSBs) consump- ing (4 vs 6 months), especially if exclusive, tended to be associated
tion; eating behaviour (eg, skipping breakfast, family dinners, etc); with faster weight gain in high-income country settings, during
meal frequency and composition (fast foods, snacking), portion the first year of life (16,17). It is important to note that an individ-
size; dietary modulators of gut microbiota (including pre-, pro-, and ual-level meta-analysis found that each +1 unit increase in weight
synbiotics); physical activity; and sedentary behaviour (see Table 1). standard deviation (SD) scores between 0 and 1 year conferred a
In order to retrieve references specifically related to children 2-fold higher risk of childhood obesity and a 23% higher risk of
and relevant to the scope of this position paper, the database Med- adult obesity, after adjustment for sex, age, and birth weight (17).
line (via PubMed) and Cochrane Database of Systematic Reviews The systematic review and meta-analysis by Giugliani et al
were searched for keywords of publications up to March 2020 (18) included BF promotion intervention studies. Of 35 studies,

Address correspondence and reprint requests to Elvira Verduci, Department life Nutrition. He also served as member of Advisory board for Danone
of Health Sciences, University of Milan, Milan, Italy (e-mail: elvira. Early life Nutrition and received payment/honorarium for lectures from
verduci@unimi.it). Danone Early life Nutrition, Nestle Nutrition Institute, Baxter, and
Supplemental digital content is available for this article. Direct URL cita- Fresenius Kabi. K.G. reports receipt of research grants, speakers and
tions appear in the printed text, and links to the digital files are provided consultancy fees, and hospitality from Nestle Health Sciences, Nutri-
in the HTML text of this article on the journal’s Web site (www.jpgn. cia-Danone, Baxter, Mylan, DrFalk, and Abbott. F.I. reports receipt of
org). payment/honorarium for lectures from Biogaia, Nestle, Danone, Abbot,
Chair of CoN: Magnus Domellöf; Secretary of CoN: Jiri Bronsky. and consultancy fees from Biogaia. A.L. reports receipt of lecture fees
Disclaimer: ESPGHAN is not responsible for the practices of physicians and/or non-financial support from Baxter, Fresenius, Nestle, and Nead
and provides guidelines and position papers as indicators of best prac- Johnson Nutrition. S.J.M. reports receipt of research support from DSM
tice only. Diagnosis and treatment is at the discretion of physicians or Nutritional Products and payment/honorarium for lectures from Baxter.
the healthcare provider. M.D. reports a research grant from Baxter and speaker fees from Sem-
The authors report no conflicts of interest relevant to this article but many per, Baxter, Nutricia, and Abbvie. The remaining authors (J.K., Bd.K.,
of the authors have received research funding or honorarium for lectures and L.N.) report no conflict of interests.
from pharmaceutical or infant food/milk companies. E.V. reports grant/ Copyright © 2021 by European Society for European Society for Pediatric
research support from Nutricia Italia Spa, Nestle Health Science—Vita- Gastroenterology, Hepatology, and Nutrition and North American Soci-
flo Italy, FoodAR srl Italy, PIAM Pharma, and Integrative Care. J.B. ety for Pediatric Gastroenterology, Hepatology, and Nutrition. This is an
reports personal fees and non-financial support from AbbVie, Nutricia, open access article distributed under the Creative Commons Attribution
Biocodex and personal fees from MSD, Nestlé, Ferring, Walmark. N.E. License 4.0 (CCBY), which permits unrestricted use, distribution, and
reports receipt of grants/research supports from National Institutes for reproduction in any medium, provided the original work is properly cited.
Health Research (UK), Prolacta Bioscience (US), and Danone Early DOI: 10.1097/MPG.0000000000003075

770 www.jpgn.org

Copyright © ESPGHAN and NASPGHAN. All rights reserved.

JPGN • Volume 72, Number 5, May 2021 Role of Dietary Factors, Food Habits, and Lifestyle in Childhood Obesity

CONTEXT INPUT OUTPUT OUTCOMES

• Maternal nutrional &

metabolic status (obesity &
diabetes)

• Intrauterin growth Infant

Pre-pregnancy • Preterm/term delivery Lean & Fat mass deposion Weight excess
(body composion)
• Mode of delivery

• Socio-economic status

• Maternal smoking & alcohol

• Anbiocs
Pregnancy and delivery
• Body weight at birth

• Macro-nutrients intake Child

Growth & BMI trajectory
• Type of milk feeding Weight excess / obesity
• Complementary feeding Metabolic syndrome
(ming and quality)
Early life
• Dietary pa ern
(0 - 2 years of age)
• Poron size

• Meal composion

• Meal frequency

• Modulators of gut microbiota

Time of adiposity rebound Adolescents
Aer 2 years of age • Eang behavior Obesity
onwards
• Parental behavior Metabolic syndrome
• Physical acvity

• Sedentary behaviour
Genec & Epigenec
• Sleeping Factors
DEI = MF * PS * mED

FIGURE 1. Logic model proposed to examine the complex systems of the child and adolescent obesity development. DEI = Daily Energy
Intake; mED = meal's Energy Density; MF = Meal Frequency; PS = Portion Size.

16 provided sufficient growth data and were included in the meta- overweight and obesity and recommended limiting the protein
analysis. A modest but significant decrease in mean body mass intake to 15% of total energy for infants and toddlers (20). After this
index (BMI) (or weight for length/height) z score (mean difference publication, Appleton et al (22) published a systematic review exam-
−0.06 [95% CI: −0.12 to 0.00]) in low- and high-income countries ining the association between formula feeding practice and excess
was observed, although this effect was not seen in low-medium or rapid weight gain. Where infants are not receiving breastmilk,
income countries. Therefore, intervention studies were inconclu- using a formula with a lower protein content (1.25 g/100 mL standard
sive, and more studies are needed. formula, then 1.6 g/100 mL follow on) from birth might reduce the
Moreover, it should be noted that breast-feeding rates in the risk of rapid or excess weight gain and the childhood risk of being
United States differ significantly depending upon the income of the overweight or obese (23,24). The authors, however, highlight that
mother (19), suggesting that economically poor settings could be the review only included 2 RCTs (23–25) (total n = 1262 formula-
early negative marker of the pediatric obesity risk, combining with fed infants), with the Inostroza study (25) including infants born
low breast-feeding rates. from overweight or obese mothers and using a low-protein formula
(1.04 g/100 mL) also containing probiotics. On the other hand, prob-
ably the diet's protein content is not only the variant to be considered
Complementary Feeding (26) but also the possible role of the early fat/protein balance should
Macronutrient Intake be taken into account as the role of excess energy intakes in case of
According to the ESPGHAN Position Paper on complemen- formula-fed infants, irrespective of the macronutrient balance (27).
tary feeding (CF) (20), complementary foods (solids and liquids A recent systematic review and meta-analysis of cross-
other than breast milk or infant formula) should not be introduced sectional and prospective cohort studies evaluated the relation-
before 4 months but should not be delayed beyond 6 months. In ship between regular cow-milk fat consumption (defined as daily
2019, the EFSA Panel on Nutrition, Novel Foods and Food Aller- or ≥4 times per week) and adiposity in healthy children ages 1
gens concluded that there was no effect of introduction of CFs at 3 to 18 years (28). Among children who tended to consume whole-
to 4 months of age, compared with 6 months of age, on body weight, fat (3.25% fat) compared with reduced-fat (0.1–2%) milk, the
body length, head circumference, BMI, and body composition (21). adjusted OR of overweight or obesity was 0.61 (95% CI: 0.52--
It should be noted, however that the limit of 3 months would have a 0.72; P < 0.0001). It, however, seems likely that some of the asso-
negative impact on breast-feeding in Europe. ciation might be because of confounders or because of reverse
In the ESPGHAN Position Paper, the CoN concluded that causation, for example, because of parents of overweight children
a high protein intake during CF may increase the risk of later choosing lower fat milks.

www.jpgn.org 771

Copyright © ESPGHAN and NASPGHAN. All rights reserved.

