*Article* **Lifestyle Changes and Determinants of Children's and Adolescents' Body Weight Increase during the First COVID-19 Lockdown in Greece: The COV-EAT Study**

**Odysseas Androutsos 1,\*, Maria Perperidi 1, Christos Georgiou <sup>1</sup> and Giorgos Chouliaras <sup>2</sup>**


**Abstract:** Previous studies showed that the coronavirus disease 2019 (COVID-19) lockdown imposed changes in adults' lifestyle behaviors; however, there is limited information regarding the effects on youth. The COV-EAT study aimed to report changes in children's and adolescents' lifestyle habits during the first COVID-19 lockdown and explore potential associations between changes of participants' lifestyle behaviors and body weight. An online survey among 397 children/adolescents and their parents across 63 municipalities in Greece was conducted in April–May 2020. Parents self-reported changes of their children's lifestyle habits and body weight, as well as sociodemographic data of their family. The present study shows that during the lockdown, children's/adolescents' sleep duration and screen time increased, while their physical activity decreased. Their consumption of fruits and fresh fruit juices, vegetables, dairy products, pasta, sweets, total snacks, and breakfast increased, while fast-food consumption decreased. Body weight increased in 35% of children/adolescents. A multiple regression analysis showed that the body weight increase was associated with increased consumption of breakfast, salty snacks, and total snacks and with decreased physical activity. The COV-EAT study revealed changes in children's and adolescents' lifestyle behaviors during the first COVID-19 lockdown in Greece. Effective strategies are needed to prevent excessive body weight gain in future COVID-19 lockdowns.

**Keywords:** COVID-19; obesity; children; lifestyle; determinants; diet; physical activity; sedentary behavior; COV-EAT

#### **1. Introduction**

Since December 2019, the world is facing a new disease (coronavirus disease 2019 (COVID-19)), caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The WHO declared COVID-19 a pandemic in March 2020. In Greece, a number of regulatory measures have been implemented, including the closing of schools on 11 March and finally a lockdown, imposed on 23 March as a "last-resort" preventive measure to halt the spreading of the disease until COVID-19 vaccines would be available. This unprecedented situation led to significant changes in children's daily routine, who no longer attended school and out-of-school activities (e.g., participation in sports, free play at playgrounds, etc.), but were isolated at home with their families.

Studies conducted during the COVID-19 pandemic in adults have shown that selfisolation at home due to the lockdown was associated with lower level of physical activity, longer sedentary time, modifications in eating behavior, and sleeping disturbances [1–3]. Furthermore, an increase of food purchased before the pandemic was reported in some countries, which increased the availability of foods during the lockdown [4]. Self-isolation has been also linked to boredom and stress, which in turn may lead to higher energy

**Citation:** Androutsos, O.; Perperidi, M.; Georgiou, C.; Chouliaras, G. Lifestyle Changes and Determinants of Children's and Adolescents' Body Weight Increase during the First COVID-19 Lockdown in Greece: The COV-EAT Study. *Nutrients* **2021**, *13*, 930. https://doi.org/10.3390/nu 13030930

Academic Editor: Amelia Martí

Received: 7 January 2021 Accepted: 10 March 2021 Published: 13 March 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

intake, consumption of energy-dense "comfort foods", and emotional eating [5]. To date, there is a lack of studies focusing on changes of children's lifestyle behaviors during the COVID-19 lockdown. However, there are studies which have examined the impact of school closing in the summer period, which is a condition similar to the COVID-19 lockdown, on children's eating behavior and weight status [3,6–9]. The majority of these studies concluded that children's body mass index (BMI) increased more rapidly during summer vacation compared to school period [7]. Moreover, it was revealed that children tend to consume less vegetables and more sugar, spend more time on TV-watching, and be more active during summer breaks compared to school time [8]. Interestingly, the study of Von Hippel et al. showed that the prevalence of childhood obesity increased only during summer vacations and not during any other time period of the year [6].

Factors that may influence children's eating behavior have been previously reported and include certain lifestyle behaviors, such as sedentary behavior [10–15], insufficient sleep duration [10,12–14], and inactivity [10–13,15–17], as well as determinants from children's social and physical environment, such as parental modeling [10,11,16,17] and availability of foods at home [11,16]. It is also known that retaining higher energy intake compared to energy expenditure for long periods of time increases the risk of overweight/obesity [18].

The scientific community raises concerns about the health implications that may be caused by the COVID-19 lockdown [1,2,19–22]. The present study aims to report changes in children's and adolescents' lifestyle behaviors during the first lockdown that was implemented in Greece due to COVID-19 and explore potential associations with changes of their body weight.

#### **2. Materials and Methods**

#### *2.1. Study Design and Participants*

The COV-EAT study was a cross-sectional study, which was conducted across 63 municipalities in Greece. Parents having children aged 2–18 years were invited to participate. The following inclusion criteria were applied: living in Greece, being able to complete the study questionnaire in Greek language, having children aged 2–18 years, and providing a consent form. The COV-EAT study adhered to the Declaration of Helsinki and the conventions of the Council of Europe on human rights and Biomedicine. The study protocol was approved by the Ethical Committee of the Department of Physical Education and Sport Science in the School of Physical Education, Sport Science, and Dietetics, University of Thessaly, and registered at clinicaltrials.org (NCT04437121). All parents electronically signed an informed consent form prior to their participation in the study. Only one child per family was included in this study.

#### *2.2. Instruments and Variables*

An online survey was conducted in a sample of families, who were invited to electronically fill in a questionnaire which included 70 questions divided in 3 sections. The first section focused on participants' sociodemographic data and included 15 questions about child's gender, parents' and child's age, number of children in the family, city of residence, and socioeconomic status. The second section contained 20 questions about parents' dietary and lifestyle habits before and during the lockdown. Specifically, parents had to answer questions about the frequency of cooking at home and their fast-food consumption, the number of meals and snacks consumed per day, the reasons of snacking, and their body weight and height. The third section contained 35 questions about their child's eating habits and its sedentary behavior. Parents were asked to report their child's body weight and height, the number of meals and snacks during the day, the frequency of breakfast, fast-food, fruits, juices, vegetables, dairy, red meat, poultry, fish, pasta, legumes, sweets, salty snacks, and beverages consumption, the servings of fruits, juices, vegetables, and dairy consumed per day, vitamin supplementation, as well as screen time, sleep duration, and changes in physical activity. All answers were given about the period before and during the lockdown. According to the difference between the values of each lifestyle

behavior before and during the lockdown, each parameter was categorized as "decrease" (i.e., before-lockdown value of lifestyle behavior higher than during-lockdown value), "stable" (i.e., same values of lifestyle behavior before and during the lockdown), or "increase" (i.e., before-lockdown value of lifestyle behavior lower than during-lockdown value).

The questionnaire is available as a Supplementary Materials.

#### *2.3. Procedure*

All data were collected between 30 April and 24 May 2020. The questionnaire was created using the Google Forms tool and was distributed electronically via networks of dietitians/nutritionists in Greece, personal networks, and social media.

#### *2.4. Statistical Analyses*

Continuous data are presented as mean ± standard deviation (SD). Categorical variables are presented as absolute (n) and relative (%) frequencies. Mean values of consumption of food groups before and during the quarantine were compared using the Student's test for paired data. Associations between categorical data were evaluated with Fisher's exact test. For paired pre- vs. post-comparisons of sleeping time, the extended McNemar's test (allowing for 3 × 3 contingency tables in matched observations) was used. For paired pre- vs. post-comparisons of screen time, the Wilcoxon matched-pairs signed-rank test was applied. A stepwise, backwards, regression approach was utilized to assess the effect of dietary data on the probability of body weight increase versus no change or decrease (logistic regression). Participants that answered "don't know" were treated as missing values. The initial, full model included dietary data, gender, age, area of residency, change in sleep and screen time, and change in physical activity. For logistic regression analysis, results are reported as odds ratios, along with 95% confidence intervals (CI) and *p*-values. The level of statistical significance was set to 0.05 for analyses, and in cases of multiple comparisons, the Bonferroni correction was applied. All analyses were run in Stata 11 MP statistical software (StataCorp., College Station, TX, USA).

#### **3. Results**

In total, 397 dyads of children/adolescents (51.4% boys) with an average age of 7.8 (4.1) years were recruited. Families' characteristics are presented in Table 1, and detailed demographic and socioeconomic data are presented in Table 2.

Tables 3 and 4 present children's/adolescents' changes of lifestyle behaviors before vs. during the lockdown. Specifically, during the lockdown, more children/adolescents tended to sleep longer than 10 h/night, and fewer slept less than 8 h/night than before the lockdown. Similarly, the children/adolescents who spent more than 3 h/day in front of a screen were more during home isolation, whereas those who did not spent time on a screen were less during the confinement than before the confinement. Moreover, 66.9% of the parents reported that their child's physical activity level was decreased during the lockdown, and 35% that their child's body weight increased. Regarding children's eating behavior, the consumption of fruits and fresh fruit juices, vegetables, dairy products, pasta, sweets, total snacks, and breakfast significantly increased (*p* < 0.05). In contrast, fast-food consumption was significantly decreased (*p* < 0.001). No significant changes were observed in core foods used for lunch and dinner, such as red meat, poultry, fish, and legumes. Similarly, no significant changes were observed in the consumption of prepacked juices and sodas and salty snacks.


**Table 1.** Families' age and anthropometric characteristics. The COV-EAT study.

**Table 2.** Family socio-demographic status. The COV-EAT study.


Data are shown as absolute (*n*) and relative (%) frequencies. Missing values were not included in the statistical analyses.



\* Data are shown as absolute (*n*) and relative (%) frequencies. \*\* Extended McNemar's test: *p* < 0.001. \*\*\* Wilcoxon matched-pairs signed-rank test: *p* < 0.001. \*\*\*\* Seven missing values for the variable (1.8% "don't know" responses of the total). \*\*\*\*\* Thirty-four missing values for the variable (8.6% "don't know" responses of the total).


**Table 4.** Changes of children's \* and adolescents' eating habits in the first COVID-19 lockdown in Greece. The COV-EAT study.

\* Data presented as mean (SD) consumption (in servings/day) of each food. Breakfast consumption indicates the frequency of consumption on a weekly level. \*\* *t*-test for paired data/students test.

Table 5 presents the bivariate correlations between body weight increase and changes in lifestyle behaviors. More specifically, body weight increase was correlated with increase of consumption of salty snacks and red meat (*p* < 0.05). Similarly, increase of sleep duration (*p* = 0.012) and screen time (*p* < 0.001) and decrease of physical activity (*p* < 0.001) were associated with body weight increase. No significant associations were observed between body weight change and consumption of fresh fruits and fruit juices, vegetables, poultry, fish, pasta, and legumes.

**Table 5.** Associations between dietary or lifestyle changes and children's and adolescents' body weight (BW) changes during the first COVID-19 lockdown in Greece. The COV-EAT study.



**Table 5.** *Cont.*

Data are shown as absolute (n) and relative (%) frequencies. \* Non-increased body weight means either stable or decreased. \*\* According to the difference between the values of each lifestyle behavior before and during lockdown, they were categorized as "decrease", "stable, or "increase".

> The results of a multiple, stepwise, backwards, logistic regression analysis of the associations between children's and adolescents' body weight increase and changes of their lifestyle behaviors during the lockdown are shown in Table 6. Based on these findings, increase of consumption of breakfast, salty snacks, and total snacks along with decrease of physical activity were significantly associated with increase of children's and adolescents' body weight.


**Table 6.** Multiple regression analysis between the probability of children's and adolescent's body weight increase and dietary and lifestyle changes during the first COVID-19 lockdown in Greece. The COV-EAT study.

Data are shown as odds ratios (OR), along with 95% confidence intervals (95% CI) and *p*-values.

#### **4. Discussion**

The COV-EAT study is the first study in Greece and one of the few globally that examined changes of lifestyle behaviors in children and adolescents during the first lockdown that was implemented due to COVID-19 and explored their associations with children's/adolescents' body weight gain. The main findings of the present study showed that during the lockdown period, children and adolescents in Greece: (1) increased their consumption of certain foods, such as fruits and fresh juices, vegetables, dairy, pasta, sweets, total snacks, and decreased their fast-food consumption, (2) increased their screen time, (3) increased their sleep duration, (4) decreased their physical activity, and (5) 35% of them gained body weight. According to the results of a multiple regression analysis conducted in this study, increased consumption of breakfast, salty snacks, and total snacks along with decreased physical activity was significantly associated with increase of children's and adolescents' body weight.

The findings of the present study suggest that the COVID-19 lockdown, with the concomitant closure of schools, negatively affected children's lifestyle behaviors, which are some of the predominant risk factors for obesity [3]. In line with the findings of the current study, Pietrobelli et al. showed that during the lockdown, children and adolescents with obesity in Italy significantly increased their consumption of certain foods (chips, red meat, and sugary drinks), their sleep duration, and the time they devoted to screen activities, while they decreased the time they spent in sports [3]. Similarly, Ng et al. in a sample of 1214 Irish adolescents, showed that half of the participants tended to decrease their physical activity during the lockdown, especially those with overweight or obesity [23]. Furthermore, Jia et al. conducted a survey among 10,082 participants from high schools, colleges, and graduate schools (aged 19.8 ± 2.3 years) and showed that individuals' BMI, screen time, and sedentary and sleeping time on weekdays and weekends increased, while the frequency of engaging in active transport, moderate/vigorous-intensity housework, leisure-time moderate/vigorous-intensity physical activity, and leisure-time walking were decreased [24]. Also, the study by Ruiz-Roso et al. in a multinational sample of adolescents from Italy, Spain, Chile, Colombia, and Brazil indicated that families had more time to cook and improved eating habits by increasing legumes, vegetables and fruits intake and reducing fast-food consumption, but that was not enough to increase the overall diet quality, because of the higher sweet food and fried food consumption [25]. Similarly, in

our study, fast-food consumption decreased (*p* < 0.001), which might have resulted from the fear of being affected by the coronavirus that could be transmitted from the person delivering the food. The COV-EAT study was conducted during the first quarantine, where ignorance of the protective measures against COVID-19 was excessive. Studies in adults are also in line with the findings of the COV-EAT study. According to the preliminary results of the ECLB-COVID19 international study in 1047 adults, the COVID-19 lockdown had a negative effect on physical activity and eating behavior and led to a significant increase in sitting time [26].

The alterations in children's and adolescent's lifestyle behaviors may be explained in different ways. The decrease of physical activity may be attributed to home confinement, which does not allow individuals to attend sport clubs and organized physical activity or visit schoolyards, parks, and recreational areas. Ng et al. reported that Irish adolescents with overweight were more likely to be less physically active during the COVID-19 lockdown [23]. In contrast, the increase of screen time may be due to the longer duration of distance learning replacing both school lessons and private lessons, in addition to more free time at home and to boredom. The increase of sleep duration may be linked to the fact that children did not have to go to school in the morning. Changes of eating behavior may be caused by several different factors. First, the insecurity caused by COVID-19 may have led families to change the home food environment and feeding practices. Indeed, Adams et al. reported that families experiencing food insecurity exerted greater pressure to their children to eat, while 30% of the families increased the amount of high-calorie snack foods, desserts/sweets, and fresh foods, and almost half of the study sample increased the availability of non-perishable processed foods in their homes [27]. In addition to the physical environment, the social environment at home may have changed because of COVID-19. In the present study, a number of parents (5.1% of mothers and 2.9% of fathers) lost their job, while others (4.2% of mothers and 3.2% of fathers) experienced an increase of working time during the COVID-19 lockdown. These changes might be associated with parental stress and disturbances of family interactions, parental modeling, and parenting feeding practices at home [28]. Furthermore, the psychological impact of self-isolation may have triggered boredom and stress, which are determinants of the consumption of energy-dense "comfort foods" and emotional eating [5]. Especially children and adolescents with obesity may be more susceptible to overeating, as observed in Polish adults with overweight and obesity [29]. It is also noted that cooking and preparation of new recipes for snacks might be used as a recreational activity by the family during the lockdown, which in turn increases the availability of home-made sweets, snacks, and foods. Additionally, a home does not always provide a steady environment for mealtimes, physical activity, and sleep schedule [3].

