Introduction

Childhood obesity remains a pressing public health issue on a global scale, with its prevalence steadily increasing across various regions of the world [1,2]. Available data suggest that the Eastern Mediterranean Region (EMR) harbors one of the highest rates of childhood overweight and obesity [3], particularly in countries of the Gulf Cooperation Council [3]. The early school age has been reported as a key stage for the development of obesity [4], as it is often associated with significant changes in the child’s lifestyle. For primary school-children, the family’s daily routine undergoes fundamental restructuring and the demands on children typically increase, such as the requirement to be punctual and to be sitting most of the day [5]. The dietary habits of children also change in this life stage with a higher reliance on snacks and foods available within the school setting [4].

Pediatric obesity may lead to adverse physical and psychological effects that appear in childhood and tend to track into adulthood [6]. Cardiometabolic complications such as impaired glucose regulation, dyslipidemia, high blood pressure and metabolic syndrome are amongst the well-described short-term health consequences of childhood obesity [7]. Pediatric obesity is also likely to persist into the adult life stage, increasing the risk for chronic diseases such as type 2 diabetes, cardiovascular diseases, and specific cancers in the long-term [8].

The etiology of childhood obesity is multifactorial and often involves a complex interplay between genetic, environmental and behavioral factors [9]. A positive energy balance, which tends to result from the regular consumption of energy-dense, high fat and/or high sugar foods, is considered amongst the main causes of obesity [10,11]. In this respect, nutrition research has initially focused on examining individual nutrients or foods in relation to obesity, but more recently, attention has shifted towards identifying and analyzing holistic dietary patterns [10,11]. This approach offers a comprehensive understanding of the overall diet, encompassing the combination of various food groups and taking into consideration the complex interactions that may exist between the different foods and nutrients [8,12]. Studies conducted amongst school-aged children from various parts of the world have suggested that certain dietary patterns characterized by frequent consumption of refined cereals, energy dense, micronutrient poor foods such as fast foods and sweets correlate with higher body mass index (BMI) and adiposity [13,14]. For instance, a study conducted amongst 6–16 year old children from various European countries showed that the majority of children who were overweight and obese (39% and 56%, respectively) were allocated to the “Refined Cereals” cluster [15]. A study conducted in China amongst 6–17 year old children [16] showed that the Western dietary pattern was significantly associated with an increased risk of obesity (OR: 1.49; 95% 1.21, 1.84) after adjusting for potential confounders, whereas the traditional Chinese pattern was not associated with obesity in this age group. Studies conducted in the Eastern Mediterranean Region are scarce. In Iran, Bahreynian et al (2013) reported that lower adherence to the “sweet and dairy” pattern was associated with lower BMI among girls aged 7–11 years (OR = 0.42, CI = 0.21, 0.85), whereas no significant associations were found for boys [17].

In the Gulf Council Coperation countries, studies investigating dietary patterns as a determinant of childhood obesity are lacking, despite the high burden of obesity in the region. Particularly, data stemming from a study conducted in 3 Emirates of the UAE showed a high prevalence of child obesity in school-aged children (14.3% in 4–13 year olds) [18]. It is in this context that this study was conducted, with the aim of 1) deriving and characterizing dietary patterns amongst young school-aged children aged 4–9 years in the UAE; 2) investigating the association of the identified dietary patterns with obesity in this age group and 3) evaluating the association of the derived patterns with energy, macro- and selected micronutrients intakes. The results of this study can inform tailored interventions aimed at combating childhood obesity and fostering a culture of healthy eating practices among children.

Materials and methods

Results

The baseline characteristics of the study population, categorized by BAZ, are summarized in Table 1. The study included 426 children (wasted: 10.3%, normal weight: 65.2%, overweight (including at risk of overweight): 19.5%, obese: 5%). The mean age of the children was 5.72 years (±1.37), almost evenly distributed between males (47.9%) and females (52.1%). Emirati nationality constituted the majority (64.3%) of the study population. Maternal age (mean 35.82 ± 6.09) and BMI (mean 28.12 ± 5.08) were significantly associated with children’s BAZ categories (p-values 0.027 and 0.038, respectively). Notably, mothers of children in the “obese” BAZ category were older and had a higher BMI compared to other categories. About half of the parents had a university education or higher (48.6% for mothers and 51.4% for fathers). Employment rates were 90% for fathers and 46.2% for mothers. Concerning living conditions, 64.9% of children were born in families with a crowding index of <2 persons per room. Supplement use was reported by 23.9%. Adherence to physical activity and sleep recommendations was observed among 40.8% and 67.6%, respectively. Picky eating behavior was prevalent (52.2%), and significantly associated with BAZ (p-value <0.001), with a higher proportion of picky eaters in the “obese” BAZ category. The presence of a live-in helper was reported by 60.6% of participants.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Baseline characteristics of the study population, by BAZ categories (n = 426).