Verduci et al JPGN • Volume 72, Number 5, May 2021

TABLE 1. Flowchart with eligibility criteria for inclusion in this Position Paper

 

Limited evidence is present on the effect of young child for- better development of child self-regulation. It should be noted that
mula on health outcomes in toddlers (29). the present study has been carried out in white, middle-income,
predominantly well-educated mothers, and attrition could be more
Method of Feeding likely to occur among higher risk participants.
A modified version of baby-led weaning called Baby Led
Introduction to SolidS (BLISS study) was developed to avoid
energy and iron deficiency, and the choking hazard potentially FROM 2 YEARS OF LIFE ONWARDS
associated with baby-led weaning, and was compared with the
traditional spoon-feeding approach in a RCT (30). The mean BMI
Dietary Patterns
z score was not significantly different at 24 months between the 2 Mediterranean Diet
groups, although a lower satiety responsiveness and less parent The Mediterranean diet (MD) is a dietary pattern rich in
food fussiness was reported in BLISS infants. Also, a more recent plant-based foods (vegetables, fruits, whole grain cereals, legumes,
RCT (31) failed to show any influence of method of feeding on nuts, seeds), moderate-to-high intake of fish and seafood, moderate
overweight and obesity at 2 years of age. consumption of eggs, poultry, and dairy products (milk, yoghurt,
and cheese), and low consumption of red meat. Olive oil, rich in
Parenting Style unsaturated n-9 fatty acids, is the main source of added fat (Fig. 2).
A recent single-centre RCT evaluated the effect of a respon- MD adherence varies widely within the Mediterranean countries
sive parenting intervention among 279 mother-child dyads on BMI for both children and adolescents, and it can be evaluated through
z score at 3 years (32). Full-term singleton newborns delivered to quantifying scores or indexes, such as KIDMED, MDS, fMDS, or
primiparous mothers ≥20 years and with a birthweight >2500 g MediLIFE—questionnaires that vary in the number of category
were recruited. The intervention consisted of parental advice on classification systems, components, questions’ content, scoring
their child's behavioural states, focusing on feeding, sleep, interac- system, and contribution (positive or negative) of a single compo-
tive play, and emotional regulation. The control group received an nent to the total score—that can be self-administered or conducted
intervention on safety that was dose-matched to ensure equivalent by interview led by a paediatrician, dietitian, and so forth.
time and intensity. The children in the responsive parenting group Two systematic reviews (33,34) aimed to summarise the
showed a lower mean BMI z score at 3 years compared with controls available literature on the dietary intake of European children and
(absolute difference, −0.28 [95% CI: −0.53 to −0.01]; P = 0.04). Fur- adolescents and investigate the association between a Mediterra-
ther studies with sufficient sample size are needed in order to deter- nean-like diet, MD adherence, and nutritional status. Pereira-da-
mine whether this method is effective for the prevention of obesity. Silva et al (33) focused more specifically on preschool children
The responsive parenting intervention used in this study may affect in the Mediterranean countries of the European Union. A lower
multiple domains of child behaviour and could be associated with adherence to a Mediterranean-like dietary pattern, evaluated by

772 www.jpgn.org

Copyright © ESPGHAN and NASPGHAN. All rights reserved.

JPGN • Volume 72, Number 5, May 2021 Role of Dietary Factors, Food Habits, and Lifestyle in Childhood Obesity

FIGURE 2. Visual representation of the Mediterranean diet composition.

KIDMED index, was associated with a higher prevalence of being MD, the ND mainly focuses on local, organic, and wild food items.
overweight and obese, and greater maternal education and famil- Indeed, the main difference of ND compared with other dietary pat-
ial socioeconomic status was associated with better quality diets. terns is related to the recommended type of oil, vegetables, and
Iaccarino Idelson et al (34) found that the association between MD fruit.
adherence (assessed with 3 different indices: KIDMED, MDS, There are very few attempts to explore associations between
fMDS) and weight status was not consistent in 2- to 20-year-old ND adherence and obesity in the pediatric population. Besharat
children and adults even if a better diet quality was observed. Nev- Pour et al (36) estimated the association between parental migra-
ertheless, a positive association was found with physical activity tion background, nutrition, physical activity, and weight in 8-year-
and a negative association with sedentary behaviour. old children born in Stockholm, including offspring of immigrants.
More recently, a Greek cross-sectional study (35) involv- Using a Food Frequency Questionnaire, ND adherence was assessed
ing ∼170,000 children and adolescents (6–18 years) published data based on the compliance of nutrients intake as stated in the “Nordic
on the association between MediLIFE Index scores and anthropo- Nutrition Recommendations of 2004” guidelines (37). Offspring
metric measurements. This study considered 4 components (the of immigrants complied more fully with nutritional recommenda-
KIDMED index, physical activity level, sedentary time, and sleep tions but had a higher risk of having low physical activity and hence
duration), and anthropometric measurements. Higher MediLIFE being overweight compared with children of Swedish origin (36).
scores were associated with lower BMI, waist circumference and An intervention study (38) investigated the effects of intro-
waist-to-height ratio (WtHR), and lower prevalence of overweight, ducing hot school-meals, following the principles of the ND, instead
obesity, and abdominal obesity, by 6% (OR 0.94; 95% CI: 0.92– of packed lunch (traditionally based on cold sandwiches) in a cohort
0.98), 30% (OR 0.70; 95% CI: 0.67–0.75), and 20% (OR 0.80; 95% of Danish children ages 8 to 11 years. ND adherence was established
CI: 0.77–0.83), respectively. It is not, however, possible to draw using a self-administered internet-based interactive food record tool.
any conclusion from the present study, regarding the role of MD No difference was found in the average daily energy intake but bev-
adherence on the risk of overweight and obesity as an index based erage intake was lower. Moreover, a lower energy intake from fat
on different factors was used. and a higher energy intake from protein, vitamin D and iodine was
observed (38), although the effect on obesity was not evaluated.
Nordic Diet
The Nordic diet (ND) is a dietary pattern, which refers to a Vegetarian Diet
modern dietary profile commonly available in the Nordic regions Two main types of vegetarian diet are recognized (39)
and acknowledged worldwide only in the last few years. ND is high (Fig. 4).
in fruits and vegetables (especially berries, cabbages, root vegeta-
bles, and legumes), plants and mushrooms collected in the wild, Lacto-ovo-vegetarianism
fresh herbs, potatoes, nuts, whole grains (mainly barley, rye, and Lacto-ovo-vegetarianism (LOV) excludes meat but includes
oats), rapeseed oil, fatty fish (especially salmon, herring, and mack- dairy products, eggs, and honey, together with a wide variety of
erel) and shellfish, seaweed, low-fat and white meat, game, and also plant foods. Subcategories are lacto-vegetarianism (LV), which
emphasises the consumption of low-fat dairy products and avoid- excludes eggs, and ovo-vegetarianism (OV), which excludes dairy
ance of sugar-sweetened products (Fig. 3). Additionally, similar to products.

www.jpgn.org 773

Copyright © ESPGHAN and NASPGHAN. All rights reserved.

Verduci et al JPGN • Volume 72, Number 5, May 2021

FIGURE 3. Visual representation of the Nordic diet composition.

FIGURE 4. Visual representation of the vegetarian diet composition.

Veganism VEG in certain countries (2016–2019 data). The exact number of

Veganism (VEG) excludes meat, dairy products, eggs, and vegetarian children is not known but it is likely that many vegetarian
honey. parents would raise children following a similar dietary pattern.
Vegetarians were estimated in 2018 to represent about 8% of There is limited literature that explores the effect of veg-
the world population (40). Table 2 reports the prevalence of LOV and etarian diets in children on the risk of overweight and obesity.

774 www.jpgn.org

Copyright © ESPGHAN and NASPGHAN. All rights reserved.

JPGN • Volume 72, Number 5, May 2021 Role of Dietary Factors, Food Habits, and Lifestyle in Childhood Obesity