Changes of lifestyle behaviors may lead to an increase of energy intake over energy expenditure, a condition which results in body weight gain when lasting for long periods of time. As expected, increased consumption of energy-dense foods (i.e., total snacks, including sweets) and decreased physical activity were associated with an increase of children's and adolescents' body weight. Still, the present study also showed that the increases of breakfast consumption and total snacks were associated with an increase of participants' body weight. These observations may be attributed to the fact that children/adolescents increased the number of meals consumed per day and that, possibly, they consumed unhealthy foods/snacks in these extra meals. Future studies should shed more light on these associations. Moreover, increased body weight in 35% of children and adolescents could be a natural trend due to children's growth if the increase of body weight was about 0.5 kg. The mean body weight increase of 2 kg indicates an abnormal weight gain.

Since obesity and its complications (diabetes, heart disease, pulmonary disease, hypertension, etc.) can worsen the implications of COVID-19, it is critical to implement measures, during the lockdown, to promote healthy eating and physical activity and prevent obesity [3,9,19,30–32]. To achieve these goals, an umbrella of telehealth (e-health and m-health) obesity prevention and treatment actions should be implemented, in addition

to the measures taken to tackle the expansion of COVID-19 [30,33]. Policy interventions to oversee food advertisements and behavioral strategies to promote nutrition education, appetite control, and family meal planning should be applied. Vulnerable groups, such as children and adolescents with overweight or obesity, lower socioeconomic groups, and families with food insecurity should be prioritized.

The findings of the current study should be interpreted under the light of its strengths and limitations. Regarding the strengths, the COV-EAT study was the first study in Greece to explore the potential effect of the first COVID-19 lockdown in Greece on children's and adolescents' lifestyle behaviors, using data from 397 families from urban, semi-urban, and rural areas. Regarding the limitations, firstly due to its cross-sectional design, no causal relationships could be established. Secondly, the study sample was not representative of all children and adolescents in Greece; therefore, the results cannot be generalized to the whole population of Greek children and adolescents. The sampling procedure, which was based on an online survey, may have also produced a selection bias regarding the recruited participants. Moreover, the questionnaire used in the COV-EAT study was not validated, while data were self-reported, and thus subject to recall bias and socially desirable answers, and only weight change was reported. It was also not feasible to conduct comparisons between maternal and paternal reports, which may have an effect on the results. Still, 90% of the reports were taken from mothers, which limits the possibility of such bias.

#### **5. Conclusions**

The COV-EAT study reported unfavorable changes in children's and adolescents' lifestyle behaviors during the first COVID-19 lockdown that was implemented in Greece in spring 2020. Certain lifestyle changes were associated with children's/adolescents' body weight gain. Considering that the COVID-19 pandemic may lead to further lockdowns, effective e-health and m-health strategies and programs to tackle the adoption of unhealthy lifestyle behaviors and prevent excessive body weight gain are urgently needed.

**Supplementary Materials:** The supplementary materials are available online at https://www.mdpi. com/2072-6643/13/3/930/s1.

**Author Contributions:** Conceptualization, O.A. and M.P.; methodology, O.A., statistical analysis, G.C.; data curation, O.A., M.P., and C.G.; writing—original draft preparation, O.A.; writing—review and editing, M.P., C.G., and G.C.; supervision, O.A. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki and the conventions of the Council of Europe on human rights and Biomedicine. The study protocol was approved by the Ethical Committee of the Department of Physical Education and Sport Science in the School of Physical Education, Sport Science, and Dietetics, University of Thessaly (protocol code 1655 and date of approval: 6 June 2020), and registered at clinicaltrials.org (NCT04437121).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. The data are not publicly available due to ethical restrictions.

**Acknowledgments:** The authors would like to thank the study participants.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Review* **The Influence of Parental Dietary Behaviors and Practices on Children's Eating Habits**

**Lubna Mahmood 1, Paloma Flores-Barrantes 1, Luis A. Moreno 1,2,3,4,\*, Yannis Manios 5,6 and Esther M. Gonzalez-Gil 1,4,7**


**Abstract:** Poor dietary habits established during childhood might persist into adulthood, increasing the risk of developing obesity and obesity-related complications such as Type 2 Diabetes Mellitus. It has been found that early modifications in eating habits, especially during childhood, might promote health and decrease the risk of developing diseases during later life. Various studies found a great influence of parental dietary habits on dietary behaviors of their children regardless of demographic characteristics such as gender, age, socioeconomic status and country; however, the exact mechanism is still not clear. Therefore, in this review, we aimed to investigate both parents' and children's dietary behaviors, and to provide evidence for the potential influence of parents' dietary behaviors and practices on certain children's eating habits. Family meals were found to contribute the most in modeling children's dietary habits as they represent an important moment of control and interaction between parents and their children. The parental practices that influenced their children most were role modeling and moderate restriction, suggesting that the increase of parental encouragement and decrease of excessive pressure could have a positive impact in their children's dietary behaviors. This narrative review highlights that parental child-feeding behaviors should receive more attention in research studies as modifiable risk factors, which could help to design future dietary interventions and policies to prevent dietary-related diseases.

**Keywords:** parents; dietary intake; feeding practices; children; family meals; breakfast; snacking habits

#### **1. Introduction**

Obesity is a complex condition influenced by both genetic and environmental factors [1]. Dietary intake has been linked with obesity in terms of volume, composition, meals' frequency, snacking habits and diet quality [2]. Additionally, there is indication that children are likely to maintain their dietary habits into adulthood [2]. Thus, understanding children's eating habits is very important in terms of children's health [3]. There are some factors that could influence children's eating habits such as the home food environment, as well as the social environment, contexts where perceptions, knowledge and eating habits are established [4]. However, parental dietary patterns seem to affect children most,

**Citation:** Mahmood, L.;

Flores-Barrantes, P.; Moreno, L.A.; Manios, Y.; Gonzalez-Gil, E.M. The Influence of Parental Dietary Behaviors and Practices on Children's Eating Habits. *Nutrients* **2021**, *13*, 1138. https://doi.org/10.3390/ nu13041138

Academic Editor: Marilyn Cornelis

Received: 11 February 2021 Accepted: 26 March 2021 Published: 30 March 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

as parents are the ones who shape the home food environment, influence how a child thinks about food, and, accordingly, start forming their own food preferences and eating behavior [4].

Out of the dietary habits, family mealtime becomes the main social context in which children can eat with their parents, who are considered as their main role-models [5]. Sharing meals with children, having breakfast together regularly and encouraging children to have healthy snacks with moderate restrictions have shown positive impacts on children's dietary behaviors [6]. Furthermore, one review study evaluated these practices and found that they were associated with higher consumption of dairy products, fruits and vegetables (FV), along with healthier breakfast patterns among children [7]. Also, the same review stated that encouragement practice gives children a chance of making decisions, whereas the moderate restriction practice help parents to imply clearer instructions to their children. Therefore, it was recommended to use a combination of the two practices, so that both parents and children would have the ability to contribute to determining food choices [8]. In this narrative review, we focus on the effect of parental dietary habits on children's eating behaviors, including family meals, breakfast routine and snacking habits.

#### **2. Method for Literature Search**

Serial literature searches for articles of interest were performed between August and December 2020. PubMed, Scopus, Education Resources Information Center (ERIC), Science Direct and Google scholar databases were searched using the following keywords: "parents", "dietary intake", "feeding practices", "children", "family meals", "breakfast", "snacking habits", "food choices", "food consumption", "role model", "diabetes", "parenting style", "behavior". We included both original researches and review articles published between 2000 and 2020. Studies were eligible if they were published in English and included preschoolers (ages 2–5 years), or school-age children (ages 6–13 years). A total of 2590 studies were identified, 508 duplicates were removed, 455 related titles were chosen, 92 articles met the inclusion criteria and 83 studies were included in this review.

#### **3. Definitions**

"Eating habits" can be defined as the conscious and repetitive way a person eats, and this includes what types of food are eaten, their quantities and timing of consumption, in response to cultural and social influences [9]. On the other hand, "eating behaviors" have been considered as a group of actions starting from a simple food chewing to food shopping, food preparation and food policy decision-making [10]. Food patterns or dietary patterns refer to the quantity, quality and variety of foods and beverages consumed as well as the frequency with which they are habitually consumed, and it refers to the diet as a whole [10]. A balanced diet is characterized by high intake of fresh FV, whole grains, legumes, nuts, fiber, polyunsaturated fatty acids and low in both refined grains as well as saturated fatty acids [11]. However, guidelines may differ in their recommendations regarding the consumption of processed meat and dairy products, probably relating to the national food culture, sustainable food choices and food safety [11].

#### **4. Children's Eating Habits**

Dietary habits from childhood track into adulthood, so understanding children's eating habits is very important in terms of children's health [12]. Nutrition is the main factor of interaction between parents and children, especially during the first year of life, starting by breastfeeding [12]. By the end of the first year of life, children start learning to feed themselves and make the transition to the family diet and meal patterns [12]. A review study that assessed both national and international research articles on child nutrition and eating behaviors concluded that as children switch to the family diet, recommendations from parents address not only food, but also the eating context, which refers to the immediate environment of each eating occasion [12]. Moreover, the same study suggested that a

variety of healthy food items provided to children can promote their diet quality and food acceptance [12].

A study across 11 countries suggested that the nutritional status of children from birth to the age of 2 years was positively associated with dietary variety [13]. Furthermore, exposures to FV in early childhood have been associated with higher acceptance of these foods at later ages [13]. A longitudinal study of 120 2-year-old children and their parents followed for 9 years found that around 25% of children experienced some eating problems such as being hesitant to try new foods or insist on a limited number of food items (no variety), concluding that those problems may lead them to become picky eaters [14].

However, children with eating problems (i.e., picky eaters, meal skippers) may be at risk for behavioral problems, as well as impaired growth and development [14], whereas repeated exposure was found to be the main way for children to recognize the food. Thus, parents are advised to keep introducing food items more than once, and to avoid getting discouraged or giving up [12].

#### **5. Home Food Environment**

The home food environment includes the availability and accessibility of food, as well as other factors such as frequency of eating out, and parents' perception of food costs [15]. In addition, the home food environment was found to have remarkable effects on eating behaviors of parents and their children as most of the food consumed is stored and prepared at home [15]. Although children are likely to be influenced by their home food environment and the community, they may have limited control over it [16]. Results from the Quebec Longitudinal Study of Child Development, which included 1492 children, found that children who had a better family environment, i.e., less family pressure to eat, had low levels of soft-drinks consumption (unstandardized β = −0.43, *p* < 0.001, 95% confidence interval (CI), –0.62 to −0.23), and high level of fitness (unstandardized β = 0.24, *p* < 0.001, 95% CI, 0.12–0.36) [16]. In the same vein, the baseline survey of the Identification and prevention of Dietary- and lifestyle-induced health Effects in Children and infants (IDEFICS) study, which included 1435 families from eight European countries, found that home food environment plays a stronger role in shaping children's intake of healthy foods than unhealthy foods, especially for younger children [17].

As previously mentioned, the home food environment determines what kind of foods are available and accessible to children [15]. While availability and accessibility are often merged into a single construct, the content map presented in Vaughn's study [18] considered them separately because they may have differential effects on children's diet and eating behaviors. Accordingly, these diverse definitions could explain the differences found in the results of studies. Availability is related to the physical presence of food [18], whereas accessibility refers to parental actions to control how easy or difficult it is for children to access food by themselves or with limited assistance [18]. A review about the availability and accessibility of FV at home found that both availability and accessibility were associated with FV consumption among children and adolescents and inversely associated with children's total energy and fat intake [19].

Besides, low consumption of nutrient-poor, energy-dense food items, like sugarsweetened beverages, cookies, packed snacks, food high in saturated/transfat, simple sugars and sodium, were noticed when these items were and were not available at home [19]. However, low-income families seem to have low access to healthy foods and possibly greater access to fast food due to dietary costs, which could explain some of the relationships between Socioeconomic Status (SES) and nutrient density of consumed foods [19].

Frequency of eating out is one of the dietary habits that are most influenced by the household environment [20]. Ready-to-eat and out-of-home (OH) foods include vending machines, take-away, cafes, restaurants, supermarkets and convenience stores [20]. Nowadays, families seem to prepare less food at home and spend more money on foods prepared away from home [20], and food prepared OH tends to be more energy-dense than food prepared at home, particularly in terms of fat and sugar content [20]. In addition, focus

groups among the urban community in the US found that parents desire easy, convenient and tasteful meals that are culturally appropriate and low-cost, while some families may believe that food eaten out is lower in cost and tastier [21]. These beliefs would encourage parents to eat out and thus perpetuate the cycle of decreased home-prepared meals. Consequently, children may have less opportunities to learn culinary skills, have access to healthy diet, or reinforce healthy eating habits [21]. Cross-sectional data from the UK National Diet and Nutrition Survey Rolling Program of 4636 children and adolescents aged 1.5–18 years showed that consuming food prepared outside the home was associated with a greater intake of foods with high levels of fat and sugar in children [20]. Also, a systematic review documented the nutritional characteristics of eating away from home and its relations with the diet quality and energy intake. The results of this review concluded that eating outside the home is associated with lower diet quality and micronutrients intake, like vitamin C, Fe and Ca. However, the conclusion needed further confirmation as the review was based on studies from national surveys from Belgium and the United States only [22]. Similar results were obtained in a cross-sectional study conducted in Japan among 4258 caregivers, where children with obesity had a lower frequency of shared home-made meals, after adjusting for confounding factors.

However, validity and reliability of the questionnaire used to assess the frequency of cooking were not examined [23]. Unfortunately, these studies have only considered the effect of eating out without concerning the effect of ready-to-eat (unhealthy) meals prepared at home.

#### **6. Parenting Styles and Feeding Practices**

In the literature, parenting styles have been defined as psychological constructs representing the more general interactions between parents and children, whereas parental feeding practices includes specific rules or behaviors used by parents to control when, what and how much their children eat [24,25].

It has been previously stated by Horst and Sleddens [26] that according to Baumrind's taxonomy, parenting styles have been divided into three categories: authoritarian, permissive and authoritative. Whereas authoritarian styles are highly demanding but less responsive, permissive styles include less demanding but high responsiveness, and authoritative styles present both demanding and responsive [26].

Studies examining the direct role of parenting styles on children's eating behaviors are limited. However, a recent review of the evidence found that less parental monitoring was presented in the permissive style, whereas more restrictive food and high pressure on children to eat were linked to authoritarian parenting style. On the other hand, preferable parental monitoring of the child's food intake was associated with the authoritative parenting style [27]. Another two systematic reviews concluded that children tend to eat more healthily with a healthy body mass index (BMI) if they raised in authoritative households. However, the effects of these generic parenting styles were generally indirect and weak [28,29].