A recent systematic review (50) of prospective cohort stud-

TABLE 2. Prevalence of lacto-ovo-vegetarianism and veganism in
ies and RCTs found a positive association between consumption
some worldwide countries (2016--2019 data)
of SSBs and weight in children. In particular, 16/17 prospective
Country LOV, % VEG, % cohort studies, none of which were funded by industry, showed a
positive association of SSBs on obesity, and 3 RCTs demonstrated
India* 31
that SSB consumption had an effect on BMI or BMI z score. The
USA† 5–8 3 RCTs explored the effect of 2 nutritional education programs at
Germany‡ 10 1.6 school (1 giving general health and advice on healthy eating and
Italy§ 5.4 1.9 1 focused on reducing SSBs consumption by encouraging water),
and 1 healthy lifestyle education program. Two out of 3 of these
Spain¶ 1.3 0.2
trials adjusted for physical activity but no study adjusted for dietary
UK∥ 7 1.16 energy intake.
France# 5 0.25 A systematic review of 27 intervention studies aiming to
Greece** 2 0.8 reduce SSBs consumption in children ages 0 to 5 years showed that
the interventions conducted in preschool/daycare settings, specifi-
Switzerland†† 14 3 cally targeting only SSBs or only oral hygiene, were able to reduce
LOV = lacto-ovo-vegetarianism; VEG = veganism. the SSB consumption (51). The variation in study characteristics,
*http://censusindia.gov.in/vital_statistics/BASELINE%20 design, and reporting of results, however, make it difficult to com-
TABLES07062016.pdf. pare effectiveness of strategies across studies.
†https://news.gallup.com/poll/238328/snapshot-few-americans- Fruit juices (100% fruit part) are not considered SSBs. No
vegetarian-vegan.aspx. evidence was found to support a positive association between fruit
‡https://proveg.com/de/confirmation/veggie-fakten-nicht-mehr-
juice consumption and weight gain (52,53), and there is currently
verfuegbar/.
§https://eurispes.eu/rapporto-italia-2019-vegetariani-e-vegani-le- a lack of RCTs on this topic. Even if fruit juices tend to have a
nuove-diete-si-consolidano/. superior nutritional composition compared with SSBs containing
¶https://web.archive.org/web/20170625180431/http://www.lantern.es/ minerals and vitamins, the amount of free sugar and energy is, how-
wp-content/uploads/2017/02/infog.jpg. ever, similar to those of SSBs, which may have similar effects on
∥https://www.plantbasednews.org/news/veganism-skyrockets-to-7-of- weight gain (49).
uk-population-says-new-survey.
#https://web.archive.org/web/20180703120048/https://o.nouvelobs. Dietary Modulation of Gut Microbiota
com/food/20180703.OBS9086/en-france-le-vegetarien-est-plutot-une-
Gut microbiota play an important role in the absorption, stor-
femme-trentenaire.htmL.
**https://www.protothema.gr/ugeia/article/870879/go-vegan-80000- age, and utilization of energy obtained from diet. Furthermore, the
atoma-aspazodai-ti-diatrofi-stin-ellada/. gut microbiota is also involved in the regulation of food intake by
††https://www.swissveg.ch/veggie_survey?language=en. affecting hormones that influence metabolic function and specific
brain areas associated with eating behaviour (54). This so-called
“gut microbiota-brain axis” represents a bidirectional signalling
The limited evidence available indicates that the growth of LOV axis that may contribute to body weight by influencing appetite,
children (41,42) and adolescents (41,43–45) is comparable to that storage, and energy expenditure (55,56).
of their omnivoric (OMN) peers. Furthermore, data suggest that One of the key activities of the gut microbiota is through the
VEG children tend to grow in a similar pattern to non-VEG chil- harvesting of energy for the host through fermentation of otherwise
dren (41,46). indigestible nutrients, including oligo- and polysaccharides, and
A more recent study (47) conducted on 215 healthy adoles- production of short-chain fatty acids (SCFAs), that is, acetate, pro-
cents, attending 5 Adventist secondary schools in Australia, showed pionate, and butyrate. Among their multiple actions, butyrate acts
that students consuming predominantly vegetarian foods had a sig- as an energy substrate for colonocytes, acetate contributes to de-
nificantly lower BMI. In this study, adolescents were classified as novo lipogenesis, and propionate is metabolized in the liver, regu-
vegetarians if they consumed red meat, chicken, and fish less than lating production of appetite hormones and cholesterol metabolism
once a week. It is important to note that no adjustment for con- (57).
founders was performed and the classification used for vegetarians The causal relation between gut microbiota composi-
and nonvegetarian adolescents does not enable an exploration of the tion and energy homeostasis is, however, complex and is largely
effects of vegetarian status following either a LOV or a VEG diet. based on preclinical research and/or association studies. Multiple
Inadequate intake of energy, protein, calcium, zinc, iron, other factors are important in the aetiology of obesity, including
vitamin B12, and vitamin D may occur on a vegetarian diet because genetic, epigenetic, and gene-nutrient interactions, which may also
of a limited variety and sub-optimal choice of foods (39,48). There- be important modifiers of gut microbiota structure and function
fore, whenever a vegetarian diet is used for children, appropriate (58,59). Also, the maternal pre-pregnancy BMI may have an impact
nutritional planning and monitoring is recommended to be super- on infants’ gut microbiome (60).
vised by an adequately trained health care professional. Childhood obesity, as in adults, has been associated with
an increased Firmicutes/Bacteroidetes ratio in faeces and reduced
Sugar-sweetened Beverages microbial diversity and richness in the gastrointestinal tract com-
SSBs are beverages or drinks that contain added caloric sweet- pared with normal-weight, although it is difficult to determine the
eners (ie, sucrose, high-fructose corn syrup, fruit juice concentrates). effect of residual confounding or reverse causation (61,62).
The ESPGHAN CoN recently published a position paper (49) that Recently a 4-year prospective study of 70 school-age chil-
evaluated outcomes related to the intake of sugar in infants, children, dren, evaluated the association between clusters of dietary habits
and adolescents, and provide recommendations and concluded that a and gut microbiota diversity and whether the interaction of micro-
higher than recommended intake of free sugars (ie, mono and disac- biota-host-diet may predict obesity (63). Out of 70 normal-weight
charides), particularly SSBs in children and adolescents, is associ- children at start, 34 remained normal-weight and 36 became obese.
ated with an increased risk of excess weight gain. The combination of “high carbohydrate/high fat” or “high protein/

www.jpgn.org 775

Copyright © ESPGHAN and NASPGHAN. All rights reserved.

Verduci et al JPGN • Volume 72, Number 5, May 2021

high fat” dietary pattern and low diversity of microbiome was asso- than a glass of milk or fruit juice). The 2 categories “daily break-
ciated with the onset of obesity. Furthermore, the study suggests fast consumption” (7 days in a week) and “less than daily” (less
that the individual gut microbiome configuration and long-term than 7 days in a week) were considered. Authors found that girls
dietary habits together can be considered as a predictive tool for the more often skipped breakfast than boys (“less than daily”: 55.9%
development of obesity in children (63). vs 48.6%). Among all age groups and both in girls and boys, “less
Considering that lifestyle changes are difficult to implement than daily” breakfast consumption was associated with overweight
over the longer term, research efforts have also explored alternative (OR = 1.33, 95% CI: 1.16–1.51 in boys; OR = 1.58, 95% CI: 1.38–
strategies, such as probiotics and/or prebiotics in order to modulate 1.82 in girls). Summarizing, no daily breakfast consumption was
gut microbiota and prevent obesity. associated with overweight including obesity (68).
Luoto et al (64) conducted an RCT in 159 women of Lac- The positive association between skipping breakfast and
tobacillus rhamnosus GG supplementation (1 × 1010 CFU) com- weight status, reported in these studies, may nevertheless reflect
menced 4 weeks before expected delivery combined with treatment “reverse causality,” considering that children with greater body
of the child during the first 6 months of life and showed slower mass may eat less frequently.
weight gain in children at 1 and 4 years of age in a secondary post
hoc analysis, although no evidence of long-term effects (10 years) Family Dinner
was found. It is, however, important to note that the primary aim In a systematic review conducted by Valdés et al (69), fre-
of the study was to investigate the preventive effect of probiotics quent family meals (and therefore, dinner) have been associated
on the onset of allergic diseases, and therefore, multiple different with a healthier and more varied dietary pattern. Six out of 11
confounding factors that may influence the obesity risk may have cross-sectional studies and 1 out of 4 longitudinal studies found
been present. A more recent RCT (65) randomized 179 term-born, statistically significant inverse associations between frequent fam-
vaginally delivered infants to receive cereals without or with probi- ily dinner and being overweight. Most of the cross-sectional studies
otic (Lactobacillus paracasei ssp F19 [1 × 108 CFU]) between 4 and showed this inverse association was more consistent among chil-
13 months of whom 120/179 were followed until 8 to 9 years. No dren than adolescents. These results also showed how the potential
significant differences were found in BMI z score and body compo- protective effect of family meals may be limited to younger children
sition, measured using dual energy X-ray absorptiometry. (4–7 years old).
We found no published studies on the effect of prebiotics or Valdés et al (69), however, noted that 1 of the 3 longitudi-
synbiotics on prevention of childhood obesity. nal studies found an inverse association between frequent family
dinner and overweight that approached significance among mid-
dle-school girls. This potential association could reflect the greater
Eating Behaviour incidence amongst girls, compared with boys, of having an eating
Skipping Breakfast disorder, such as binge eating and dieting. Irrespective of the find-
Skipping breakfast, defined as not eating between 06:00 and ings, however, all studies suffered from 2 major limitations: the lack
09:00 a.m., has always been considered a risk factor for obesity as of a standard definition and characteristics of family meal.
it is believed to have a critical role in energy balance and dietary More recently, a sub-study of a project done by Roos et al
regulation. (70) among 11-year-old European children, found that having family
A recent systematic review (66) of observational studies meals (mainly breakfast and dinner) and TV viewing during dinner
aimed to summarise the association of skipping breakfast with was not associated with overweight. When these associations were,
body weight and metabolic outcomes in children. The total sample however, stratified by region, results showed that in Northern Europe,
included 286,804 participants (2–18 years) living in 33 countries. children who had family breakfast or dinner less than once weekly
Data were appeared consistent perhaps as the studies represented and TV viewing during the dinner were more likely to be overweight,
children living in multiple different locations; however, definition while there was no association between family breakfast or din-
and assessment of overweight/obesity and of skipping breakfast ner and adiposity status (according to BMI) in Southern and East-
were highly heterogeneous. Regardless of this, most of the studies ern European countries (70). These country discrepancies may be
reported that at least 10% to 30% of children and adolescents never explained by different dietary patterns and lifestyles across Europe.
ate breakfast, and there was an increasing trend in skipping break- In 890 young Japanese adolescents, Shirasawa et al (71)
fast from childhood to adolescence, as well as reporting higher val- observed no difference in terms of eating dinner alone between
ues in girls than in boys. Overall, studies representing around 94% overweight and nonoverweight boys. Compared with girls not eat-
of all subjects reported a positive association between skipping ing dinner alone, girls who ate dinner alone ≥1 time/week showed
breakfast and obesity. In conclusion, this review supports that skip- an increased risk of overweight (adjusted OR = 2.78; 95% CI:
ping breakfast may be a useful predictor of the risk of overweight/ 1.21–6.38).
obesity, even if the mechanism of weight excess may be because of Haghighatdoost et al (72) studied a sample of 5528 Iranian
a higher energy intake during the following hours in children who adolescents and showed that no significant differences were found
skipped breakfast (66). in dietary intake between family dinner “consumers” (≥5 times/
Traub et al (67) collected data about skipping breakfast week) and “skippers” (<5 times/week). After controlling for con-
within a population of German primary schoolchildren. Regression founders, family dinner consumers were, however, associated with
model for the prevalence of abdominal obesity, overweight, and reduced odds for central obesity (according to WtHR) by more than
obesity at 1 year follow-up showed a significant association between 30% and reduced odds for obesity (OR = 0.67, 95% CI: 0.5--0.96).
skipping breakfast and abdominal obesity (according to WtHR) and
overweight (OR = 3.36, 95% CI: 2.23–5.07; P = 0.006 and OR = 2.30, Meal Frequency and Composition, Portion
95% CI: 1.54–3.45; P = 0.034, respectively) (67). Girls skipped
breakfast significantly more often than boys (15.2% vs 10.6%).
Size
An Italian study (68) included a representative sample of Eating Frequency
11- to 15-year-old children from 20 Italian regions who completed Two large observational studies and a meta-analysis of 11
a self-reported anonymous questionnaire indicating, in a typical cross-sectional studies showed that a higher number of daily meals
week, how many days they had breakfast (defined as having more is associated with a lower risk of obesity in children.

776 www.jpgn.org

Copyright © ESPGHAN and NASPGHAN. All rights reserved.

JPGN • Volume 72, Number 5, May 2021 Role of Dietary Factors, Food Habits, and Lifestyle in Childhood Obesity

Zurriaga et al (73) conducted a matched case-control study in Children [ISAAC] Phase Three) found an association between
on 1188 Spanish children, ages 2–14 years, and observed that con- increasing frequency of fast-food consumption and higher BMI in 6-
suming 5 meals per day was associated with lower childhood obe- to 7-year-old children but in female adolescents, a higher frequency
sity risk (73). was associated with a lower BMI (82). This latter result could be
In a cross-sectional, multi-centric survey within 13,486 Ira- because of bias, particularly underreporting of fast-food consump-
nian children and adolescents, ages 6 to 18 years, Kelishadi et al tion and reverse causality in adolescents.
(74) noted that as age increased, eating frequency (EF) decreased: It is important to point out that multiple confounders in the
13-year-old students ate ≤3 meals and/or snack, whereas 11 years food environment, such as large supermarkets, convenience stores,
old participants ate ≥6 meals and/or snacks during the day. Anthro- fast-food outlets, and children's age may influence this association.
pometric indices, such as weight, waist circumference, and BMI, Recently, a large cross-sectional study (83), using data from
were higher among those who had an EF of ≤3 compared with those the Yorkshire Health Study (n = 22,889), suggested that the associa-
with EF ≥6. An obese status was observed in 14% of students who tion between fast-food outlets and obesity varies by age. Indeed,
reported EF ≤3, and in 9.5% of those with EF ≥6. Concerning the with increasing age, the highest availability of fast-food outlets has
risk of central obesity, a significant inverse association with EF has been associated with risk of obesity.
been shown; having an EF of 4, 5, or ≥6 decreased the abdominal A previous systematic review investigated the associations
adiposity risk, even though, higher EF might be expected to lead to between food outlets near schools and children's food consump-
excess weight through higher daily caloric intake. The authors of tion but no firm conclusions were made (84). A subsequent study
this study suggested a role in the reduction of hunger provided by conducted in Arkansas, USA (one of the poorest and least healthy
higher and regular EF (74). US states), suggested that the number of FFR within 1 mile can
A large meta-analysis of 11 cross-sectionals studies involv- increase school-level obesity rate (85). This might also apply in
ing 18,849 children, 2 to 19 years of age, showed a modest but nega- European settings, where school days typically end mid or late
tive association of daily EF with weight status (75). In 5 out of the afternoon when hungry children leave school and either go to a
11 studies, the EF associated with lowest weight was 5 meals per nearby FFR or consume fast-food on the way home. Well planned
day. It is important to note that many of the studies in this meta- studies are, however, needed to test this hypothesis.
analysis suffered from limitations inherent to dietary assessment The association of snacking with overweight/obesity is
methods. unclear, particularly for young children for whom snacks are
Snacks are defined as a small portion of food given or con- believed to be the most nutritionally important (86). The prob-
sumed in-between main meals, frequently with an intention of lem is not only the frequency but also portion size and the type of
reducing or preventing hunger until the next mealtime (76). The foods consumed during snacking. Indeed, snacking contributes to
American Academy of Pediatrics (AAP) recommends 2 snacks increased energy intake in children, and the energy density (ED) of
daily for preschool-aged children as part of obesity prevention (77). snacks has increased in recent decades (87,88). Snack foods eaten
Snacking frequency (more than recommended by AAP) has by children increasingly consist of foods, such as desserts, sweet-
been positively associated with weight among preschool children in ened beverages, and salty snacks that also tend to be high in ED,
the 2005 to 2014 National Health and Nutrition Examination Sur- saturated fat, and refined sugars.
vey (NHANES), taking into account the dietary reporting bias (78). A RCT (89) reported the efficacy of a school-based inter-
Normal-weight children tended to snack less frequently than chil- vention, aiming to improve the nutritional value of snacks on the
dren with adiposity excess when considering all foods/beverages dietary intake and waist circumference in 1433 Ecuadorian adoles-
eaten between meals (78). The observed mean effects were very cents. A decreased consumption of table sugar, sweets, salty snacks,
small (3.2 vs 3.3) but these findings raise the possibility that small fast foods, soft drinks, and packaged foods parallel to a reduction of
differences in snacking may accumulate over time to influence obe- waist circumference were observed after 28 months in the interven-
sity risk among young children. Similar results were observed in tion group.
a recent NHANES analysis of snacking and weight status among Recently, a systematic review of cross-sectional, longitudi-
older children (6--11 years old) (79). nal, and experimental studies (90), observed that parental restric-
tive feeding and home access to foods with high ED, saturated
Meal Composition: Consumption of Fast Foods, Snacks fats, and added sugars were consistently associated with snacking
Fast-food consumption is increasingly considered a contrib- among children ages 2 to 18 years.
uting factor for increasing obesity prevalence in childhood. Despite
this, a recent systematic review and meta-analysis of longitudinal and Portion Size
cross-sectional studies (80), focusing on fast-food restaurant (FFR) Three different systematic reviews (91–93) evaluated por-
access and childhood obesity, reported a lack of association in most tion size as a determinant of obesity risk. In Rolls’ systematic
studies when BMI-related continuous measures were used. When- review (91) several experimental studies testing the responsiveness
ever using overweight/obesity outcomes, about half of the cohort to increasing portion size were reported. Three years old children
studies and one-third of the cross-sectional studies reported a posi- appeared largely unaffected, whereas 5 years old children consume
tive association but no significant results were observed in separate more as portion size increases but additional studies have failed to
meta-analyses between various measures of FFR access and body clearly demonstrate such developmental changes in the suscepti-
weight. The authors conclude that this systematic review was limited bility to portion size. Of note, 1 study showed that children who
by methodological diversity of the different studies (80). A study per- were allowed to serve food themselves, ate 25% less of a large
forming comparative analyses in 2 German pediatric cohorts (total main course compared with those who were served the large por-
n=670 children), the “Kiel Obesity Prevention Study” (KOPS) and tion by an adult. In addition, 4 years old children taught to focus on
the “Identification and prevention of Dietary- and lifestyle-induced self-regulatory satiety cues (such as the fullness of their stomachs)
health EFfects In Children and infantS” (IDEFICS-Germany) showed better self-regulation of energy intake than those who were
studies, however, confirmed an adverse impact of fast-food con- rewarded for completing their plates. Although there is convincing
sumption on excess weight gain during primary school years (81). evidence that portion size has persistent effects on energy intake,
Furthermore, a secondary analysis from a multi-centre, international data do not prove that portion size alone plays a role in the aetiol-
cross-sectional study (International Study of Asthma and Allergies ogy of obesity.

www.jpgn.org 777

Copyright © ESPGHAN and NASPGHAN. All rights reserved.