One review critically summarized previous research on parental feeding practices and found that role models can play a really important part in shaping children's eating habits. Therefore, role modeling behaviors were recommended for parents such as: providing healthy foods, modeling healthy eating and increasing encouragement to eat healthy foods [30]. Results from a study that used the Parental Feeding Style Questionnaire (PFSQ), which included 100 children (aged 2–5) in Hong Kong, showed that encouraging children to consume a variety of foods was associated with healthier eating behaviors, like meal frequency, better food choices and higher intake of fruits (Odd Ratio (OR) = 1.357; 95% confidence interval (CI) = 1.188 to 1.551) and vegetables (OR = 1.335; 95% CI = 1.128 to 1.579) [31]. Whereas, using foods as rewards could increase the child's preferences for these food items. Thus, using unhealthy foods as rewards may promote children's consumption of unhealthy energy-dense palatable foods [31]. Likewise, a cross-sectional study conducted in 17 primary schools in Dunedin city in New Zealand found that through

a good parental role modeling, higher parental diet quality was associated with lower consumption of cakes, chocolate, biscuits and savory dishes in children [32]. A crosssectional study included 13,305 children in nine European countries and found associations (OR between 1.40 and 2.42, *p* < 0.02) between parental role modeling of healthful foods with children's dietary habits, food intake and preferences for fruits and vegetables [33].

The results of these studies highlighted the importance of parental modeling in terms of their dietary behaviors and food choices on the diet of their children. However, parental role modeling studies have employed different methods, with varying validity, to measure children's dietary intake, such as 24 h dietary recalls, food frequency questionnaires, parent report of child dietary intake and child report of parental role modeling. This could explain why correlations between parent and child reports for these studies have also been mixed, whereas studies that have utilized both parent and child report are very limited.

A review study summarizing previous results on parental strategies and practices concluded that a "moderate restriction" could be beneficial as children of moderately restrictive parents were found to consume fewer calories, eat more fruits, and eat less fatty snacks and sweets [34]. Besides, the "prompting and encouragement" feeding practice made by parents could help their children to have healthier dietary habits [34]. The term "moderate restrictions" indicates a careful use of restrictions by parents in which unhealthy food items were gradually decreased and limited rather than being strictly forbidden, whereas the word 'encouragement' refers to the situation when parents offer more types of food with positive messages, but at the same time, children can still make decisions in combination with their parents [31].

On the other hand, restricted parental feeding practice seemed to be related to overeating, especially among preschool-age children [35]. One longitudinal study assessed the maternal influences on picky eating behaviors and diet of 173 9-year-old non-Hispanic white girls [36]. The results of this study suggested that with mothers who were less likely to pressure their children to eat, their children were less likely to be picky eaters or overweight [36]. While, when parents highly restrict energy-dense foods from their children's diet hoping children choose healthful alternatives, children usually increase their desire for it and start to eat when they are not hungry [34]. Therefore, various research studies discourage pressuring practice as it can create a negative family eating environment and make children pickier eaters [37,38].

Evidence suggests that high involvement and role-modeling practices are more favorable for supporting positive food-related behaviors, especially among young children. But unfortunately, these studies cannot be taken as proof of causality. Thus, long-term studies are needed to determine the causal link between parental feeding practices and children's eating habits.

Household food rules is another factor which is usually established by parents to guide youth behaviors and achieve goals for their growth [39]. To explain further, for example, both "limited fast food" and "limited portion sizes at meals" were significantly linked with improved food consumption and weight status [39]. Whereas a rule of "no fried snacks" was positively associated with percent body fat (PBF), however, the link between fat intake, snacking and excess weight was unclear as snack foods are often grouped as one item (e.g., chips, candies, ice cream and cookies) [39]. Besides, the "no snacking while watching television" rule was found to be an effective one as children tend to eat more when they are distracted and eating while watching TV also prolongs the eating period [39]. In a School of Public Health Project, Eating and Activity over Time (EAT) researchers found that children in families who watch TV while eating meals had a lower-quality diet than the children of families who turned the TV off during meals [40]. In the same study, children who watched TV while eating family meals seemed to consume fewer grains and vegetables, and more soft drinks, than those who did not watch TV. Similar results were also found among Australian children in which watching TV was associated with the consumption of energy-dense foods and drinks [41]. However, these studies do not definitively prove

direct causal effects of household food rules on unhealthy food preferences and overall unhealthy diet.

#### **7. Parental Dietary Behaviors Influence on Children's Eating Habits**

Dietary preferences are formed by a combination of a complex interplay of genetic, familiar and environmental factors. However, parents seemed to have a high degree of control in modeling their children's eating behaviors [42]. During the first year of life, children's dietary patterns undergo a rapid evolution since parents are the ones who select the foods of the family and serve as models of eating. Thus, children tend to imitate their parents' behaviors as well as eating habits [42]. As illustrated in Figure 1, children's eating behaviors are affected by social, physical and intra-individual factors. In the family environment, parents establish more than 70% of their children's dietary behaviors by their own intake and the methods followed to socialize their children [42]. To systematically assess the effect of parental dietary patterns on children, several studies have been revised, which summarize how parental eating habits and feeding styles have been significantly associated with children's eating behaviors, food preferences, intake and consumption (Supplementary Table S1).

**Figure 1.** Summary of home/family-related determinants of children's eating habits.

Parental dietary behaviors refer to the passive processes that influence their children's dietary behaviors and food environment [43]. Various cross-sectional studies have indicated the close similarity between parents and children in the intakes of some healthy and unhealthy foods and beverages, as well as dietary composition, especially when more meals are eaten together [44–48]. Although this association has demonstrated that parents' dietary behavior might influence children's intake, these studies cannot be used to conclude causality. Therefore, the process by which parents affect their children's food intake remains largely unclear.

Four focus groups were conducted in Belgium among parents and caregivers showing that the influence of parental practices differs by age. The younger the child, especially at preschool age and first years of primary education, the stronger the role of parental practices [43]. The same study found that children may consider parents' norms and perceptions as a reference for what is appropriate to consume [43].

Various cross-sectional studies found showed a significant positive association and substantial correlation between children's and parent's intake of various foods [49–51]. Thus, parents' eating behaviors have proven to be a part of the whole process of establishing and promoting healthy or unhealthy dietary patterns among children and adolescents [42]. A Parent Mealtime Action Scale (PMAS) was developed among 439 fathers and 541 mothers in the USA to examine the dimensions of mealtime behaviors used by parents on children's diet and weight status. The results showed that parents could be influenced by their environment and culture, which may also affect their food choices, suggesting that their children's dietary patterns and nutritional status may also be altered accordingly [52]. Whereas the same study found that obliging a child to accept healthy food through giving advice only, without eating it themselves, is a dead end in nutrition education [52].

Previous studies concluded that parents' influence is thought to be strongest during childhood, especially in early ages, when parents act as role models, enforcers and providers. Therefore, intervention programs should consider what parents consume as well as the parental influence in terms of what parents feed their children and how they feed them.

#### *7.1. Family Meals*

Family meal has been defined as a meal being shared with family members or when one, or both, of the parents are present [53]. There are differences when analyzing the frequency of family meals: some considered it as having ≥3 and others ≥5 family meals taken weekly [53]. Thus, the lack of specificity and consistency in measuring, analyzing and defining family meals makes it difficult to come up with definite results and to compare results [53].

A systematic review [54] that focused on the effects of family meal frequency and psychosocial consequences in youth concluded that more frequent family meals were inversely associated with disordered eating, violent behaviors and depression in children. Additionally, in the same review, it was found that more frequent family meals were positively associated with an increased self-esteem among children [54]. It is agreed that family meals represent an important moment of both control and interaction in the family [55]. A study of family mealtime characteristics of Australian families with children aged 6 months to 6 years old showed that parents place high value on mealtime when they share meals with their children, which helps children to promote healthy eating behaviors. An important strength of this study was the reliable survey measures, but the used online, self-report surveys can be affected by respondent interpretation bias [55]. The presence of parents during mealtime has been linked to decreased meal skipping and increased consumption of dairy products and FV [43]. Correspondingly, results of the Next Generation Health Study in the US showed a higher FV consumption among children whose parents were eating the same food items and sharing their meals with them. This study included a large, nationally representative and generalizable sample; however, the self-estimation and self-report assessment were susceptible to recall bias [56].

In Scotland, a cohort of young children followed-up for 10 years suggested that determining the characteristics of family mealtime practices is needed to increase diet quality and improve children' eating behaviors, such as reduced access to TV viewing during meals, portion sizes, sitting at a table, besides social engagement between parents and children [57]. Similarly, the Quebec Longitudinal Study of Child Development investigated the effect of frequent family meals on children, and results showed that children who had a better family meal environment at the age of 6 years had lower levels of soft-drinks consumption and higher levels of fitness when they reached 10 years [16]. In the same vein, a Harvard cohort study found that children who eat together with their parents are twice as likely to eat their five servings of FV compared to families who do not share their meals. Moreover, in the same study, family meals seemed to help parents to perform as role models and be considered as an example of polite table manners and healthy eating habits [58]. In addition, results from the same study also showed that shared meals seem

to help in childhood obesity prevention as children tend to eat less when they eat in the presence of their parents [58]. Participants in this study were children of nurses, hence, they all came from highly educated families compared to the general population [58].

One meta-analysis concluded that higher frequency of shared family meals in children and adolescents was significantly associated with a normal body weight and healthier dietary habits when they shared family meals 3 or more times per week [59]. Additionally, home cooking and shared family meals have been considered as a key strategy to promote healthy dietary habits and prevent obesity among children [60,61]. A family meals-focused randomized controlled trial in 160 families of 12-year-old children followed-up for about 5 years. Data were collected at baseline, post-intervention and follow-up, and results indicated that promoting healthy shared family meals could lead to a moderate reduction in excess body weight, especially among young children.

Despite the rigorous design, quality measurement and strong theoretical framework used in this study, the generalizability of study findings is limited [61], while engagement in family meals has been considered as the simplest and easiest independent intervention that could be applied to establish a healthy family environment [61]. Therefore, eating environment should be taken into account as it usually affects family communication, parents and children interactions, what kind of food is served, how much is eaten at meals and frequency and lengths of meals. However, it seems that the specific mechanisms of how family mealtimes influence children's nutritional outcomes are yet unclear and should be investigated.

#### *7.2. Breakfast Routine*

"Breakfast" refers to the first meal of the day, or a meal often eaten in the early morning [62]. The findings from the "Anthropometry, Intake and Energy Balance" (ANIBES) Study [62] reported that around 85% of the Spanish population (9–75 years) were regular breakfast consumers, although one in five adolescents were breakfast skippers. It has also been found in the same study that breakfast provides only 16–19% of the daily energy intake. Among the specific foods, the most commonly consumed breakfast foods among children and teenagers were chocolate, pastries and milk [62]. Additionally, a review studied the benefits of breakfast by involving national dietary survey data from various countries including Spain, the UK, Canada, the USA, Denmark and France. Its results found that a healthy regular breakfast has been associated with improved cognitive health, nutritional status and lower plasma cholesterol levels among children and adolescents [63]. These results were supported by a cross-sectional study conducted among 126 children in four elementary schools in Indonesia. Results from that study found that breakfast habits of children were significantly associated with the parent's breakfast habits [64]. Moreover, in the same study, 23% of fathers and 15.9% of mothers were not having breakfast daily, whereas, 17% of children reported that they are not taking their breakfast because no food was available at home in the morning [64].

One of the most wide-reaching reports is that of the European branch of the World Health Organization, who conducted a health behavior survey of over 200,000 male and female schoolchildren, 11–13 and 15 years of age in 39 European states in 2009/2010 [63]. Overall, 61% of 13-year-olds consumed a breakfast on each school day, while the figure fell to 55% among 15-year-olds. In general, breakfast consumption was most common among boys and declined with lower socio-economic status [63]. These data showed that about half to one third of children do not have breakfast every day, although the data does not reveal the actual frequency of breakfast intake [63]. The report also indicates that regular breakfast consumption is associated with higher intakes of micronutrients, a better diet that includes FV and less frequency of consumption of soft drinks [63]. According to the Health Sponsorship Council (HSC), there are more than 100,000 children worldwide aged 1–5 years missing breakfast at least once per week, while their parents are also skipping this meal. Besides, over 36,000 children worldwide never consume breakfast at home. It has been revealed that children of parents who skip breakfast are more likely to skip

their breakfast, consume more energy-dense nutrient-poor food and are more likely to be overweight [65]. A cross-sectional study including 426 children aged 10–14 years from 4 local schools in Queensland found that skipping breakfast among children was associated with the lack of perceived parental emphasis on consuming breakfast (OR = 3.67, 95% CI: 1.75–7.68) [66].

Another cross-sectional survey conducted among preschoolers aged 2–5 years in Hong Kong showed that most children were having their breakfast daily but less than half of them consumed the recommended number of dairy products and FV [31]. Consequently, these studies suggested that parental breakfast-skipping habits are strongly associated with breakfast skipping among their children. Thus, findings underline the importance of addressing parental habits and their children's in the intervention plan.

#### *7.3. Snacking Habits*

"Snack" has been defined as a small portion of foods or drinks that is taken between regular meals [67]. Another study considered snacks as food items consumed at different times of the day [68]. A study conducted in Spain defined snacking as the process of consuming any food intake outside the three main meals, including mid-morning snack "between breakfast and lunch" and mid-afternoon snack "between lunch and dinner", and nibbling, "disorganized and without defined timing" [69]. The term "snack" seems not to have a static definition [67]. Thus, the impact of snacking is difficult to be assessed due to the variety of its definitions in the literature [67].

In Spain, it has been found that 84.4% of younger and 78.3% of older children were mid-afternoon snack consumers. Specifically, sandwich was the most common food item consumed [69]. Excessive consumption of soft drinks and high-fat-containing snacks and low intake of fruits and vegetables was reported among Mexican children in five Baja California counties [70]. Similar findings were found in a cross-sectional study which involved 109 students and their parents in Milan. Results showed that more than 35% of snacks consumed by school-age children were sweets, 23.8% sugary drinks, 9.4% savory snacks, whereas consumption of nuts, yogurt and fresh fruits was very low [71].

Despite limited data, a systematic review concluded that parents' eating behaviors, whether positive or negative, have an impact on the quality of snacks consumed by their children [72]. Whereas consumptions of lower-quality snacks were associated with increased prevalence of overweight among children [73–75]. Some research studies found that the influence of parents on children's snacking habits depends on the children's life stage and age. For instance, parental influence decreases in the transition from childhood to adolescence [76,77]. The nationally representative surveys of food intake in US children demonstrated a positive association between parents' and children's snack consumption, where children tend to consume more snacks if their parents prefer to have more snacks throughout the day [78]. A cross-sectional study which included 1632 elementary school children in Japan showed that their snacking habits were affected by paternal eating habits, for example, children did not consume vegetables as snacks as it was not offered by their parents. Nonetheless, since data were collected only from children in Takaoka city in Japan, the results may not be generalizable to a global population [79], whereas children's consumptions of FV as snacks were high in homes with greater FV intake among parents as well as FV availability [79]. These results were confirmed by another study which used a Web-based survey among 9842 students in Australia and found that when parents offered more snacks, children consumed more snacks [80]. Another cross-sectional study conducted among 667 students selected from schools in West-Flanders (Belgium) confirmed that parental monitoring and child's eating schedule or routine set by parents were associated with more FV intake among girls (*p* ≤ 0.001, 95% CI: −1.8 to −0.5) and boys (*p* ≤ 0.001, 95% CI: −1.7 to −0.5), and reduced negative eating behaviors such as less unhealthy snacking [81]. Results of comprehensive questionnaires, completed by parents of children aged 4–8 years (*n* = 203) in New Zealand, found that the lack of rules regarding the offering of foods to children was associated with a higher intake of fatty snacks [82].