Verduci et al JPGN • Volume 72, Number 5, May 2021

In a further systematic review, Small et al (92) focused on 15 to 20 minutes of vigorous physical activity every day could fur-
2 main themes: to determine the effect of varying portion size and ther help to reduce childhood overweight status (99). In the same
ED of food on energy or food intake and to determine the child's cohort, data showed that over the course of 5 years (6–11 years),
age at which variable portion sizes can affect the dietary intake. children who spent a longer time in sedentary behaviour had a
Even this review, however, pointed out that overall energy intake higher BMI, even when adjusting for time spent in MVPA. This
was affected by the size of the served portion, with larger served observation supports the concept that inactivity is an independent
portions resulting in greater daily energy intake in many of the risk factor for childhood obesity (100). In the same cohort, a longer
children. In addition, ED and portion size positively affected the daily screen time from the ages of 3 to 6 years is associated with
daily energy intake, as the serving and consumption of energy- a higher BMI z score (P = 0.002) and WtHR (P = 0.001) at 6 years
dense foods resulted in increased energy intake. An interest- of age. Specifically, for every additional hour per week of inactiv-
ing finding was that larger portion sizes of vegetables resulted ity, the risk for overweight and obesity increased by 7% and 10%,
in greater vegetable intake, even if not affecting the amount of respectively (101).
the other foods subsequently consumed. The age of the studied A meta-analysis of observational studies published in 2014
populations, included in this systematic review, ranged from 2 to (102) concluded that the degree of physical activity and sedentary
9 years. The age at which young children could override internal behaviour are independent risk factors for obesity, although there
self-limiting mechanisms and might become sensitive to larger is an inverse but weak correlation between the 2 behaviour patterns
portions was not well described as study findings were equivocal (r = −0.108, 95% CI: −0.128 to −0.087).
for children who were 2 and 3 years of age. Children 4 years and A cross-sectional study by the WHO European Childhood
older, however, demonstrated increased energy intake when large Obesity Surveillance Initiative (103), including primary-school
portions were served. Another important finding is that a parent- children (6–9 years) in 5 European countries, evaluated a “food-
directed intervention regarding the children portion education risk behaviour score” and a “physical activity-risk behaviour score,”
appears to be successful only if parents are able to learn how to created for each child based on the presence of 8 food-related and
estimate portion sizes. Interestingly, school-aged children (8–12 5 physical activity-related (including screen time and sleep dura-
years) who directly underwent the same intervention did not learn tion) health-risk behaviours, respectively. Surprisingly, only 4 out
this skill. of 13 health-risk behaviours were directly associated with obesity
A review by Birch et al (93) suggested that, although there and 3 were even found to be negatively associated with obesity or
was a positive correlation between portion size and weight sta- overweight. In contrast “physical activity-risk behaviour score” cor-
tus, data did not support the idea that large portions were caus- related directly with obesity, confirming that both physical activity
ally implicated in the development of greater BMI and obesity and sedentary behaviour have a key role in the energy expenditure
onset, and the positive relation between portion size and weight balance.
status could reflect “reverse causality.” Children with greater body Data show that adolescents tend to have sedentary behav-
mass consumed large portions as their energy requirements were iour and low physical activity more frequently than children, as
greater. Although children's preferences (likes and dislikes) are pri- indicated by a cohort study of 2312 people (104). With regard to
mary determinants of what and how much food is consumed, the sedentary behaviour, a meta-analysis of prospective studies (105)
effects of preferences and palatability have not been systematically showed a linear dose-response relationship between television
studied. Although additional research is needed, findings suggested (TV) watching and childhood obesity, with an increased risk of
that “liking” or palatability may play a role in determining whether 13% for each 1 hour/day increment in TV time. A more recent
increasing portion size increases children's food intake. long-term study (106) showed that TV viewing of at least 2 hours/
Moreover, a variety of factors, including media, marketing, day versus no TV at the age of 3 to 5 years was associated with an
observational learning of parents and others’ eating behaviour, increased risk of overweight and obesity at 5 and 10 years (for 2,
parents’ feeding practices, and postprandial feelings following the 3, and ≥4 hours of TV viewing per day, adjusted ORs were 1.16
consumption of various portion sizes, are likely to be involved in [95% CI: 1.00--1.35], 1.39 [95% CI: 1.15--1.69], and 1.61 [95%
determining how children learn about portion size. CI: 1.20--2.17], respectively). Furthermore, a behavioural pattern
of high TV viewing time/low physical activity level versus low
Physical Activity and Sedentary Behaviour TV viewing time/high physical activity level at ages 3 to 5 years
A negative association between levels of physical activity has been with a risk of overweight/obesity at 5 and 10 years (106).
and overweight/obesity in pre-school, school-age children, and in Moreover, eating while TV viewing has been positively associ-
adolescents has been shown in prospective studies (94,95). ated with childhood and adolescence overweight (OR = 1.28; 95%
According to WHO recommendations (96,97), based on CI: 1.17--1.39) in a recent systematic review and meta-analysis
observational and intervention trials, moderate-vigorous physical of observational studies (107). Subgroup analyses showed similar
activity (MVPA) should be encouraged for at least 60 minutes daily positive associations in children who ate dinner while watching
in all subjects of 3–17 years of age to maintain healthy status. TV (107).
In pre-school aged children, a RCT (98) conducted in Swed- Two meta-analysis of randomised intervention trials
ish child health centres in the context of routine health care, aimed (108,109) and a recent review of systematic reviews (110) of stud-
to reduce the prevalence of obesity as the primary outcome. The ies aimed at reducing sedentary behaviour in children and adoles-
intervention program was performed in 1355 children, starting cents, showed a pooled significant but small, reduction of BMI and
from 8 to 9 months of age and ending at age 4 years, and was based BMI z score. It is important to note that in the Azevedo meta-anal-
on the promotion of healthy food and physical activity habits using ysis (109) and in the Reilly review (110) interventions targeting
motivational interviewing and principles from cognitive behavioral exclusively sedentary behaviour, or sedentary behaviour combined
therapy. After 1 year of follow-up, there were no differences in BMI with physical activity, or sedentary behaviour with other behaviours
between the 2 arms of intervention. (eg diet, sleep) were included. Research about physical activity in
In a European multi-centre study cohort, a secondary analy- the prevention of obesity has key gaps. Many studies differ consid-
sis of 419 11-year-old children revealed that meeting WHO physi- erably in methodology meaning direct comparison is difficult, and
cal activity guidelines appeared adequate to prevent excess weight even well-designed studies will fail to evaluate the impact of every
in children. Moreover, the authors suggested that recommending possible aetiological factor in the obesogenic environment. Thus,

778 www.jpgn.org

Copyright © ESPGHAN and NASPGHAN. All rights reserved.

JPGN • Volume 72, Number 5, May 2021 Role of Dietary Factors, Food Habits, and Lifestyle in Childhood Obesity

many studies were conducted on cohorts including normal weight, • Larger portion size, especially of energy-dense foods, is associated
overweight, or obese children. with greater daily energy intake, a predictor of body weight excess.
• Children can effectively self-regulate energy intake from at least
the age of 4 years and up to 12 years, although energy intake
CONCLUSIONS tends to increase with increasing portion size.

General
Considering the complexity of the development of obesity in Physical Activity, Sedentary Behaviour
children and adolescents, an integrated multi-component approach • A combination of increased physical activity and decreased sed-
is required for obesity prevention. entary time may have an important role in obesity prevention.
Parental eating behaviours are a key determinant of child-
hood obesity.
RECOMMENDATIONS FOR ROUTINE CLINICAL
First 2 Years of Life PRACTICE
• Breast-feeding compared with not breast-feeding has been as- Considering the currently available evidence, the Committee
sociated with a preventive effect on later overweight/obesity. of Nutrition of ESPGHAN recommends:
• In high-income country settings, 6 versus 4 months breast-feed-
ing duration is associated with slower growth rates during in- First 2 Years of Life
fancy, especially if exclusive breast-feeding.
• Compared with breast-feeding, formula feeding is associated • Breast-feeding should be promoted as long as possible during
with altered body composition in infancy. infancy.
• Very small (if any) effects of breast-feeding promotion interven- • High protein intake must be avoided during complementary
tions on growth are reported. feeding.
• A high protein intake during complementary feeding increases • There is insufficient evidence to make any firm recommenda-
the risk of later overweight or obesity. tions about a baby-led weaning approach in terms of childhood
• After 1 year, a whole fat cow's milk has been associated with a obesity prevention.
lower risk of childhood overweight or obesity, compared with • Between 1 and 2 years of life, there is no evidence to recom-
reduced fat cow's milk (although this may be because of reverse mend a reduced-fat cow's milk to prevent childhood overweight
causation). or obesity.
• A modified version of baby-led weaning is associated with bet- • An early response parenting intervention may be included in the
ter food intake self-regulation, without an effect on BMI z score multi-component approach for childhood obesity prevention.
at 24 months of life.
• A responsive parenting intervention during infancy is associated From 2 Years of Life Onwards
with lower BMI z scores at 3 years.
Dietary Patterns, Sugar-sweetened Beverages
Consumption
From 2 Years Onwards • Dietary patterns based on the principles of the Mediterranean
diet, can be used as the best approach for obesity prevention
Dietary Patterns, Sugar-sweetened Beverages in the paediatric age group. This recommendation is based on
Consumption expert opinion as evidence is lacking.
• Mediterranean and Nordic diets may be a promising approach • Public health policies, including correct information for parents
for obesity prevention, but no firm conclusions can be drawn through schools, TV, and other media, and by local health profes-
from the available literature. sionals, should aim to reduce the consumption of SSBs in pre-
• There is insufficient evidence to determine the role of vegetarian school and school children, and in adolescents, while encouraging
diets for the prevention of childhood obesity. healthy alternatives, such as water. Removal of vending machines
• SSBs consumption is positively associated with the develop- selling SSBs where children have access during school breaks, and
ment of obesity. providing the children with free drinking water, should be the goal.

Dietary Modulators of Gut Microbiota

Eating Behaviour
• There is insufficient evidence to recommend the use of pro-,
• Skipping breakfast is associated with obesity, possibly because pre-, or synbiotics for obesity prevention.
of its role in energy balance and dietary regulation.
• Regular family meals are associated with positive health out- Eating Behaviour
comes and weight excess prevention.
• Children and adolescents should be encouraged to consume
breakfast every day.
Meal Frequency and Composition, Portion Size • Family meals should be promoted; in particular for adolescents,
• A higher number of daily meals is associated with a lower risk and family dinners should be consumed at least 5 times per week.
of obesity in children perhaps because of better modulation of
hunger. Meal Frequency and Composition, Portion Size
• During snacking, the consumption of high energy density foods
contributes to obesity in childhood, by increasing daily energy • Children up to the age of 12 years are encouraged to eat at least
intake. 5 meals per day, including a mid-morning and a mid-afternoon

www.jpgn.org 779

Copyright © ESPGHAN and NASPGHAN. All rights reserved.

Verduci et al JPGN • Volume 72, Number 5, May 2021

snack. Whether eating 6 or more meals per day provides an ad- • Intervention studies to identify how to modify/change the psy-
ditional contribution to the prevention of overweight/obesity chological process involved in eating behaviour (decision-mak-
remains to be elucidated. ing, impulsive behavior, …).
• Healthy food options (fruit and vegetables) should be promoted
for snacking while avoiding consuming high energy density
foods (chips, cookies, sweets). Meal Frequency and Composition, Portion Size
• Parents need to be educated on healthy food choices and appro- • Determining the key aspects of snacking that may influence the
priate food portion sizes and they must share this information risk of obesity, including the quality, timing, and portion sizes
with their children. of snacks offered.
• Identifying the key parenting behaviours around childhood
Physical Activity, Sedentary Behaviour snacking, which may be used as targets for promoting good
health.
• Children (>3 years) and adolescents should spend on average • Longitudinal studies evaluating the food environment especially
60 minutes a day on moderate-to-vigorous physical activity. in school neighbourhoods. Policymakers should consider the
• Screen time and sedentary behaviour, should be limited in children impact of any planned fast-food outlet interventions according
and the use of screen devices should be avoided during mealtimes. to the likely presence of children of different ages.
• High-quality studies that evaluate the role of factors, such as
age, race, ethnicity, income, food insecurity, liking, palatability,
and weight status (of both children and parents) and how these
RECOMMENDATIONS FOR FUTURE RESEARCH may interact with portion size.
Considering the currently available evidence, the Commit- • Evaluation on approach of shifting from restriction to more-
tee of Nutrition of ESPGHAN recommends future research should positive messages relating to the increase of healthy, low-ED
focus on: foods intake, on obesity risk.

First 2 Years of Life

Physical Activity, Sedentary Behaviour
• Additional studies to assess the role of excess energy intakes,
irrespective of macronutrients balance, on later overweight/obe- • High-quality studies that evaluate the combined effects of physi-
sity in case of formula-fed infants. cal activity and sedentary behaviour interventions in normal
• Well-designed intervention trials that evaluate the effect of re- weight children and impacts on obesity prevalence.
duced-fat intake on childhood overweight or obesity prevention. • High-quality studies that evaluate vigorous physical activity to
• Conducting large studies to determine whether responsive par- reduce the risk of overweight and obesity to determine the types
enting interventions are associated with better development of of physical activity that are most beneficial according to the
child self-regulation. child's age.
Studies to design individualised interventions for obesity preven-
tion from early life.
REFERENCES
1. Kerry S. National Heart forum. Healthy weight, healthy lives: a toolkit
From 2 Years of Life Onwards for developing local strategies. http://www.fph.org.uk/uploads/full_
obesity_toolkit-1.pdf. Accessed June 20, 2018.
Dietary Patterns, Sugar-sweetened Beverages 2. World Health Organization. Report of the commission on ending child-
Consumption hood obesity. Geneva, Switzerland: World Health Organization; 2016.
3. World Health Organization: Childhood overweight and obesity. http://
• Conducting high-quality intervention studies to evaluate the ef- www.who.int/dietphysicalactivity/childhood/en/. Accessed June 20,
fect of childhood obesity prevention in different populations and 2018.
according to different sex. 4. Styne DM, Arslanian SA, Connor EL, et al. Pediatric obesity-assess-
• Development of a unified and universally validated score for ment, treatment, and prevention: an Endocrine Society Clinical Practice
MD adherence in terms of reproducibility and consistency. Guideline. J Clin Endocrinol Metab 2017; 102:709–757.
• The effect of taxation on SSBs on the reduction of obesity devel- 5. Zimmet P, Alberti KG, Kaufman F, et al. IDF Consensus Group. The
opment in different countries. metabolic syndrome in children and adolescents - an IDF consensus
report. Pediatr Diabetes 2007; 8:299–306.
• Studies that determine how best to reduce the amount of free 6. Ahrens W, Moreno LA, Mårild S, et al. IDEFICS consortium. Metabolic
sugars in solid foods. syndrome in young children: definitions and results of the IDEFICS
study. Int J Obes 2014; 38:S4–S14.
7. Wells JC, Sawaya AL, Wibaek R, et al. The double burden of malnutri-
Dietary Modulation of Gut Microbiota tion: aetiological pathways and consequences for health. Lancet 2019;
395:75–88.
• Identify the different dysbiosis patterns in the first 2 years to de- 8. Rudolf M. Predicting babies’ risk of obesity. Arch Dis Child 2011;
velop individualised interventions. 96:995–997.
• High-quality RCTs that evaluate the effect of dietary modula- 9. Silano M, Agostoni C, Fattore G. Italy's unsolved childhood obesity
tors of gut microbiota (pre-, pro-, synbiotics) on obesity risk in crisis. Arch Dis Child 2019; 104:202–203.
children. 10. ESPGHAN Committee on Nutrition. Role of dietary factors and food
habits in the development of childhood obesity: a commentary by the
ESPGHAN Committee on Nutrition. J Pediatr Gastroenterol Nutr
Eating Behaviour 2011; 52:662–669.
11. Gale C, Logan KM, Santhakumaran S, et al. Effect of breastfeeding
• Interventional studies that confirm whether skipping breakfast compared with formula feeding on infant body composition: a system-
in childhood and adolescence is causally related to adiposity. atic review and meta-analysis. Am J Clin Nutr 2012; 95:656–669.

780 www.jpgn.org

Copyright © ESPGHAN and NASPGHAN. All rights reserved.

JPGN • Volume 72, Number 5, May 2021 Role of Dietary Factors, Food Habits, and Lifestyle in Childhood Obesity