Based on previous studies, it is suggested that during school age, parents play an important role in the control of children's food intake and food choices. Thus, the whole family is encouraged to be involved in the educational interventions to prevent imbalanced snacking behaviors in children.

#### **8. Conclusions**

Multiple parental factors influence a child's dietary habits and are reciprocally interacting, so they cannot be considered separately. The family environment that surrounds a child's domestic life has an active role in establishing and promoting behaviors that will persist throughout their life. Family meals seem to represent an important moment of both control and interaction, which contributes the most in modeling children's dietary habits. Parents should avoid excessive pressure or restriction as it can create a negative social and emotional experience that could affect children's acceptance of the food. Instead, parents should encourage their children on healthy snacking as well as not to skip their breakfast. This can be achieved through positive and active social modeling as well as moderate restriction. Given the considerable evidence for the strong effect of parents on their children's dietary habits, we believe that parents' child-feeding behaviors should receive more attention in childhood obesity prevention policies. We recommend that parents should be provided with information and guidance on how, as well as what, to feed their children, and these promotion strategies should be particularly aimed at parents' unhealthy eating too so they can improve their diet and so their children will imitate them.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/nu13041138/s1, Table S1: Studies assessing the influence of parental dietary behaviors on children's eating habits.

**Author Contributions:** L.M. completed the literature searches, review and drafted the manuscript. P.F.-B. revised the manuscript. E.M.G.-G., Y.M. and L.A.M. reviewed, edited and approved the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** No new data were created or analyzed in this study. Data sharing is not applicable to this article.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Antenatal Determinants of Childhood Obesity in High-Risk Offspring: Protocol for the DiGest Follow-Up Study**

**Danielle Jones 1,2, Emanuella De Lucia Rolfe 3, Kirsten L. Rennie 3, Linda M. Oude Griep 3, Laura C. Kusinski 2, Deborah J. Hughes 2,5, Soren Brage 1,3, Ken K. Ong 1, Kathryn Beardsall 4,5 and Claire L. Meek 2,5,\***


**Abstract:** Childhood obesity is an area of intense concern internationally and is influenced by events during antenatal and postnatal life. Although pregnancy complications, such as gestational diabetes and large-for-gestational-age birthweight have been associated with increased obesity risk in offspring, very few successful interventions in pregnancy have been identified. We describe a study protocol to identify if a reduced calorie diet in pregnancy can reduce adiposity in children to 3 years of age. The dietary intervention in gestational diabetes (DiGest) study is a randomised, controlled trial of a reduced calorie diet provided by a whole-diet replacement in pregnant women with gestational diabetes. Women receive a weekly dietbox intervention from enrolment until delivery and are blinded to calorie allocation. This follow-up study will assess associations between a reduced calorie diet in pregnancy with offspring adiposity and maternal weight and glycaemia. Anthropometry will be performed in infants and mothers at 3 months, 1, 2 and 3 years post-birth. Glycaemia will be assessed using bloodspot C-peptide in infants and continuous glucose monitoring with HbA1c in mothers. Data regarding maternal glycaemia in pregnancy, maternal nutrition, infant birthweight, offspring feeding behaviour and milk composition will also be collected. The DiGest follow-up study is expected to take 5 years, with recruitment finishing in 2026.

**Keywords:** gestational diabetes mellitus; pregnancy; study protocol; randomised controlled trial; large for gestational age; complex intervention; calorie restriction; maternal weight gain; childhood obesity; adiposity; type 2 diabetes; prevention

#### **1. Introduction**

Gestational diabetes, a common complication of pregnancy, is associated with shortterm and long-term health implications for the baby [1,2]. Infants are commonly largefor-gestational-age at birth (LGA; >90th centile) and have a higher risk of obesity in childhood [2,3]. Unfortunately, very few interventions are available with proven efficacy to reduce the likelihood of childhood obesity in these high-risk children. The early development of obesity in children with existing environmental and genetic susceptibilities to type 2 diabetes is a major public health concern [4].

Events in pregnancy, perinatal and early postnatal periods may be important for future childhood obesity, but are relatively understudied, particularly in specific high-risk populations [5]. Babies born to mothers with gestational diabetes often have multiple risk factors

**Citation:** Jones, D.; De Lucia Rolfe, E.; Rennie, K.L.; Griep, L.M.O.; Kusinski, L.C.; Hughes, D.J.; Brage, S.; Ong, K.K.; Beardsall, K.; Meek, C.L. Antenatal Determinants of Childhood Obesity in High-Risk Offspring: Protocol for the DiGest Follow-Up Study. *Nutrients* **2021**, *13*, 1156. https://doi.org/10.3390/nu13041156

Academic Editor: Louise Brough

Received: 30 January 2021 Accepted: 29 March 2021 Published: 31 March 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

for childhood obesity, which appear to have an additive effect upon risk. Maternal obesity in pregnancy [6,7], maternal excessive gestational weight gain [6], maternal postnatal weight retention [6], exposure to hyperglycaemia in utero [8–10], perinatal complications such as large-for-gestational age at birth [9,11,12], infant formula feeding [13] and increased growth trajectory in early life [14] are all established risk factors for childhood obesity or adiposity and are common features of a pregnancy affected by gestational diabetes.

It is therefore possible that an intervention which addresses maternal weight in pregnancy may reduce obesity rates in offspring. Family interventions (which target at least one parent to improve obesity rates in children) are already well-established in the prevention of childhood obesity [5]. However, many interventions to reduce the risk of childhood obesity target older children (2–10 years) and may miss the opportunity to intervene in early years [5,15]. The DiGest study, a currently ongoing dietary intervention in pregnant diagnosed with gestational diabetes, provides the opportunity to study the influence of a pregnancy weight intervention upon risk factors for childhood obesity in early life [16].

The DiGest Study is a randomised, double-blind controlled trial of a reduced calorie diet using a novel dietary intervention to assess the benefits of controlling maternal weight gain in late pregnancy in gestational diabetes. The trial is described in detail elsewhere [16]. Briefly, women are randomised to receive a weekly dietbox containing all meals and snacks and are blinded to the overall calorie content (1200 kcal/day for intervention group and 2000 kcal/day for control group; 40% carbohydrate, 25% protein, 35% fat). The dietbox commences at enrolment, typically 28–32 weeks' gestation, and continues until delivery of the infant. The clinical care team and research team are also blinded to calorie allocation. Dietboxes are nutritionally balanced and low in glycaemic index, low in saturated fat, high in vegetables and protein and suitable for use in pregnancy. Data will be collected to assess the impact upon maternal weight gain, infant birthweight and a range of obstetric and glycaemic outcomes during late pregnancy up to 3 months postnatally [16]. The design of the DiGest trial provides the opportunity for a controlled and blinded dietary study and reduces potential bias due to differences in maternal educational level, cooking ability, income and kitchen facilities.

In this manuscript we describe a follow up study to the DiGest trial which investigates the effect of the reduced calorie dietary intervention in pregnant women upon the development of obesity in a high-risk population of children from birth to 3 years of age. The hypothesis is that a reduced calorie diet in late pregnancy in women with gestational diabetes reduces offspring adiposity and improves maternal weight at 1, 2 and 3 years postpartum.

#### **2. Materials and Methods**

Study design and ethical approval: The DiGest Follow up study is an observational study on the effects of a multicentre, prospective, randomised double-blind controlled dietary intervention trial conducted in late pregnancy. In summary, participants of the follow up study will have been exposed to either the intervention diet of 1200 kcal/day, or the control diet of 2000 kcal/day as part of the DiGest study. Macronutrient ratios were identical for each diet; 40% carbohydrate, 25% protein, 35% fat. Meals were prepared from the same recipes, with a factor of 1.667 used to convert portion size to obtain meals of two different sizes. This diet it provided from enrolment (typically 28–32 weeks' gestation) to delivery of the infant. Participants will have attended 4 study visits in total to provide blood samples, blood pressure, body weight and anthropometry measurements, and to complete a series of questionnaires. Randomisation for the original DiGest study was stratified for centre. Throughout the DiGest intervention and Follow-up study, both mothers and children will receive standard NHS care, as described in the NICE guidelines [17]. The study is being conducted in accordance with the Declaration of Helsinki, and the protocol has been submitted to the Research Ethics Committee (UK Bloomsbury REC 21/PR/0213) and the NHS Health Research Authority (IRAS 281062).

Recruitment: The DiGest trial recruitment occurs at 5 hospital trusts in East Anglia, UK. The same study sites will be used for the follow-up study. At 3 months postpartum, DiGest trial participants will be given information about the follow-up study and invited to participate by the research midwives, nurses, clinical research staff or by their physician/obstetrician. For training purposes, students in healthcare disciplines (e.g., medicine, biomedical science, nursing, midwifery) may also occasionally recruit patients under appropriate supervision. Informed consent will be obtained at the final visit of the DiGest dietary intervention, with the mothers providing consent on their infant's behalf. Participants (or mother-infant dyad) can withdraw from the study at any time without reason without affecting their clinical care. There is no financial incentive for this study, but a small token of appreciation is provided for the child at each visit in line with guidelines of the Royal College of Paediatrics and Child Health [18].

Eligibility criteria: All women from the DiGest cohort (confirmed gestational diabetes and BMI 25 kg/m<sup>2</sup> or above at enrolment) are eligible to enrol in the follow-up study, however, they must be recruited within 12 months of the baby's birth. Mothers would be excluded if they are unwilling or unable to provide informed consent, if they experienced stillbirth, neonatal death or had an infant born with severe congenital anomaly.

Follow-up visit structure: The study timeline is outlined in Figure 1. Study visits will be carried out in the participants' home, local hospital or at another place convenient for the participant. The initial follow-up visit will coincide with the final DiGest visit at 3 months after the birth, where the consent form will be signed for both the mother and infant. Maternal and infant anthropometry will be measured at this visit as part of the DiGest study. Further Follow-up visits will take place at 1, 2 and 3 years postnatally and will take approximately 45 min.

**Figure 1.** Summary of protocol and links between the DiGest trial and the follow up study.

Anthropometry Measurements: At all visits, maternal height and weight will be measured using a routinely calibrated stadiometer and weight scale (Seca Hammer Steindamm, Birmingham, U.K.). Waist and hip circumference will be measured to the nearest 0.1 cm with a fibreglass tape, in accordance with the World Health Organisation criteria [19]. Waist circumference is located at the midpoint between the lowest palpable rib and the iliac crest. Hip circumference is measured at the greater trochanters, or at the widest extension of the

buttocks. Other maternal anthropometry that will be measured include mid upper arm circumference and skinfold thickness, using Harpenden calipers recorded to the nearest 0.2 mm. Infant length and weight will be taken in a supine position, measured to the nearest 0.1 cm using fibreglass tape, and 0.01 kg using scales (SECA 757 Infant digital scale, Seca, Birmingham, UK). Infant abdominal circumference, head circumference, skinfold thickness (Holtain calipers, Crosswell, Wales, U.K.), mid upper arm circumference will also be measured by trained research staff according to methods described elsewhere [16]. All equipment used to measure anthropometry are routinely calibrated.

Maternal Glucose Assessment: Due to the COVID-19 pandemic, a home-based OGTT using continuous glucose monitoring (CGM) will replace the gold standard OGTT for assessment of maternal glucose tolerance postnatally. An HbA1c will also be performed to replace mothers' annual diabetes check in primary care. We have previously assessed the feasibility and efficacy of the home-based OGTT with good results (Kusinski et al., submitted to press). In brief, a Dexcom G6 CGM sensor is sited during the study visit with a masked receiver so participants do not see their glucose results in real time. On day 3, participants are asked to eat normally, and fast overnight for at least 10 h. On the morning of day 4, at 09.00, participants are asked to drink a sachet of Rapilose (Galen, Craigavon, UK) containing 75 g of anhydrous glucose. Participants can have sips of water but are asked to consume no other foods or drinks for 3 h after the test. The timing of the home OGTT is chosen to coincide with peak sensor accuracy. Glucose readings are taken automatically every 5 min and transmit to the CGM receiver. Results from the OGTT at 0, 1 and 2 h are included in the analysis. Other CGM metrics will also be used to assess glycaemia as described in a recent CGM consensus statement. CGM metrics will be reported using both adult non-pregnant and pregnant ranges to allow comparison with pregnancy data gathered in the DiGest trial (also using a Dexcom G6 system).

#### *Physical Activity Assessment*

Participants will be asked to wear a wrist-worn accelerometer continuously for 7 days concurrently with the CGM. The triaxial accelerometer is waterproof and does not have a visual display, nor any auditory or vibrational cues, which means that participants will not be able to influence their activity level based on what is recorded by the device and nor will they be prompted to move about during periods of inactivity. These accelerometers have been used in in women during and after pregnancy to assess their daily physical activity with high compliance and produce reliable estimates of energy expenditure, overall physical activity and moderate-vigorous intensity activity [20–22]. Accelerometry data at 100 Hz will be collected and downloaded from the monitors for analysis. At the end of the recording period, mothers will be asked to complete the Recent Physical Activity Questionnaire (RPAQ), a self-completion questionnaire designed to assess an individual's physical activity over the previous four weeks. The questionnaire contains questions about physical activity in four domains: at home, at work, commuting and during leisure time. RPAQ has been validated against doubly labelled water and individually calibrated heart rate and movement sensing to assess physical activity energy expenditure (PAEE) in adults [23,24]. It has been used in diabetes prevention trials [25] and in longitudinal studies of pregnant women [26].

Other Biochemistry samples: A blood spot sample will be taken from mothers and frozen at −80 ◦C for future batch analysis of C-peptide and metabolomics. An optional heelprick blood spot will also be taken from infants, for future batch analysis of C-peptide and metabolomics. If a genetic sample has not been taken already as part of the DiGest trial, a cheek swab will be taken from both mothers and infants. Mothers will be asked to provide a sample of milk (formula or breast) which will be collected onto filter paper for assessment of infant nutrition including lipidomic profiling. To protect participant's privacy, this can be performed after the visit.

Questionnaires: Mothers will be asked to complete validated questionnaires about quality of life (EuroQuol EQ5D), eating behaviour (three factor eating questionnaireTFEQ-18) [27], physical activity (RPAQ) [24] and web-based multiple pass 24 h dietary recalls to assess habitual dietary intake (Intake24; [28,29]). These questionnaires have been used during the DiGest trial and participants will be familiar with them. In addition, mothers will be asked to complete questionnaires about parental feeding style (PFSQ) and their baby or child's eating behaviour (CEBQ) [30–33]. Information will be collected about infant feeding choice and if relevant, duration of breastfeeding.

#### **3. Results**

The aim of the study is to investigate the effects of a reduced calorie diet in late pregnancy in women diagnosed with gestational diabetes upon longer-term maternal and offspring metabolic outcomes. The primary outcome for child health is standardised weight at 1, 2 and 3 years of age. The primary outcome for the maternal population is maternal weight at 1, 2 and 3 years postpartum.

Offspring secondary outcomes at 1, 2 and 3 years of age: There are multiple secondary outcomes for children including weight, BMI, growth trajectory, and blood spot biomarkers such as C-peptide or metabolomics at 1, 2 and 3 years. Questionnaire data will be assessed to identify effects of the intervention in pregnancy upon child eating behaviour, with assessment for confounding factors including maternal BMI, maternal eating behaviour and parental feeding style.