12. Horta BL, Loret de Mola C, Victora CG. Long-term consequences of 34. Iaccarino Idelson P, Scalfi L, Valerio G. Adherence to the Mediterranean
breastfeeding on cholesterol, obesity, systolic blood pressure and type diet in children and adolescents: a systematic review. Nutr Metab
2 diabetes: a systematic review and meta-analysis. Acta Paediatr 2015; Cardiovasc Dis 2017; 27:283–299.
104:30–37. 35. Katsagoni CN, Psarra G, Georgoulis M, et al. EYZHN Study Group.
13. Yan J, Liu L, Zhu Y, et al. The association between breastfeeding High and moderate adherence to Mediterranean lifestyle is inversely
and childhood obesity: a meta-analysis. BMC Public Health 2014; associated with overweight, general and abdominal obesity in children
14:1267. and adolescents: the MediLIFE-index. Nutr Res 2019; 73:38–47.
14. Ip S, Chung M, Raman G, et al. A summary of the Agency for 36. Besharat Pour M, Bergström A, Bottai M, et al. Effect of parental
Healthcare Research and Quality's evidence report on breastfeeding in migration background on childhood nutrition, physical activity, and
developed countries. Breastfeed Med 2009; 4 Suppl 1:S17–S30. body mass index. J Obes 2014; 2014:406529.
15. Patro-Golab B, Zalewki BM, Polaczek A, et al. Duration of breast- 37. Becker W. New Nordic nutrition recommendations, 2004. Physical
feeding and early growth: a systematic review of current evidence. activity as important as good nourishing food. Lakartidningen 2005;
Breastfeed Med 2019; 14:218–229. 102:2757–2758.
16. Lakshman R, Elks CE, Ong KK. Childhood obesity. Circulation 2012; 38. Andersen R, Biltoft-Jensen A, Christensen T, et al. Dietary effects of
126:1770–1779. introducing school meals based on the New Nordic Diet – a randomised
17. Druet C, Stettler N, Sharp S, et al. Prediction of childhood obesity by controlled trial in Danish children. The OPUS School Meal Study. Br J
infancy weight gain: an individual-level meta-analysis. Paediatr Perinat Nutr 2014; 111:1967–1976.
Epidemiol 2012; 26:19–26. 39. Agnoli C, Baroni L, Bertini I, et al. Position paper on vegetarian diets
18. Giugliani ER, Horta BL, Loret de Mola C, et al. Effect of breastfeeding from the working group of the Italian Society of Human Nutrition. Nutr
promotion interventions on child growth: a systematic review and meta- Metab Cardiovasc Dis 2017; 27:1037–1052.
analysis. Acta Paediatr 2015; 104:20–29. 40. An exploration into diets around the world. Game Changers, Ipsos.
19. Graham GN. Why your ZIP code matters more than your genetic code: 2018.
promoting healthy outcomes from mother to child. Breastfeed Med 41. Hebbelinck M, Clarys P, De Malsche MA. Growth, development t,and
2016; 11:396–397. physical fitness of Flemish vegetarian children, adolescents, and young
20. Fewtrell M, Bronsky J, Campoy C, et al. Complementary feeding: a adults. Am J Clin Nutr 1999; 70:579S–585S.
position paper by the European Society for Paediatric Gastroenterology, 42. Sabate J, Lindsted KD, Harris RD, et al. Anthropometric parameters
Hepatology, and Nutrition (ESPGHAN) Committee on Nutrition. J of schoolchildren with different life-styles. Am J Dis Child 1990;
Pediatr Gastroenterol Nutr 2017; 64:119–132. 144:1159–1163.
21. Castenmiller J, de Henauw S, et al. EFSA Panel on Nutrition. Novel 43. Cooper R, Allen A, Goldberg R, et al. Seventh-day adventist adoles-
Foods and Food Allergens (NDA). Appropriate age range for intro- cents e life-style patterns and cardiovascular risk factors. West J Med
duction of complementary feeding into an infant's diet. EFSA J 2019; 1984; 140:471.e7.
17:e05780. 44. Persky VW, Chatterton RT, Van Horn LV, et al. Hormone levels in vege-
22. Appleton J, Russell CG, Laws R, et al. Infant formula feeding practices tarian and nonvegetarian teenage girls: potential implications for breast
associated with rapid weight gain: a systematic review. Matern Child cancer risk. Cancer Res 1992; 52:578–583.
Nutr 2018; 14:e12602. 45. Nathan I, Hackett AF, Kirby S. A longitudinal study of the growth
23. Koletzko B, von Kries R, Closa R, et al. European Childhood Obesity of matched pairs of vegetarian and omnivorous children, aged
Trial Study Group. Lower protein in infant formula is associated with 7-11 years, in the north-west of England. Eur J Clin Nutr 1997;
lower weight up to age 2 y: a randomized clinical trial. Am J Clin Nutr 51:20–25.
2009; 89:1836–1845. 46. O’Connell JM, Dibley MJ, Sierra J, et al. Growth of vegetarian children:
24. Weber M, Grote V, Closa-Monasterolo R, et al. European Childhood the Farm study. Pediatrics 1989; 84:475–481.
Obesity Trial Study Group. Lower protein content in infant formula 47. Grant R, Bilgin A, Zeuschner C, et al. The relative impact of a vegeta-
reduces BMI and obesity risk at school age: follow-up of a randomized ble-rich diet on key markers of health in a cohort of Australian adoles-
trial. Am J Clin Nutr 2014; 99:1041–1051. cents. Asia Pac J Clin Nutr 2008; 17:107–115.
25. Inostroza J, Haschke F, Steenhout P, et al. Low-protein formula slows 48. Melina V, Craig W, Levin S. Position of the Academy of Nutrition and
weight gain in infants of overweight mothers. J Pediatr Gastroenterol Dietetics: vegetarian diets. J Acad Nutr Diet 2016; 116:1970–1980.
Nutr 2014; 59:70–77. 49. Fidler Mis N, Braegger C, Bronsky J, et al. ESPGHAN Committee on
26. Agostoni C, Guz-Mark A, Marderfeld L, et al. The long-term effects Nutrition. Sugar in Infants, Children and Adolescents: a position paper
of dietary nutrient intakes during the first 2 years of life in healthy of the European Society for Paediatric Gastroenterology, Hepatology
infants from developed countries: an umbrella review. Adv Nutr 2019; and Nutrition Committee on Nutrition. J Pediatr Gastroenterol Nutr
10:489–501. 2017; 65:681–696.
27. Ferguson MC, O'Shea KJ, Hammer LD. Can following formula-feeding 50. Luger M, Lafontan M, Bes-Rastrollo M, et al. Sugar-sweetened bever-
recommendations still result in infants who are overweight or have obe- ages and weight gain in children and adults: a systematic review from
sity? Pediatr Res 2020; 88:661–667. 2013 to 2015 and a comparison with previous studies. Obes Facts 2017;
28. Vanderhout SM, Aglipay M, Torabi N, et al. Whole milk compared with 10:674–693.
reduced-fat milk and childhood overweight: a systematic review and 51. Vercammen KA, Frelier JM, Lowery CM, et al. A systematic review
meta-analysis. Am J Clin Nutr 2020; 111:266–279. of strategies to reduce sugar-sweetened beverage consumption among
29. ESPGHAN Committee on Nutrition. Young child formula: a posi- 0-year to 5-year olds. Obes Rev 2018; 19:1504–1524.
tion paper by the ESPGHAN Committee on Nutrition. J Pediatr 52. Crow-White K, O’Neil CE, Parrott JS, et al. Impact of 100% fruit juice
Gastroenterol Nutr 2018; 66:177–185. consumption on diet and weight status in children: an evidence based
30. Taylor RW, William SM, Fangupo LJ, et al. Effect of a baby-led approach review. Crit Rev Food Sci Nutr 2016; 56:871–874.
to complementary feeding on infant growth and overweight: a random- 53. Shefferly A, Scharf RJ, DeBoer MD. Longitudinal evaluation of 100%
ized clinical trial. JAMA Pediatr 2017; 171:838–846. fruit juice consumption on BMI status in 2–5-year-old children. Pediatr
31. Morandi A, Tommasi M, Soffiati F, et al. Prevention of obesity in tod- Obes 2016; 11:221–227.
dlers (PROBIT): a randomised clinical trial of responsive feeding pro- 54. Cryan JF, O’Riordan KJ, Cowan CSM, et al. The microbiota-gut-brain
motion from birth to 24 months. Int J Obes 2019; 43:1961–1966. axis. Physiol Rev 2019; 99:1877–2013.
32. Paul IM, Savage JS, Anzman-Frasca S, et al. Effect of a responsive 55. De Clercq NC, Groen AK, Romijn JA, et al. Gut microbiota in obesity
parenting educational intervention on childhood weight outcomes at 3 and undernutrition. Adv Nutr 2016; 7:1080–1089.
years of age: the INSIGHT Randomized Clinical Trial. JAMA 2018; 56. Zmora N, Suez J, Elinav E. You are what you eat: diet, health and the gut
320:461–468. microbiota. Nat Rev Gastroenterol Hepatol 2019; 16:35–56.
33. Pereira-da-Silva L, Rêgo C, Pietrobelli A. The diet of preschool chil- 57. Brahe LK, Astrup A, Larsen LH. Is butyrate the link between diet,
dren in the Mediterranean countries of the European union: a system- intestinal microbiota and obesity-related metabolic diseases? Obesity
atic review. Int J Environ Res Public Health 2016; 13:572. Rev 2013; 14:950–959.

www.jpgn.org 781

Copyright © ESPGHAN and NASPGHAN. All rights reserved.