Maternal secondary outcomes at 1, 2 and 3 years postpartum: Maternal outcomes to be studied include maternal weight and weight change, BMI, anthropometry measures of adiposity, glycaemia (CGM metrics, HbA1c, OGTT results, indices of insulin production and sensitivity, including HOMA-IR and HOMA-B, Matsuda score and Stumvoll index [34,35], cardiometabolic health (blood pressure, heart rate, lipids, fasting insulin, fasting glucose), maternal food intake, food nutritional content and quality, eating behaviour, quality of life, and incidence of type 2 diabetes or gestational diabetes in a future pregnancy.

Analysis Plan: An intention to treat analysis of the primary outcome for child health (standardised weight at 1, 2 and 3 years of age) will be based on linear regression with adjustment for the stratification variable of study centre through a fixed effects model. The potential role of other explanatory variables such as pre-pregnancy BMI, infant nutrition, infant postnatal growth trajectory or information from the questionnaires will be investigated. A per protocol analysis will also be performed in participants with >80% compliance and at least 4 weeks' exposure to the intervention. Secondary outcomes will also be examined through regression analyses (linear or logistic) appropriate for the type of outcome being considered.

Power calculation: All eligible women and their infants will be invited to join the follow-up study. However, calculations are based on assuming a 50% recruitment rate (*n* = 250 women and their infants) and a 20% withdrawal rate. For the maternal primary endpoint, data from earlier work suggest that typical values for maternal BMI outside of pregnancy in women with a history of gestational diabetes is mean 28.7 kg/m2 (SD 7.1; *n* = 416) and maternal postpartum HbA1c 37.5 mmol/mol (SD 7.5; *n* = 157) [36]. Using these figures, recruitment of 250 women, will give 90% power to identify a 3 kg/m2 difference in BMI (e.g., 29 vs. 32 kg/m2) and a 3 mmol/mol difference in HbA1c postnatally while allowing for a 10–20% withdrawal rate. At 80% power, this sample size is sufficient to identify a 2 kg/m2 difference in BMI (e.g., 30 vs. 32 kg/m2) and a 2 mmol/mol difference in HbA1c postnatally.

For the offspring primary endpoint, assessment of infant weight will be based upon z-(SD) scores. At the sample size of 250 infants, there will be 90% power to identify a 0.45 SD increase in weight with 80% power to identify 0.4 SD increase in weight. At the age of 2 years old, a z-score of 0.4 is equivalent to 0.5 kg.

#### **4. Discussion**

This follow-up study of the DiGest randomised controlled trial provides a unique opportunity to assess the potential benefits of a dietary intervention in late pregnancy upon the development of obesity in children with multiple risk factors. The availability of data from mid pregnancy until the age of 3 years also allows detailed characterisation of the relative importance of pregnancy and postnatal risk factors in the development of adiposity in early childhood.

Rates of maternal obesity are increasing in the antenatal population throughout the world, and pre-pregnancy BMI is a strong predictor of both birthweight and future childhood obesity. A recent metanalysis identified that maternal obesity was significantly associated with overweight/obesity in early, mid and late childhood with odds ratios 2.43, 3.12 and 4.47, respectively [6]. Weight gain in pregnancy is also important and has repercussions for women's BMI for 15 years or more after the pregnancy [37]. Landon and colleagues found that gestational weight gain was strongly related to obesity in children aged 5–10 years old [10].

In addition to the effects of maternal obesity, exposure to intrauterine hyperglycaemia appears to further increase the risk of childhood obesity. There is evidence that maternal glycaemia in gestational diabetes is associated with childhood obesity at 10–14 years [2] and altered anthropometry at 5–10 years, favouring obesity [10]. Maternal hyperglycaemia can also indirectly increase childhood obesity rates, by increasing the risk of LGA in offspring. Data from the UK and Canada suggest that childhood obesity rates in LGA infants are at least twice that of children born appropriate for gestational age [11,38]. The exact mechanisms behind these intrauterine exposures and later life obesity are unclear. It is possible that altered placental secretary function, offspring hyperinsulinism and genetic susceptibilities all play a role.

The design of the DiGest and DiGest follow-up studies also allows longitudinal assessment of the effects of other pregnancy exposures upon longer-term offspring growth and health. For example, metformin use in pregnancy has been associated with lower birth weight but increased postnatal catch-up growth, but the consequences of this upon longer-term offspring cardiometabolic outcomes remain less clear [39,40]. The collection of anthropometric measures in offspring exposed to metformin in utero with paired blood samples, and a comparable unexposed control group, provides opportunity to explore this issue in greater depth.

Serum and cord blood stored for biomarkers such as leptin, adiponectin and placental hormones provides opportunities to identify infants at an earlier stage who are at risk of obesity in childhood. Previous work has demonstrated that cord blood leptin levels are associated with pregnancy diet, physical activity and neonatal body composition in a comparable population [41,42]. Cord blood adiponectin has also been associated with body composition effects which may be distinct in male and female neonates [43] and may additionally provide information about neonatal beta cell function [44]. Placental growth factors and metabolic function have also shown relevance for pregnancy outcomes [45,46]. Taken together, it is feasible that biomarkers in cord blood or maternal serum may facilitate early identification of offspring at risk of obesity and diabetes in later life, who could be prioritized for health interventions.

Although maternal physical activity levels in pregnancy and postpartum are likely to be vital for determining offspring habitual exercise levels, relatively few modifiable factors have been identified in children's physical activity levels in the very young [47,48]. Findings to date suggest that parents' physical activity levels are associated with children's activity levels in pre-school aged children and role-modelling by mothers appears to be one of the strongest associations [48]. However, relatively few studies have examined exercise after gestational diabetes in mothers and children. The DiGest Follow-Up study uses both questionnaires and accelerometers to assess physical activity, information which could inform future interventional studies.

Infant feeding and growth trajectory in the first year of life are also important. Although randomised studies of feeding modality in early life are not possible, observational analyses in unselected populations suggest consistent benefits of breastfeeding upon rates of childhood obesity [49,50]. There is also evidence that breastfeeding reduces childhood

obesity risk in offspring of mothers with gestational diabetes and obesity [13,51]. Stettler and colleagues reported that similar benefits may persist until adulthood in a study of offspring to age 20 years [52]. The study also includes questionnaires about child eating behaviour, child food preferences and parental feeding style to examine behavioural associations with obesity and feeding behaviour in children aged up to 3 years.

The aetiology of childhood obesity is therefore complex and multifactorial. In infants of mothers with gestational diabetes, multiple risk factors are often evident at birth. Successful interventions are urgently needed to reduce the risk of obesity and future metabolic disease in these high-risk children.

#### **5. Conclusions**

The DiGest follow-up study provides the opportunity to assess pregnancy and postnatal risk factors for the development of childhood obesity, and to describe the potential impact of a dietary intervention in pregnancy. Early intervention in offspring with existing environmental and genetic susceptibilities to type 2 diabetes will be vital to break the intergenerational cycle of obesity.

#### **6. Patents**

No patents are relevant to the study described in this manuscript.

**Author Contributions:** C.L.M. designed the study, developed the methodology and wrote and revised the manuscript. C.L.M. acquired funding for the study and has overall responsibility for study conduct. D.J. contributed to writing the manuscript and read and revised the final manuscript. E.D.L.R., K.L.R., L.M.O.G., L.C.K., S.B., D.J.H., K.K.O. and K.B. contributed to aspects of study design, and read and revised the final manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This paper presents a study protocol for a follow-up study which was funded by the European Foundation for the Study of Diabetes and Novo Nordisk Foundation through the Future Leaders' Award (NNF19SA058974). The DiGest trial is funded by Diabetes UK (17/0005712).

**Institutional Review Board Statement:** The study will be conducted according to the guidelines of the Declaration of Helsinki, and is under review by the Bloomsbury Research Ethics Committee (protocol v.1; REC 21/PR/0213 and date 3/3/2021).

**Informed Consent Statement:** Informed consent will be obtained from all subjects involved in the study. Consent for infants was given by parents.

**Data Availability Statement:** No applicable.

**Acknowledgments:** We thank the National Institute of Health Research (NIHR) Clinical Research Network (CRN Eastern) for supporting research personnel at study sites who are involved in performing this research study. We are grateful to funding to E.D.L.R., K.R. and L.O.G., who are supported by the NIHR Cambridge Biomedical Research Centre (IS-BRC-1215-20014). Thank you also to Søren Brage for constructive comments and input to the physical activity aspects to the study.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **Complementary Feeding and Overweight in European Preschoolers: The ToyBox-Study**

**Natalya Usheva 1,\*, Sonya Galcheva 2, Greet Cardon 3, Marieke De Craemer 4,5, Odysseas Androutsos 6, Aneta Kotowska 7, Piotr Socha 7, Berthold V. Koletzko 8, Luis A. Moreno 9, Violeta Iotova 2, Yannis Manios <sup>10</sup> and on behalf of the ToyBox-study Group †**


**Abstract:** Complementary feeding (CF) should start between 4–6 months of age to ensure infants' growth but is also linked to childhood obesity. This study aimed to investigate the association of the timing of CF, breastfeeding and overweight in preschool children. Infant-feeding practices were self-reported in 2012 via a validated questionnaire by >7500 parents from six European countries participating in the ToyBox-study. The proportion of children who received breast milk and CF at 4–6 months was 51.2%. There was a positive association between timing of solid food (SF) introduction and duration of breastfeeding, as well as socioeconomic status and a negative association with smoking throughout pregnancy (*p* < 0.005). No significant risk to become overweight was observed among preschoolers who were introduced to SF at 1–3 months of age compared to those introduced at 4–6 months regardless of the type of milk feeding. Similarly, no significant association was observed between the early introduction of SF and risk for overweight in preschoolers who were breastfed for ≥4 months or were formula-fed. The study did not identify any significant association between the timing of introducing SF and obesity in childhood. It is likely that other factors than timing of SF introduction may have impact on childhood obesity.

**Keywords:** complementary feeding; solid food; breastfeeding; overweight; obesity

#### **1. Introduction**

Obesity is an increasing worldwide problem with an estimate of 340 million overweight or obese children and adolescents aged 5–19 in 2016 and 38.2 million children under the age of 5 years being overweight or obese worldwide in 2019. Moreover, health expenditures for the adult population are constantly increasing, €70 billion per year in Europe (2017) and \$342.2 billion in the US (2013). Direct medical costs related to childhood

**Citation:** Usheva, N.; Galcheva, S.; Cardon, G.; De Craemer, M.; Androutsos, O.; Kotowska, A.; Socha, P.; Koletzko, B.V.; Moreno, L.A.; Iotova, V.; et al. Complementary Feeding and Overweight in European Preschoolers: The ToyBox-Study. *Nutrients* **2021**, *13*, 1199. https:// doi.org/10.3390/nu13041199

Academic Editors: Maria Luz Fernandez and Jennifer T. Smilowitz

Received: 31 January 2021 Accepted: 30 March 2021 Published: 5 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

obesity alone were approximately \$14 billion in 2013 and they are expected to rise significantly, especially because today's obese children are likely to become tomorrow's obese adults [1–5].

One of the risk factors for childhood obesity is inappropriate nutrition during infancy. The advantages of exclusive breastfeeding (EBF) compared to partial breastfeeding in the first months of life have been recognized. The World Health Organization's global public health recommendations promote "exclusive breastfeeding for 6 months" with continued breastfeeding up to the age of two years or beyond. Complementary feeding (CF) should occur when a baby is both developmentally ready and when breast milk is no longer able to fulfil the nutritional requirements of the child [6,7]. The recommended period for starting CF as stated by the European Society for Pediatric Gastroenterology, Hepatology and Nutrition is at week 17–26 (between the beginning of the fifth month and the beginning of the seventh month of life) [8].

The protective role of breastfeeding (BF) against overweight and obesity has been reported in many studies, showing a greater effect with longer duration of breastfeeding [9–11]. However, the association of timing of CF introduction and the quantity and quality of CF with childhood obesity is controversial [12–18]. Some studies reported an association of very early CF introduction before four months of age with later obesity in formula-fed infants whereas there was little effect in breastfed infants [19]; overweight and obesity at 2–12 years; obesity at three years [20,21]. An association between early CF introduction and overweight in children aged 1–17 has also been reported as being modified by the duration of breastfeeding in a birth cohort study in the Netherlands (Prevention and Incidence of Asthma and Mite Allergy—PIAMA) [22]. These findings are not consistent with the conclusions of the systematic review by Pearce et al. who found no consistent association between very early introduction of CF (prior to the age of four months) and childhood body mass index (BMI) [23].

Since there is no consensus yet on a possible relationship between introduction of solid foods (SF) and overweight in childhood, the aim of this study was to investigate the association between the timing of CF, breastfeeding status and overweight in a large pan-European sample of preschool children.

#### **2. Materials and Methods**

The ToyBox-study was conducted between May and June 2012 in six European countries (Belgium, Bulgaria, Germany, Greece, Poland and Spain) among parents/caregivers of preschoolers born between January 2007 and December 2008. The ToyBox-study (www.toybox-study.eu; accessed on 20 September 2020) adhered to the Declaration of Helsinki and the conventions of the Council of Europe on human rights and biomedicine. All the countries (Belgium, Bulgaria, Germany, Greece, Poland and Spain) obtained ethical clearance from the relevant ethics committees and local authorities and all parents/caregivers provided a signed consent form before being enrolled in the study. Detailed information about the ToyBox-study design has been previously reported [24].

Data about perinatal information of preschoolers (including anthropometric measurements at birth, breastfeeding and complementary feeding practices during the first year of life) were obtained through a standardized self-administered questionnaire for primary caregivers. They also aimed to include sociodemographic characteristics of the participants. The questions about children's nutrition during the first year of life were formulated with a focus on the presence/absence of breastfeeding at each month after birth and the age at which water, tea, juice, formula milk and solid/semi-solid foods were introduced. For the purpose of limiting the recall bias, parents/caregivers were advised to use the child's medical records for the questions in the perinatal section, resulting in an excellent value of ICC (intraclass correlation coefficient) in the test–retest reliability study—0.75, whereas the questions on parental weight and height had "moderate-to-excellent reliability" (i.e., ICC ranged from 0.489 to 0.911) [25]. Data on health-related behaviors (dietary habits, physical activity and sedentary behavior) of preschoolers and their parents were collected

using validated questionnaires and the results regarding this topic are presented in other papers [26,27].

Family socioeconomic status (SES) was categorized according to maternal years of education as "low SES" (≤12 years), "medium SES" (13–16 years) and "high SES" (≥16 years of education). Preterm birth was defined as <37 gestational weeks, full-term birth—as ≥37 gestational weeks. Breastfeeding status was defined according to the WHO indicators [28]. Exclusive breastfeeding indicated breastfeeding with no other food or liquid given, except for medical drops and syrups (vitamins, minerals, medicines). Predominant breastfeeding applied if the infant received additional water or water-based liquids. The inclusion of other milks and foods (formula milk and/or semi-solids) was considered partial breastfeeding. The ever breastfed rate was the proportion of infants aged less than 12 months who were ever breastfed. Complementary feeding included liquids and SF (fruit juice, fruits, vegetables, meat, fish, eggs, milk products, creams and soup). Timely complementary feeding rate was defined as the proportion of infants 4–6 months of age who received breast milk and CF. Children without any information about feeding in the first two months (*n* = 454) and without information about the time of solid foods' introduction (*n* = 300) were excluded from the analyzed study sample (*n* = 7554). Both the analyzed and the excluded samples have similar distribution by country and participating status (intervention or control groups).

#### *2.1. Anthropometric Data*

Children's weight (to the nearest 0.1 kg) and height (to the nearest 0.1 cm) were measured using a standardized protocol and standardized equipment which was calibrated before and during the period of data collection [29]. All measurements were taken by research assistants who were thoroughly trained before the initiation of the study to achieve very good intra- and interobserver reliability agreement [30]. Overweight including obesity was defined on the basis of the WHO criteria as BMI z-score > 2 standard deviations (SD) and BMI z-score > 3 SDs, respectively, for children aged < 5 years. For children aged > 5 years, overweight was defined as BMI z-score > 1 SD and obesity as BMI z-score > 2 SDs. Calculation of the ponderal index (PI = weight/height3) was used for assessment of the weight status of children at birth, with a PI range of 2.0–3.0 g/cm3 considered normal. Children with a PI > 3.0 were considered overweight, and those with PI < 2.0 were classified as small for gestational age (SGA). Parental weight and height were self-reported by parents/caregivers and their BMI was calculated. Parents/caregivers were categorized according to their BMI as "under-/normal weight" (≤24.9 kg/m2), "overweight" (≥25 and ≤29.9 kg/m2) or "obese" (≥30 kg/m2) [31].

#### *2.2. Statistical Analyses*

Normal distribution of variables was tested with Shapiro–Wilk tests. Continuous variables are presented as the means ± standard deviation in case they were normally distributed (e.g., age of preschoolers, age of mothers, introduction of solid foods) and as the medians and IQR (interquartile range) for non-normally distributed variables (duration of breastfeeding, introduction of tea, introduction of fruit juices). Statistical analysis of parameters' distribution of the original samples by country was not applied as their number was small (*n* = 6). Post-sampling or bootstrapping were not considered.

Categorical variables were analyzed using the χ<sup>2</sup> test regarding country and children's BMI categories. An independent samples *t*-test was applied for comparison of the means and percentages from two samples while the one-way ANOVA (analysis of variance) for the means of more than two samples (birth weight; mother's age).

Comparison of the medians was performed using the median test. The association of feeding practices and children's BMI as well as mother's characteristics was determined by the correlation analysis. Logistic regression analysis with 95% confidence intervals (CIs) was used to estimate the odds of being overweight/obese (dependent variable) in relation to different infant-feeding practices. The results were adjusted for mother's age and BMI before pregnancy, SES, smoking habits during pregnancy and country. In order to quantify the probability of complying with current recommendations for the introduction of CF at 4–6 months of age, logistic regression analysis was performed and adjusted for mother's age and BMI before pregnancy, SES, smoking habits during pregnancy and country. Compliance to recommendations for introduction of CF at 4–6 months of age (yes/no) was considered as a dependent variable.

In the logistic regression models, the variables were selected based on their relevance for the research topic and being tested for absence of collinearity, hence the presented model coefficients correspond to variables with no significant impact as well. Thus, we can reach a conclusion about the existence of meaningful links. The regression analyses of the current data were targeted at identification of the relevant links between the study variables, but not at constituting a universal model which may be applied for analysis of other populations or for establishing new theories. The data were analyzed using the Statistical Package for Social Sciences (IBM SPSS v. 20, Chicago, IL, USA). The level of significance was set at *p* < 0.05.

#### **3. Results**

The total number of analyzed eligible questionnaires from the six countries was 6800 (mean age of the participants, 4.75 ± 0.43 years; 47.7% girls with no statistically significant difference in gender distribution between the participating countries). The sociodemographic characteristics of responders are presented in Table 1. Additional information regarding characteristics of the ToyBox-study sample were previously reported [24,31].

**Table 1.** Characteristics of participants by country (\* χ<sup>2</sup> test; \*\* ANOVA).



**Table 1.** *Cont.*

Tea (chamomile and other types of tea, especially for baby colics) was the first CF for most of the children in our study, introduced at a median age of three months (IQR, 2–5 months), resulting in a low proportion of exclusively breastfed children at four months of age (Table 2).

**Table 2.** Infant feeding practices among pre-school children from the six countries, participating in the ToyBox-study.


\* Introduction of solid foods is significantly different with exception of the following comparisons: Greece and Spain (*p* = 0.13); Greece and Poland (*p* = 0.9); Poland and Spain (*p* = 0.18) (ANOVA); † *p*-value of the χ<sup>2</sup> test; EBF—exclusive breastfeeding; SF—solid foods; IQR—interquartile range.

> In the study sample, the median introduction to fruit juices was at six months of age (IQR, 5–8 months), with the earliest introduction being among Bulgarian children (median, four months of age; IQR, 3–6 months). In the total sample, the proportion of 4–6-month-old infants who received breast milk and CF (timely complementary feeding rate) was 51.2% (35.4%, Belgians; 50.4%, Spanish; 53.7%, Polish; 67.8%, Bulgarians; 72.3%, Germans (*p* < 0.01)). The median age of SF introduction was six months (IQR, 5–6 months), with the earliest introduction in Belgium (median, four months of age, IQR, 4–5 months). Some 26.6% (*n* = 1806) introduced CF outside of the recommended age range, with 4.1% (*n* = 279) before the 16th postnatal week and 22.5% (*n* = 1527) after the 25th week. The

time of CF introduction was correlated with breastfeeding duration (Spearman's ρ = 0.2; *p* < 0.001). There was a weak positive relationship between the introduction of CF and SES (Spearman's ρ = 0.08; *p* < 0.001) and a negative relationship with smoking during pregnancy (Pearson's r = –0.04; *p* = 0.003). Stratifying by country, a negative relationship with smoking during pregnancy was identified only in the Bulgarian sample (Spearman's ρ = 0.12; *p* < 0.001).

The prevalence of overweight and obesity according to the WHO I criteria was 8.0% (*n* = 542) and 2.8% (*n* = 190), respectively (Table 1). Infant-feeding practices showed a different relationship to the prevalence of overweight and obesity at different stages of childhood. Timely introduction of CF at 4–6 months of age had a negative association with the prevalence of overweight and obesity at six and 12 months of age, with no differences between breastfed and non-breastfed children (*p* < 0.05) (Table 3).

**Table 3.** Breastfeeding practices and weight status of children (χ<sup>2</sup> and independent samples *t*-test).


EBF—exclusive breastfeeding; significant comparisons of the prevalence of overweight and obesity (independent samples *t*-test) at 6 months of age: \* introduction of CF (complementary foods) at 4–6 and ≥ 7 months of age—t = 2.71; *p* < 0.01; at 12 months of age: \* introduction of CF at 4–6 and 0–3 months of age—t = 4.32; *p* < 0.001; pre-school age: \* introduction of CF at 4–6 and 0–3 months of age—t = 2.98; *p* < 0.01.

> On the country level, the significant difference in the prevalence of overweight and obesity at preschool age was observed only in two countries—late CF introduction (after seven months) compared to earlier introduction is related to higher prevalence of obesity in Belgium (*p* < 0.001) and to higher prevalence of overweight in Poland (*p* < 0.05) (Table 4).

**Table 4.** Breastfeeding practices and weight status of preschool children by country (χ2).



**Table 4.** *Cont.*

Significant comparisons of the prevalence of overweight and obesity in preschool age (independent samples *t*-test): Belguim: obesity; \* introduction of CF (complementary foods) at 0–3 and 7–12 months of age (t = 2.06; *p* < 0.02); \*\* introduction of CF at 4–6 and 7–12 months of age (t = 2.27; *p* < 0.001); Poland: overweight; \* introduction of CF at 4–6 and ≥ 7 months of age (t = 2.12; *p* = 0.03).

> Table 5 presents results of the logistic regression analysis identifying one single risk factor connected to inappropriate timing of CF introduction (<4 months of age or >6 months of age)—lower SES (OR = 1.25; 95% CI, 1.08–1.45).

> **Table 5.** Maternal characteristics associated with non-compliance to the recommendation for introduction of complementary foods (CF) at 4–6 months of age.


<sup>1</sup> Adjusted for age and BMI before pregnancy, country and SES. <sup>2</sup> Adjusted for age before pregnancy, smoking habits during pregnancy, country and SES. <sup>3</sup> Adjusted for age and BMI before pregnancy, smoking habits during pregnancy, country.

The logistic regression analysis showed that the odds of becoming overweight at preschool age among children who had early introduction of SF (1–3 months of age) compared to those with CF introduction at 4–6 months of age was 0.69 (OR = 0.69; 95% CI,

0.41–1.16; *p* = 0.16). The children introduced to SF before four months of age had a trend for a different overweight risk at preschool age according to their BF status which was not significant when adjusted for the mother's characteristics (SES, education, pre-pregnancy weight and smoking habits during pregnancy). For the children breastfed for ≥4 months, early introduction of CF was associated with a trend for higher later overweight (OR = 1.23; 95% CI, 0.29–5.14; *p* = 0.78), while among the exclusively formula-fed breastfed children, this risk tended to be lower (OR = 0.39; 95% CI, 0.052–3.12; *p* = 0.38). Late CF introduction at 7–12 months of age was not related to a difference in later overweight and obesity risk when adjusted for country, age and gender (OR = 0.98; 95% CI, 0.83–1.21; *p* = 0.99).

#### **4. Discussion**

Our study investigated the association between the timing of CF, breastfeeding status and overweight among European preschool children. Breastfed children (any type of breastfeeding) throughout the first 4–6 months of life and after the 12th month had a lower prevalence of overweight/obesity in childhood compared to formula-fed children. This finding is consistent with the European studies reporting similar results in previous cohorts [9,14,32,33]. The main findings point at a lower prevalence of overweight/obesity at six months of age (*p* < 0.001) in children with introduction of solid foods between 4–6 months of age compared to late introduction (7–12 months of age). However, there were no significant findings for the prevalence at one year and at preschool age (*p* > 0.05). A late SF introduction is related to a higher prevalence of overweight and obesity at six months of age, 12 months of age and at preschool age. Our results are consistent with the previously reported findings [15,18].

Pluymen et al. and Huh et al. reported that the duration of breastfeeding modifies the association between CF introduction and overweight: BF for less than four months and CF introduction before four months of age increased the risk for overweight by 37% compared to those with CF introduction ≥ 4 months of age [21,22]. We found a non-significant trend for an association of early introduction of SF and preschool overweight in breastfed children but not in formula-fed children.

Different previous studies aimed at identification of predictors of children's dietary intake such as SES and geographic region [34,35]. SES is one of the most commonly identified factors associated with childhood overweight and obesity and reflects a child's living conditions. However, there is uncertainty as to the mechanisms through which SES influences the child's weight. Breastfeeding practices and timing of CF introduction are related to SES and other maternal characteristics such as BMI, age at birth, tobacco use during pregnancy, gestational weight gain, depression and use of day care [36–39]. Results from the ToyBox-study show that mothers with low SES are more likely to have overweight/obese children compared to those with medium/high SES (OR = 1.41; 95% CI, 1.17–1.71 [31].

The current analysis supports the findings of other studies that significant risk factors associated with non-compliance to the recommendation for introduction of CF at 4–6 months of age are low SES and smoking throughout pregnancy (*p* < 0.05) [14,40–42]. Maternal educational level did not modify the association of CF < 4 months of age and overweight in the PIAMA cohort as well [19]. Rose et al., based on the data of the Infant Feeding Practices Study II and Year 6 Follow-Up Studies, suggested that the mother's decisions about milk-feeding and the types and quality of solid foods introduced in infancy can shape dietary patterns and obesity risk later in childhood. Infants who were offered foods high in energy density at nine months of age had a higher intake of these foods at six years of age and a higher prevalence of overweight compared to other classes of dietary patterns [43].

Our results, which hopefully will be useful for improving effectiveness of childhood obesity prevention programs in Europe, can also be utilized in low developed and developing countries. Although malnutrition is still a major challenge across the African continent,

the largest growth of obesity among 5- to 19-year-olds in the world between 1975 and 2016 was observed in southern Africa (about 400% per decade) [44].

A methodological limitation of the report is the cross-sectional study design of our study which does not enable identifying cause–effect associations. Another limitation is the parental self-reporting of weight, height, gestational weight gain, infant's birth weight, as well as BF practices and timing of CF introduction by mothers' some 3–4 years later. The use of the mean educational level as a single indicator for SES is another limitation of the study. Furthermore, the mother's alcohol consumption during pregnancy which was not investigated as a risk factor for child health may be added to the list of the study limitations. The strengths of our study are the large number of study participants, the inclusion of children from several European countries adding external validity and the standardization of measurement approaches [25,31].

#### **5. Conclusions**

We conclude that other variables have a greater impact on the risk for childhood obesity than the timing of CF introduction. Therefore, intervention programs for childhood obesity should be conducted, including educating mothers about healthy eating practices and other possible risk factors for overweight.

**Author Contributions:** Conceptualization, V.I. and N.U.; methodology, G.C., N.U. and V.I.; formal analysis, S.G., N.U.; investigation, S.G. and M.D.C.; writing—original draft preparation, N.U., V.I. and M.D.C.; writing—review and editing, B.V.K., M.D.C., O.A., V.I., G.C., A.K., S.G., Y.M., L.A.M. and P.S.; supervision, V.I., Y.M. and the Toy-Box-study group. All authors have read and agreed to the published version of the manuscript.

**Funding:** The ToyBox-study was funded by the Seventh Framework Program (The Community Research and Development Information Service (CORDIS; FP7)) of the European Commission under grant agreement No. 245200. B.V.K. is the Else Kröner Seniorprofessor of Paediatrics at LMU— University of Munich, financially supported by the Else Kröner-Fresenius-Foundation, the LMU Medical Faculty and the LMU University Hospitals. The content of this article reflects only the authors' views and the European Community is not liable for any use that may be made of the information contained therein.

**Institutional Review Board Statement:** The ToyBox-study (www.toybox-study.eu; accessed on 20 September 2020) adhered to the Declaration of Helsinki and the conventions of the Council of Europe on human rights and biomedicine. All the countries (Belgium, Bulgaria, Germany, Greece, Poland and Spain) obtained ethical clearance from the relevant ethics committees and local authorities: 1. Ethics Committee of Ghent University Hospital (Belgium)—EC/2010/037; 2. Committee for the Ethics of Scientific Studies (KENI) at the Medical University of Varna (Bulgaria)— 15/21.07.2011; 3. Ethics Committee of the Medical Faculty at LMU Munich (Ethikkommission der Ludwig-Maximilians-Universität München) (Germany)—400-11 (2012); 4. Bioethics Committee of the Harokopio University of Athens (Greece) (28/02-12-2010) and the Ministry of Education of Greece (approval code 29447/Γ7 (05-05-2011)); 5. Ethics Committee of the Children's Memorial Health Institute (Poland)— 1/KBE/2012; 6. CEICA (Comité Ético de Investigación Clínica de Aragón (Spain))—PI11/056 30.08.2011. The ToyBox-study is registered with the clinical trials registry: clinicaltrials.gov, ID: NCT02116296.

**Informed Consent Statement:** All parents/caregivers provided a signed consent form before being enrolled in the study.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. The data are not publicly available due to restrictions of informed consent and the requirement of IRB review and approval.