Verduci et al JPGN • Volume 72, Number 5, May 2021

58. Khan MJ, Gerasimidis K, Edwards CA, et al. Role of gut microbiota 79. Murakami K, Livingstone MBE. Associations between meal and
in the aetiology of obesity: proposed mechanisms and review of the snack frequency and overweight and abdominal obesity in US chil-
literature. J Obes 2016; 2016:7353642. dren and adolescents from National Health and Nutrition Examination
59. Ridaura VK, Faith JJ, Rey FE, et al. Cultured gut microbiota from Survey (NHANES). Br J Nutr 2016; 115:1819–1829.
twins discordant for obesity modulate adiposity and metabolic pheno- 80. Jia P, Luo M, Li Y, et al. Fast-food restaurant, unhealthy eating, and
types in mice. Science 2013; 341:1241214. childhood obesity: a systematic review and meta-analysis. Obes Rev
60. Cerdó T, Ruiz A, Jáuregui R, et al. Maternal obesity is associated 2019; (Epub ahead of print).
with gut microbial metabolic potential in offspring during infancy. J 81. Wolters M, Joslowski G, Plachta-Danielzik S, et al. Dietary patterns
Physiol Biochem 2018; 74:159–169. in primary school are of prospective relevance for the development
61. Indiani CMDSP, Rizzardi KF, Castelo PM, et al. Childhood obesity of body composition in two German pediatric populations. Nutrients
and firmicutes/bacteroidetes ratio in the gut microbiota: a systematic 2018; 10:1442.
review. Child Obes 2018; 14:501–509. 82. Braithwaite I, Stewart AW, Hancox RJ, et al. ISAAC Phase Three
62. Cerdó T, García-Santos JA, Bermúdez MG, et al. The role of probiot- Study Group. Fast-food consumption and body mass index in children
ics and prebiotics in the prevention and treatment of obesity. Nutrients and adolescents: an international cross-sectional study. BMJ Open
2019; 11:635. 2014; 4:e005813.
63. Rampelli S, Guenther K, Turroni S, et al. Pre-obese children's dysbi- 83. Hobbs M, Green M, Roberts K, et al. Reconsidering the relationship
otic gut microbiome and unhealthy diets may predict the development between fast-food outlets, area-level deprivation, diet quality and body
of obesity. Commun Biol 2018; 1:222. mass index: an exploratory structural equation modelling approach. J
64. Luoto R, Kalliomäki M, Laitinen K, et al. The impact of perinatal Epidemiol Community Health 2019; 73:861–866.
probiotic intervention on the development of overweight and obesity: 84. Williams J, Scarborough P, Matthews A, et al. A systematic review of
follow-up study from birth to 10 years. Int J Obesity 2010; 34:1531. the influence of the retail food environment around schools on obe-
65. Karlsson Videhult F, Öhlund I, Stenlund H, et al. Probiotics during sity-related outcomes. Obes Rev 2014; 15:359–374.
weaning: a follow-up study on effects on body composition and meta- 85. Alviola PA4th, Nayga RMJr, Thomsen MR, et al. The effect of fast-
bolic markers at school age. Eur J Nutr 2015; 54:355–363. food restaurants on childhood obesity: a school level analysis. Econ
66. Monzani A, Ricotti R, Caputo M, et al. A systematic review of the Hum Biol 2014; 12:110–119.
association of skipping breakfast with weight and cardiometabolic 86. Casazza K, Fontaine KR, Astrup A, et al. Myths, presumptions, and
risk factors in children and adolescents. What should we better inves- facts about obesity. N Engl J Med 2013; 368:446–454.
tigate in the future? Nutrients 2019; 11:387. 87. Dunford EK, Popkin BM. 37-year snacking trends for US children.
67. Traub M, Lauer R, Kesztyüs T, et al. Research Group “Join the Pediatr Obes 2018; 13:247–255.
Healthy Boat”. Skipping breakfast, overconsumption of soft drinks 88. Pereira-da-Silva L, Rêgo C, Pietrobelli A. The diet of preschool chil-
and screen media: longitudinal analysis of the combined influence on dren in the Mediterranean Countries of the European Union: a system-
weight development in primary schoolchildren. BMC Public Health atic review. Int J Environ Res Public Health 2016; 13:572.
2018; 18:363. 89. Ochoa-Avilés A, Verstraeten R, Huybregts L, et al. A school-based
68. Lazzeri G, Giacchi MV, Spinelli A, et al. Overweight among students intervention improved dietary intake outcomes and reduced waist
aged 11-15 years and its relationship with breakfast, area of residence circumference in adolescents: a cluster randomized controlled trial.
and parents’ education: results from the Italian HBSC 2010 cross- Nutrients 2017; 16:79.
sectional study. Nutr J 2014; 13:69. 90. Blaine RE, Kachurak A, Davison KK, et al. Food parenting and child
69. Valdés J, Rodríguez-Artalejo F, Aguilar L, et al. Frequency of family snacking: a systematic review. Int J Behav Nutr Phys Act 2017; 14:146.
meals and childhood overweight: a systematic review. Pediatr Obes 91. Rolls BJ. Dietary strategies for the prevention and treatment of obe-
2013; 8:e1–e13. sity. Proc Nutr Soc 2010; 69:70–79.
70. Roos E, Pajunen T, Ray C, et al. Does eating family meals and having 92. Small L, Lane H, Vaughan L, et al. A systematic review of the evi-
the television on during dinner correlate with overweight? A sub-study dence: the effects of portion size manipulation with children and por-
of the PRO GREENS Project, looking at children from nine European tion education/training interventions on dietary intake with adults.
countries. Public Health Nutr 2014; 17:2528–2536. Worldviews Evid Based Nurs 2013; 10:69–81.
71. Shirasawa T, Ochiai H, Yoshimoto T, et al. Effects of eating dinner 93. Birch LL, Savage JS, Fisher JO. Right sizing prevention. Food portion
alone on overweight in japanese adolescents: a cross-sectional survey. size effects on children's eating and weight. Appetite 2015; 88:11–16.
BMC Pediatr 2018; 18:36. 94. te Velde SJ, van Nassau F, Uijtdewilligen L, et al. ToyBox-study group.
72. Haghighatdoost F, Kelishadi R, Qorbani M, et al. Family dinner fre- Energy balance-related behaviours associated with overweight and
quency is inversely related to mental disorders and obesity in adoles- obesity in preschool children: a systematic review of prospective stud-
cents: the CASPIAN-III Study. Arch Iran Med 2017; 20:218–223. ies. Obes Rev 2012; 13: (Suppl 1): 56–74.
73. Zurriaga O, Pérez-Panadés J, Quiles Izquierdo J, et al. Factors asso- 95. Pate RR, O’Neill JR, Liese AD, et al. Factors associated with devel-
ciated with childhood obesity in Spain. The OBICE study: a case– opment of excessive fatness in children and adolescents: a review of
control study based on sentinel networks. Recent OBICE Research prospective studies. Obes Rev 2013; 14:645–658.
Group. Public Health Nutr 2011; 14:1105–1113. 96. World Health Organization. Global recommendations on physical
74. Kelishadi R, Qorbani M, Motlagh ME, et al. Association of eating activity for health. 2010.
frequency with anthropometric indices and blood pressure in children 97. World Health Organization. Guidelines on physical activity, sedentary
and adolescents: the CASPIAN-IV Study. J Pediatr (Rio J) 2016; behaviour and sleep for children under 5 years of age. 2010.
92:156–167. 98. Enö Persson J, Bohman B, Tynelius P, et al. Prevention of childhood
75. Kaisari P, Yannakoulia M, Panagiotakos DB. Eating frequency and obesity in child health services: follow-up of the PRIMROSE Trial.
overweight and obesity in children and adolescents: a meta-analysis. Child Obes 2018; 14:99–105.
Pediatrics 2013; 131:958–967. 99. Schwarzfischer P, Weber M, Gruszfeld D, et al. BMI and recom-
76. Younginer NA, Blake CE, Davison KK, et al. What do you think of mended levels of physical activity in school children. BMC Public
when I say the word 'snack’?” Towards a cohesive definition among low- Health 2017; 17:595.
income caregivers of preschool-age children. Appetite 2016; 98:35–40. 100. Schwarzfischer P, Gruszfeld D, Socha P, et al. Longitudinal analysis of
77. American Academy of Pediatrics. HALF Implementation Guide: physical activity, sedentary behaviour and anthropometric measures
age specific content. https://www.aap.org/en-us/advocacy-and- from ages 6 to 11 years. Int J Behav Nutr Phys Act 2018; 15:126.
policy/aap-health-initiatives/HALF-ImplementationGuide/ 101. Schwarzfischer P, Gruszfeld D, Socha P, et al. Effects of screen time
Age-Specific-Content/Pages/Age-Specific-Content.aspx. <date-in- and playing outside on anthropometric measures in preschool aged
citation>Published 2016. Accessed June 20, 2018</date-in-citation>. children. PLoS One 2020; 15:e0229708.
78. Kachurak A, Davey A, Bailey RL, et al. Daily snacking occasions 102. Pearson N, Braithwaite RE, Biddle SJ, et al. Associations between sed-
and weight status among us children aged 1 to 5 years. Obesity 2018; entary behaviour and physical activity in children and adolescents: a
26:1034–1042. meta-analysis. Obes Rev 2014; 15:666–675.

782 www.jpgn.org

Copyright © ESPGHAN and NASPGHAN. All rights reserved.

JPGN • Volume 72, Number 5, May 2021 Role of Dietary Factors, Food Habits, and Lifestyle in Childhood Obesity

103. Wijnhoven TM, van Raaij JM, Yngve A, et al. WHO European among children and adolescents: a systematic review and meta-analy-
Childhood Obesity Surveillance Initiative: health-risk behaviours on sis of observational studies. Obes Rev 2018; 19:313–320.
nutrition and physical activity in 6-9-year-old schoolchildren. Public 108. van Grieken A, Ezendam NP, Paulis WD, et al. Primary prevention of
Health Nutr 2015; 18:3104–3128. overweight in children and adolescents: a meta-analysis of the effec-
104. Ortega FB, Konstabel K, Pasquali E, et al. Objectively measured tiveness of interventions aiming to decrease sedentary behaviour. Int J
physical activity and sedentary time during childhood, adolescence Behav Nutr Phys Act 2012; 9:61.
and young adulthood: a cohort study. PLoS One 2013; 8:e60871. 109. Azevedo LB, Ling J, Soos I, et al. The effectiveness of sedentary
105. Zhang G, Wu L, Zhou L, et al. Television watching and risk of child- behaviour interventions for reducing body mass index in children and
hood obesity: a meta-analysis. Eur J Pub Health 2016; 26:13–18. adolescents: systematic review and meta-analysis. Obes Rev 2016;
106. Tahir MJ, Willett W, Forman MR. The association of television view- 17:623–635.
ing in childhood with overweight and obesity throughout the life 110. Reilly JJ, Hughes AR, Gillespie J, et al. Physical activity interventions
course. Am J Epidemiol 2019; 188:282–293. in early life aimed at reducing later risk of obesity and related non-
107. Ghobadi S, Hassanzadeh-Rostami Z, Salehi-Marzijarani M, et al. communicable diseases: a rapid review of systematic reviews. Obes
Association of eating while television viewing and overweight/obesity Rev 2019; 20:61–73.

www.jpgn.org 783

Copyright © ESPGHAN and NASPGHAN. All rights reserved.