**Acknowledgments:** We gratefully acknowledge all the members of the ToyBox-study Group. Coordinator: Yannis Manios; project manager: Odysseas Androutsos; steering committee: Yannis Manios, Berthold Koletzko, Ilse de Bourdeaudhuij, Mai Chin A Paw, Luis Moreno, Carolyn Summerbell, Tim Lobstein, Lieven Annemans and Goof Buijs; external advisors: John Reilly, Boyd Swinburn and Dianne Ward; Harokopio University (Greece): Yannis Manios, Odysseas Androutsos, Eva Grammatikaki, Christina Katsarou, Eftychia Apostolidou, Anastasia Livaniou, Eirini Efstathopoulou, Paraskevi–Eirini Siatitsa, Angeliki Giannopoulou, Effie Argyri, Konstantina Maragkopoulou, Athanasios Douligeris and Roula Koutsi; Ludwig-Maximilians-Universität München (Germany): Berthold Koletzko, Kristin Duvinage, Sabine Ibrügger, Angelika Strauß, Birgit Herbert, Julia Birnbaum, Annette Payr and Christine Geyer; Ghent University (Belgium): Department of Movement and Sports Sciences: Ilse de Bourdeaudhuij, Greet Cardon, Marieke de Craemer and Ellen de Decker; Department of Public Health: Lieven Annemans, Stefaan de Henauw, Lea Maes, Carine Vereecken, Jo van Assche and Lore Pil; Vrije Universiteit Amsterdam (VU University) Medical Center, The Institute for Research in Extramural Medicine (EMGO) Institute for Health and Care Research (The Netherlands): Mai Chin A Paw and Saskia te Velde; University of Zaragoza (Spain): Luis Moreno, Theodora Mouratidou, Juan Fernandez, Maribel Mesana, Pilar de Miguel–Etayo, Esther M. González-Gil, Luis Gracia–Marco and Beatriz Oves; Oslo and Akershus University College of Applied Sciences (Norway): Agneta Yngve, Susanna Kugelberg, Christel Lynch, Annhild Mosdøl and Bente B. Nilsen; University of Durham (UK): Carolyn Summerbell, Helen Moore, Wayne Douthwaite and Catherine Nixon; State Institute of Early Childhood Research (Germany): Susanne Kreichauf and Andreas Wildgruber; Children's Memorial Health Institute (Poland): Piotr Socha, Zbigniew Kulaga, Kamila Zych, Magdalena Gózd ´ ´ z, Beata Gurzkowska and Katarzyna Szott; Medical University of Varna (Bulgaria): Violeta Iotova, Mina Lateva, Natalya Usheva, Sonya Galcheva, Vanya Marinova, Zhaneta Radkova and Nevyana Feschieva; International Association for the Study of Obesity (UK): Tim Lobstein and Andrea Aikenhead; CBO B.V. (the Netherlands): Goof Buijs, Annemiek Dorgelo, Aviva Nethe and Jan Jansen; AOK-Verlag (Germany): Otto Gmeiner, Jutta Retterath, Julia Wildeis and Axel Günthersberger; Roehampton University (UK): Leigh Gibson; University of Luxembourg (Luxembourg): Claus Voegele.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Review* **Management of Childhood Obesity—Time to Shift from Generalized to Personalized Intervention Strategies**

**Mohamad Motevalli 1,\*, Clemens Drenowatz 2, Derrick R. Tanous 1, Naim Akhtar Khan <sup>3</sup> and Katharina Wirnitzer 1,4,5,6**


**Abstract:** As a major public health concern, childhood obesity is a multifaceted and multilevel metabolic disorder influenced by genetic and behavioral aspects. While genetic risk factors contribute to and interact with the onset and development of excess body weight, available evidence indicates that several modifiable obesogenic behaviors play a crucial role in the etiology of childhood obesity. Although a variety of systematic reviews and meta-analyses have reported the effectiveness of several interventions in community-based, school-based, and home-based programs regarding childhood obesity, the prevalence of children with excess body weight remains high. Additionally, researchers and pediatric clinicians are often encountering several challenges and the characteristics of an optimal weight management strategy remain controversial. Strategies involving a combination of physical activity, nutritional, and educational interventions are likely to yield better outcomes compared to single-component strategies but various prohibitory limitations have been reported in practice. This review seeks to (i) provide a brief overview of the current preventative and therapeutic approaches towards childhood obesity, (ii) discuss the complexity and limitations of research in the childhood obesity area, and (iii) suggest an Etiology-Based Personalized Intervention Strategy Targeting Childhood Obesity (EPISTCO). This purposeful approach includes prioritized nutritional, educational, behavioral, and physical activity intervention strategies directly based on the etiology of obesity and interpretation of individual characteristics.

**Keywords:** children; adolescents; overweight; obesity; weight management; lifestyle; body composition

#### **1. Introduction**

The examination of the etiology of childhood obesity is a growing area of research aiming to yield important insights for public health [1,2]. During the last three decades, the annual growth rate of publications on childhood obesity (average of 11.6% per year) has been generally higher than other sub-areas in the pediatric field and biomedical research [3]. Given the rising prevalence of childhood obesity in most developed and developing countries, it is now considered a global pandemic [4]. Worldwide, an estimated 170 million children are considered overweight or obese currently [5], and approximately more than half of them are predicted to become obese adults [6].

These trends in excess body weight may also contribute to an increase in chronic cardiometabolic disorders, typically observed only in adults (e.g., hypertension, hyperglycemia,

**Citation:** Motevalli, M.; Drenowatz, C.; Tanous, D.R.; Khan, N.A.; Wirnitzer, K. Management of Childhood Obesity—Time to Shift from Generalized to Personalized Intervention Strategies. *Nutrients* **2021**, *13*, 1200. https://doi.org/ 10.3390/nu13041200

Academic Editors: Odysseas Androutsos and Evangelia Charmandari

Received: 30 January 2021 Accepted: 2 April 2021 Published: 6 April 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

and dyslipidemia), but are becoming increasingly common in children and adolescents with obesity [7]. Additionally, pediatric populations with obesity are known to have several psychosocial problems including discrimination, social isolation, and low self-esteem, which affect their health, education, and quality of life [6,8]. Furthermore, the crosstalk between obesity and many viral pandemics, such as the 2009 swine flu [9] or the current COVID-19 pandemic [10,11], has provided new insights into mortal characteristics of this chronic syndrome.

The etiology of obesity is complicated and multifactorial, which indicates that excess body weight results from a complex interaction of a broad range of factors [12]. In addition to genetic vulnerability as a well-recognized internal factor contributing to excess body weight, a wide range of physiological disorders, as well as modifiable environmental factors and obesogenic behaviors, play key roles in the development of obesity [11,13,14]. Amongst children, the most common obesogenic behavior includes high consumption of unhealthy foods, low levels of physical activity (PA), high levels of mental stress, high levels of screen time, and poor sleep patterns [1,2,15]. These behaviors are influenced by several factors and interactions involving genetics, interpersonal relationships, and the environment [16,17]. Additionally, some evidence indicates that obesity-related behaviors are highly context-dependent and are influenced by several biopsychosocial factors [18]. When discussing the gene–environment interplay in the etiology of obesity, it is believed that the contribution to an obese phenotype is not "nature or nurture", but rather "nature via nurture" [12]. Recently, Jackson et al. emphasized in their review that biology plays a fundamental role in determining the amount of body fat in addition to environmental factors [12]. Moreover, genetically predetermined obesogenic behavior seems to have a significant relationship with environmental influence on body weight [12], which could add to the complexity of obesity. This complex interaction is exemplified further by the role of energy flux (the rate of energy expenditure and energy intake) in the regulation of energy balance [19]. The complicated nature of childhood obesity which leads to a wide range of inter-individual differences highlights the importance of child-centered specific approaches, particularly personalized interventions, for managing childhood obesity.

Current scientific insights are limited in successfully decelerating the rise of the childhood obesity pandemic; the present review's primary goal is to establish a novel translational link between the literature and practice by introducing applied strategies targeting childhood obesity. Therefore, the purpose of this narrative review is to provide a brief overview of the current preventative and therapeutic approaches towards the management of childhood obesity (focusing on their limitations and complexity), and accordingly, to suggest an Etiology-Based Personalized Intervention Strategy Targeting Childhood Obesity (EPISTCO) as a purposeful approach to prioritize and implement nutritional, PA, and lifestyle intervention strategies based on the etiology of obesity and interpretation of individual characteristics. These objectives are based on the limited success of previous efforts targeting childhood obesity.

#### **2. Weight-Related Behaviors in Children**

Comprehensive clinical guidelines and recommendations to diagnose, prevent, and treat childhood obesity have been well documented for pediatric specialists to implement at different stages in obesity prevention and treatment programs [20,21].

Nutrition and dietary pattern not only affect anthropometry and body composition in children and adolescents, but they also influence neurocognitive and psychomotor development [22,23]. Evidence indicates that when compared to adults, children and adolescents are at a higher risk of insufficient intake of certain food groups (e.g., whole grains and/or unprocessed foods) [7,24], which may contribute to the increased risk of obesity [7]. Calorically restricted diets are widely used to target childhood obesity [22], but potentially contribute to nutrient deficiencies, which may impair growth and development [22,25]. The importance of adequate nutrient consumption for growth and development also emphasizes the need for targeting energy expenditure by ensuring sufficient PA. Even

though detrimental effects of insufficient PA on various health and weight outcomes are well-documented, results from a comprehensive analysis including 1.6 million adolescents from a pooled number of 298 school-based studies from 146 countries show that 80% of adolescents do not meet PA recommendations of at least 60 min of PA with moderate to vigorous intensity over 5 d/w, which puts their current and future health at risk [26].

In addition to poor nutrition and low PA, results from different studies show that most children and adolescents do not meet obesity-associated lifestyle guidelines, such as recommendations for sleep and screen time [27,28]. According to the Centers for Disease Control and Prevention (CDC), 60% to 70% of the American pediatric population does not meet the American Academy of Pediatrics (AAP) sleep recommendations [1,27] of 10–13 h for 3- to 5-year-olds, 9–12 h for 6- to 12-year-olds, and 8–10 h for 13- to 18-yearolds [29]. Short duration and late sleep timing can contribute to the onset and development of childhood obesity [30], particularly by altering appetite-regulating hormones and consequential eating disorders [31]. Parents can play an important role in fostering healthy sleep patterns by arranging sleep time, providing a calm atmosphere, and keeping screens away before bedtime [32]. Higher amounts of daily screen time contribute to obesity due to their association with a reduced feeling of satiety, increased consumption of unhealthy and energy-dense snacks [33], and poor sleep patterns [34]. The AAP recommends that daily nonacademic screen time (TV, video games, and mobile phone) should not exceed one hour for 2- to 5-year-olds, and two hours for ≥6-year-old children, and there should be parental supervision of content watched [1]. Available data, however, indicates that the majority of children has an extraordinarily high daily screen time [28,35], up to an average of 6 h/day among 13- to 18-year-olds [36] and 7 h/day among 8- to 18-year-olds [37] when TV, computer, mobile devices, and web-based sources are combined. Sedentary behaviors, screen time, and sleep abnormalities can be even higher during annual vacation due to the absence of a regular schedule [38].

#### **3. Strategies for the Prevention and Treatment of Childhood Obesity**

To date, the safety and efficacy of various approaches on the management of childhood obesity have been reported by numerous experimental and cross-sectional studies as well as reviews and meta-analyses [2,6,15,39,40]. Nevertheless, pediatric clinicians and researchers are often encountering several challenges when applying preventative and therapeutic programs and the characteristics of an optimal and comprehensive weight management strategy remain controversial [1,41]. Evidence suggests that the management of childhood obesity requires consideration of genetic, biological, behavioral, psychological, interpersonal, and environmental factors to induce sustainable lifestyle changes along with an in-depth understanding of these interactions to identify opportunities for intervention strategies [39].

#### *3.1. Intervention Components*

The most common preventative and therapeutic interventions applied and suggested by investigations are nutritional, PA, lifestyle, and educational methods. In adolescents with a high degree of obesity or advanced metabolic disease, clinical treatments including pharmacological and surgical strategies have also been suggested [20]. Due to the potential side effects of medical interventions, a careful evaluation and comparison of risks and benefits are necessary before implementing such interventions for pediatric patients with obesity [42]. It has been emphasized that pharmacotherapy and bariatric surgery should never be implemented in adolescents with obesity (and those with other vital untreated disorders) who have not engaged in healthy dietary and PA practices [20].

#### 3.1.1. Diet

It should be considered that nutritional approaches targeting childhood obesity are not limited to restricted energy intake but rather, the most appropriate nutrition strategy for long-term weight reduction and the promotion of metabolic and mental health is

shifting to healthy food choices [43] that include predominantly whole food plant-based sources [44,45]. This dietary pattern restricts added sugars, refined grains, sweetened beverages, fast foods, calorie-dense snacks, and high-fat processed foods and includes fruits, vegetables, nuts, and whole-grains along with well-structured meal frequencies [6,44,46]. The weight-related benefits rising from plant-based diets are attributable to a reduced caloric intake and an increase in postprandial energy expenditure by a higher thermogenic response [47,48]. Additionally, whole food plant-based diets lead to favorable changes in cardio-metabolic and digestive health, which are both associated with further weightrelated advantages as well [49,50]. Increasing nutritional literacy of children and their parents (regarding agriculture, food industries, food safety, cooking, and theoretical knowledge of energy balance, nutrition, and diets) could further promote sustainable changes that contribute to healthier dietary patterns [51]. While these are general principles and recommendations, no single diet should be prescribed or recommended as the best for all children with obesity [43]. Researchers believe that the optimal macronutrient composition depends on factors such as appetite, thermogenesis, energy homeostasis, and gut microbiota [52]. Furthermore, the ideal diet for treating overweight and obesity should be safe, efficacious, nutritionally adequate, culturally acceptable, and economically affordable [43]. To date, however, the majority of nutritional strategies targeting childhood obesity are still based on a "one-size-fits-all" model, which does not take into account the inter-individual variability [53], which often results in a reduced adherence rate to dietary changes [54].

#### 3.1.2. Physical Activity

On the other side of the energy balance equation, PA has been emphasized as a critical component for healthy body weight. Promoting PA is considered an effective intervention strategy in pediatric weight management [55,56], which attributes to the concept of energy flux [19]. Energy flux represents the rate of energy expenditure and energy intake [19,57], and a higher energy flux (obtained by increased PA) results in better regulation of energy balance during weight loss [57] and/or the prevention of weight gain [58]. PA could also result in favorable improvements in mental and physiological health, and both are indirectly associated with further weight-related advantages [59]. In children and adolescents with obesity, the most common barriers to engage in regular PA programs are lack of selfdiscipline, lack of someone to engage in PA with, self-consciousness about appearance [60], and decreased level of motivation due to the limited motor abilities and/or being out of shape [61]. Daily physical activities of children are not limited to regular physical education classes or sport/exercise engagements. Active travel to school, unstructured active play during school recess, and activities at home or playgrounds can be additional viable PA sources [32]. Currently, however, due to the COVID-19 pandemic and social lockdowns, the movement opportunities have been significantly diminished [62], and home exercises have been highly recommended [63].

#### 3.1.3. Lifestyle and Education

In addition to diet and PA, other lifestyle parameters (e.g., psychological behaviors, modifying sleep patterns) are considered effective interventions in weight management programs. Independently or along with educational interventions, additional lifestyle behaviors could further increase the efficiency of PA and dietary interventions [1]. A controlled experimental study showed that a two-year, multi-component obesity intervention focusing on a lifestyle educational curriculum resulted in beneficial changes in Body Mass Index (BMI) percentiles in the intervention groups [64]. There is also evidence suggesting that mindfulness interventions [65,66] and forethoughtfulness (defined as being oriented more towards the future than the present) [67] might be advantageous for improving obesity-related eating and behavioral patterns. Mindfulness, defined as the awareness that arises from purposefully paying attention in the present moment with non-judgment [68], is suggested as an effective intervention strategy targeting childhood obesity [69]. Evidence also indicates that sleep is an important modifiable risk factor for managing childhood

obesity, as eating and PA behaviors can be affected by the quality and duration of sleep [70]. Modifying sleep patterns in school-age children resulted in healthy dietary patterns via decreased food consumption, in particular, thus promoting favorable weight outcomes [71].

In general, it has been well-established that multi-component interventions including PA, nutritional, lifestyle, and educational strategies have been shown to yield better outcomes than single-component strategies [1,41,72]. In a systematic review of the effectiveness of lifestyle interventions targeting children's weight and cardio-metabolic health, beneficial outcomes were observed only following the multi-component interventions [73]. However, due to the complex interaction in these approaches, identifying the degree of effectiveness for each component remains controversial.

#### *3.2. Intervention Settings*

Almost all studies in the childhood obesity sector have critically investigated or discussed the environment where interventions are applied and the people who support and/or supervise weight management programs. It has been well-established that home, school, and community can all play important roles independently in shaping and stabilizing children's lifelong health- and weight-related behaviors [74,75].

#### 3.2.1. Home

Parental beliefs, attitudes, behaviors, and social support are vital for a child's health and body composition [76]. Available evidence indicates that parents are involved in about half of the interventions targeting childhood obesity, and successful improvements on children's BMI are in 75% of studies with parental involvement [74]. Results show, from a meta-analysis of 22 randomized control trials examining home-based interventions to control childhood obesity, that parents can play a crucial role in managing children's weight by facilitating, motivating, and coaching the healthy behaviors of their children [77]. Due to the close familiarity of parents and their child, parents may better understand and consider the lifestyle parameters contributing to the development of obesity in their child [78]. Although parent-only interventions may be more cost-effective compared to school- and community-based programs [79], evidence suggests to combine home-based programs with other settings to deliver more favorable effects on anthropometry and BMI in children [80]. The effectiveness of grandparental supervision, on the other hand, has been reported to be close to zero with no association between children's BMI z-scores and grandparental child care (whether as the primary caregiver or co-residence) [81].

#### 3.2.2. School

School is an important setting for improving child health behaviors, as children spend a significant part of their daily life in schools [82,83]. Moreover, the following conditions of the school environment also benefit the setting for implementing overweight/obesityrelated interventions: schools offer a structured environment for applying interventions with ease; schools may provide one or two meals per day and, therefore, potentially dictate healthy food choices in their cafeteria; schools usually provide opportunities for PA and active games during recess and daytime; schools produce and expand extracurricular educational resources and wellness policies for both children and parents; schools could run an indirect competitive and encouraging atmosphere to promote children's motivation and adherence towards interventions; schools could introduce their physical education instructors and/or athletic trainers as role-models; schools benefit from the contribution of staff and teachers to facilitate, deliver, and supervise the interventions [74,84,85]. Interestingly, successful school-based interventions are also highly effective in improving children's anti-obesogenic behaviors at home [83]. However, because of time and budget constraints, many schools are not able to implement health and weight management programs [72].

#### 3.2.3. Community and Clinics

In addition to home- and school-based strategies, the community environment and clinical settings are also common areas for managing childhood obesity and prevention [20,75]. Community interventions targeting obesity incorporate policies and strategies and aim to reduce the population risk of obesity [75]. These interventions involve but are not limited to the availability and use of health and fitness facilities, media-based activities, and health-oriented businesses by local and central administrations [1,86]. The EPODE program (Ensemble Prévenons l'Obésité Des Enfants: together, let us prevent childhood obesity) could be considered a successful example of a community-based intervention, which emphasizes a multifactorial approach targeting childhood obesity at different community levels [87]. Interventions using a community-based approach could achieve the long-term goals of reducing the prevalence of childhood obesity [88], especially for children who live in low-income societies [75]. Clinical or primary-care interventions, on the other hand, include any medical or non-medical strategies implemented by healthcare and pediatric specialists [89]. Clinical and community-level interventions can significantly improve lifestyle patterns when applied simultaneously [90]. Significant improvements in body weight have been achieved in pre-school children aged 2–5 years following multi-component clinical interventions (e.g., PA, nutrition, education) with parental involvement [91]. However, reports from different meta-analytic studies indicate poor effectiveness of primary-care programs on childhood obesity [89,92,93], which might be attributed to a dose-response relationship, where the frequency and duration of treatment contact highly affect the outcomes [1]. Additionally, the "sustainability" of intervention effects could be another limitation in clinical approaches, as the time of engagement is limited compared to other settings [94,95].

In general, it seems that due to the multi-factorial nature of childhood obesity, a maximally efficient strategy to manage childhood obesity requires integrating multiple settings for delivering multi-component interventions. Evidence shows that school-based interventions with family inclusion have the largest effect on weight outcomes when multicomponent programs are implemented, including PA and diet [84,96]. Results from a study comparing the effectiveness of home versus school settings on nutritional habits, PA behaviors, and BMI changes showed that the home environment had a stronger association with health in general compared to the school setting [97]. It should be mentioned, however, that parental involvement in many preventive studies can be more effective in pre-school and early-school children, whereas school- and community-based intervention strategies lead to more favorable outcomes in older children and adolescents [1], particularly for those who are above twelve [98]. Nevertheless, to enhance the effectiveness of strategies, it appears important that parents permanently engage in supporting and reinforcing their children's health behaviors [99].

Some limitations can affect the progress of weight management programs, similar to any preventative and therapeutic strategy. Time and financial resources are major limitations for the implementation of multi-component weight management strategies [100]. In addition, poor awareness and lack of self-discipline have been reported as personal barriers when adhering to a healthy lifestyle [60]. Furthermore, difficult-to-reach goals set by parents and clinicians are considered a vital but hidden limitation, as it is well-established that strict targets may often lead to failure in weight control programs in children [101]. Age also appears to be an important moderator for weight control outcomes as older children displayed larger and more beneficial effects than younger children following weight-management interventions [92,93]. Beyond these limitations, the obesity prevention strategies seem to follow a dose-effective manner, as more intensive and longer-lasting interventions are associated with better outcomes in children [93]. Further, it appears that purposeless and/or unsupervised strategies not based on the individual needs and personal characteristics of the targeted child could minimize intervention effects [54].

#### **4. Personalized Strategies**

Despite a wealth of scientific information on a wide range of interventions and strategies targeting childhood obesity, the translation and transfer of this knowledge into a practical approach seem highly challenging. According to a comprehensive study by the Institute of Medicine (IOM), which analyzed more than 800 scientific reports, the progress of obesity prevention was not favorable in the national trend data, and data was not translatable into clearly scalable strategies [102,103]. It has been reported that metaanalytic approaches for identifying solutions to obesity, which is a complex health problem, could not deliver favorable practical information [102]. As a result of obesity's complexity, the condition seems not only limited to its etiology, but also to intervention strategies targeting childhood obesity. Given the interaction between various components of preventative/treatment approaches, the management of childhood obesity remains highly complex. Figure 1 shows a conceptual model that describes the complexity of interactions between key aspects of four research-derived categories ("What", "When", "Who", and "Where"), which are critical in the management of childhood obesity. "What" refers to the components including diet, PA, other lifestyle interventions, education, medication, and surgery. "When" stands for different age groups that are targeted (including pre-school, school-age, and puberty). "Who" represents the involved population such as the child, parents, teachers, and specialists. Finally, "Where" appoints different settings including home, school, community, and clinic (Figure 1).

**Figure 1.** Conceptual 4W model describing the complexity of interactions between research-based modules contributing to the management of childhood obesity.

Recently, the implementation of personalized dietary approaches to managing complicated health problems (e.g., cardiovascular and metabolic disorders) has been increasing [104,105]. Personalized interventions could be defined as advanced and detailed models of clinical/primary care interventions. Given the available data, current clinical interventions seem to have some limitations. Clinical strategies focus primarily on treatment rather than prevention and thus are often conducted in close coordination with the primary healthcare system with high accessibility and frequency of visits mostly in the clinical setting [106]. Clinical approaches mainly focus on nutritional and medical interventions with lower attention on lifestyle, educational, and movement strategies [20]. Lifestyle

counseling, including suggestions for PA by health specialists, remains below an acceptable level [107–109] even though the importance of lifestyle interventions by physicians and/or health care providers has been well-documented in patients' health- and weight-related behaviors [107,108]. This may be attributed to inadequate knowledge and training, office time constraints, and poor personal habits of specialists/physicians [107,109]. Rather than providing general information, a personalized strategy uses a broad range of info on individual characteristics to develop targeted nutritional and non-nutritional advice, products, or services assisting people to reach their goals via a purposeful approach based on their current behaviors, preferences, barriers, and objectives.

Personalized dietary approaches have been reported previously as a promising topic of research in the treatment of obesity [110]. To date, evidence supporting personalized strategies to manage obesity has come from clinical and observational studies mostly in the area of nutrition. The Academy of Nutrition and Dietetics developed the personalized nutritional approach NCP (Nutrition Care Process) based on nutritional assessment, diagnosis, planning, and monitoring. This multi-step model was designed to structure individualized nutritional care targeting childhood obesity and was effective in different investigations [22,111,112]. In a review study assessing the effects of NCP-based educational programs (including education on meal planning, portion control, healthy snack selection, and cooking with plant-based sources), favorable outcomes were reported regarding childhood obesity [113].

In addition to nutritional interventions, the limited available data supports the effectiveness of other personalized interventions on weight and/or health outcomes in children and adolescents. A 3-month personalized PA intervention using an internet-based program showed significant effects on psychosocial health and PA level in adolescents [114]. Evidence consistently indicates that when compared to generalized programs, personalized technology-based PA interventions are more effective at modifying health behaviors [115]. Results from a study on adolescents with obesity or diabetes show that 16 weeks of personalized exercise (based upon baseline fitness level of participants)—with parental support and ongoing motivation—can improve PA level and result in a sense of personal health [116]. Additionally, a controlled experimental study showed beneficial effects of personalized lifestyle coaching on childhood obesity [117], in which a health coach called child-parent pairs separately by telephone for a total of 21 sessions. There is also evidence indicating children's health behaviors, particularly sleep patterns, could be improved following personalized educational interventions for mothers with 3- to 5-year-old children [118].

In general, personalized recommendations on the personal needs of a child and his/her family could be a promising approach for the prevention and treatment of obesity. A comprehensive personalized approach targeting childhood obesity may include, among others, nutritional, educational, and PA-based intervention strategies at various settings to alter lifestyle patterns and attitudes. The overall consensus is that implementing a well-proposed personalized program not only maximizes desirable outcomes but also contributes to the sustainable adherence of a healthy lifestyle pattern [54]. In addition, due to the purposeful nature of personalized interventions, time and budget could be partially saved following this approach. Similar to other successful programs, a personalized program should further combine education and motivation to obtain slow but sustainable weight and health benefits [43].

#### **5. EPISTCO Model**

Considering the complicated facts in the etiology and management of childhood obesity and in order to establish a novel translational link between the literature and practice, this narrative review presents a basis for an Etiology-Based Personalized Intervention Strategy Targeting Childhood Obesity (EPISTCO) (Figure 2), to provide a framework for the purposeful prevention and treatment of childhood obesity.

**Figure 2.** Schematic design of EPISTCO (Etiology-Based Personalized Intervention Strategy Targeting Childhood Obesity) model, which is based on four multi-stage steps.

The EPISTCO model highlights that the design of a personalized program targeting childhood obesity requires an understanding of the complex etiology of excess weight gain by assessing a series of biological, nutritional, behavioral, and environmental factors. Unlike previously described models, the EPISTCO model implements a multi-component intervention program within multiple settings and considers personalized priorities for the components and settings. In this structured multi-disciplinary and etiology-based approach, the programs are highly adaptable based on individual and environmental barriers and potentials.

The EPISTCO model includes four multi-stage steps (Figure 2). The first and probably the most important step is "discovering the etiology of obesity", which most likely requires a clinical setting. This step consists of four stages including (a) assessments, (b) interpretation of data, (c) diagnosis of the relevant causes, and (d) classification of the causes. Assessments (e.g., physical characteristics, eating habits and disorders, sleep patterns, etc.) can be made via questionnaires, field tests, and laboratory measurements and are depending on the availability of time, equipment, and specialists. Table 1 represents the most important assessment items summarized in nine general categories that provide viable information to design an EPISTCO.



Step 2 is "setting the target", in which a multi-phase goal is defined according to the information obtained from Step 1. The emphasis is to set a realistic target, as difficult-toreach targets often lead to failure. Step 3, "designing the strategy", consists of the following stages: (a) selecting the most appropriate interventions, (b) prioritizing interventions according to stage "d" from Step 1, (c) designing program schedules along with extra general recommendations, and (d) educating both the child and parents about the next step in conjugation with motivational incentives. At the end of Step 3, the next visit(s) must be set according to a time-based, target-based, or problem-based style. Accordingly, this step will also determine the role of different settings including home, schools and communities. Finally, Step 4, "supervising and supporting", consists of three stages: (a) individualized

direct or indirect coaching and psychological supporting, (b) monitoring, (c) reassessment, analysis of the progress as well as evaluating problems, and (d) revising and updating the program. This step returns most likely to the clinical setting to connect Step 4 and Step 1. Rather than circling back, the intention is to continue the procedure to stabilize health behaviors.

To further enhance the understanding of the characteristics of this personalized approach, the following examples can be considered. For a child with obesity who has a proper quantity and quality of nutrition, interventions should prioritize PA and other lifestyle patterns. A more active child with excess body weight, on the other hand, with other causes (e.g., unhealthy food choices, sleep patterns, lifestyle, biochemistry) may require a different approach that should be scrutinized during initial assessments in order to design and suggest a purposeful etiology-based program. To reach favorable results, it is, nevertheless, highly recommended that all steps and stages of the EPISTCO approach are conducted and supervised by well-experienced pediatric specialists in different sub-disciplines of health. Moreover, every stage throughout the process should be well documented. The gathered data will also provide viable information that enhances the understanding of the etiology of obesity, which is critical for the improvement of the effectiveness of such personalized approaches.

#### **6. Conclusions**

The high prevalence of childhood obesity is a major threat to future public health and available literature indicates that weight-related nutritional, PA, and lifestyle recommendations are not met by the majority of children. Given the complex and multi-factorial nature of obesity in both etiology and management, it appears that there is a fundamental need to develop and apply personalized approaches to prevent and treat childhood obesity. As a practical, purposeful, and promising suggestion, the EPISTCO model emphasizes incorporating various approaches, including nutritional, lifestyle, and PA, that are prioritized, prescribed, and supervised based on the individual needs and personal characteristics within multiple settings.

In general, the EPISTCO model offers a purposeful framework for pediatric researchers and specialists that contributes to a better understanding of the interplay between various factors associated with childhood obesity, which can increase the efficacy of interventions. While it includes a comprehensive approach towards minimizing childhood obesity, not all aspects need to be implemented in every situation.

**Author Contributions:** Conceptualization, M.M., C.D. and K.W.; writing original draft and preparation, M.M.; reviewing and editing, C.D., K.W., D.R.T., N.A.K. and M.M.; supervision and general support, K.W. and N.A.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** No new data were created or analyzed in this study. Data sharing is not applicable to this article.

**Acknowledgments:** There are no professional relationships with companies or manufacturers who will benefit from the results of the present study.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