https://doi.org/10.1371/journal.pone.0352032.t001

Table 2 provides a description of the two identified dietary patterns and their corresponding food groups, as derived from FA: “Western Dietary Pattern” and the “Healthy Dietary Pattern.” The suitability of the data for factor analysis was supported by the KMO test value of 0.536 and Bartlett’s test of sphericity, which was statistically significant (χ², p < 0.001). The number of factors retained was determined based on Kaiser’s criterion, using eigenvalues greater than 1, as well as the interpretability of the resulting factors. The two retained dietary patterns explained a total of 24.76% of the variance in the data, with the Western Dietary Pattern explaining 12.6% and the Healthy Dietary Pattern explaining 12.2%. The Western Dietary Pattern was characterized by higher intakes of animal proteins, refined grains, and starchy vegetables, as indicated by positive factor loadings of these food groups. Additionally, this pattern was negatively associated with mixed traditional dishes, legumes, and nuts. In contrast, the Healthy Dietary Pattern was marked by higher consumption of fruits and vegetables, olive and vegetable oils, whole grains, and milk and dairy products. Fast food, salty snacks, sweets and sugar-sweetened beverages had negative loadings on this dietary pattern.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Food groups intakes and their corresponding factor loading according to the two derived dietary patterns.

https://doi.org/10.1371/journal.pone.0352032.t002

Table 3 shows the association of the derived dietary pattern with BAZ. While no significant association was found between the Western Dietary Pattern and BAZ, higher scores of the Healthy Dietary Pattern were associated with lower BAZ, in both the crude and adjusted models. The adjusted model included variables associated with BAZ at p < 0.20 in Table 2, namely child’s sex, mother’s age, mother’s BMI, father’s education, supplementation, meeting daily sleep recommendations, and picky eating. The adjusted beta coefficient in the multiple linear regression analyses was −0.07; 95% CI: −0.15 to −0.01; p-value = 0.043 (Table 3). Fig 1 presents the mean scores of both dietary patterns by weight categories. Children in the normal weight category exhibited higher mean scores of Healthy Dietary Pattern compared to those in the overweight or obese categories. In turn, those in the overweight and obese categories had higher scores of the Western Dietary Pattern than children in the normal BAZ category (Fig 1).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Linear regression analyses for the associations of the derived dietary patterns with BAZ in the study population.

https://doi.org/10.1371/journal.pone.0352032.t003

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Mean scores of the Western and Healthy dietary patterns based on BAZ among the study population.

Legend: Abbreviations: BAZ: BMI for Age Z-Score.

https://doi.org/10.1371/journal.pone.0352032.g001

The mean intakes of the various food groups stratified by BAZ status are presented in Fig 2. As depicted in this figure, higher intakes of SSB, refined grains, animal proteins, fast food, legume and nuts, salty snacks were reported by children with overweight and obesity. On the other hand, these children were reported to consume less milk and dairy products, mixed dishes, fruits and vegetables and whole grains.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Mean intake of food groups in grams stratified by BAZ status.

Legend: Abbreviations: BAZ: BMI for Age Z-Score.

https://doi.org/10.1371/journal.pone.0352032.g002

Table 4 presents the relationships between the factor scores of the derived dietary patterns and the energy, macro- and selected micronutrient intake among the study participants. Significant positive correlations were observed between the scores of the Western Dietary Pattern and carbohydrates (r = 0.12), proteins (r = 0.32), cholesterol (r = 0.25), saturated and trans-fat ((r = 0.26), and iron (r = 0.23) intakes. On the other hand, the scores of this pattern were negatively associated with fat (r = −1.99), monounsaturated fat (r = −0.2), polyunsaturated fat (r = −0.24), dietary fiber (r = −0.19), and calcium (r = −0.11). In contrast, the Healthy Dietary Pattern demonstrated negative correlations with total energy intake (r = 0.2), carbohydrates (r = −0.17), fat (r = −0.18), saturated and trans-fat (r = −0.33), monounsaturated fat (r = −0.13), polyunsaturated fat (r = −0.2), and iron (r = −0.1). A positive correlation was observed between the Healthy Dietary Pattern and dietary fiber (r = 0.19).

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Pearson correation coefficients for the association of the scores of the derived dietary patterns with energy, macro- and selected micronutrients intake, among the study participants^*†.

https://doi.org/10.1371/journal.pone.0352032.t004

Discussion

This study is the first to characterize dietary patterns amongst young school-aged children in the Gulf Council Cooperation and their association with overweight and obesity in this age group. Two main dietary patterns were identified: the “Western Pattern” and the “Healthy Pattern.” Higher scores of the “Healthy” pattern were associated with lower BAZ while no significant association was observed for the Western pattern. Together, the two patterns identified in this study explained 24.76% of the total variance in dietary intakes amongst 4–9-year-old children, which is within the range of variance reported in the literature [33,34]. It remains important to note that there remains a substantial 75% of dietary intake variance unexplained, highlighting the complexity and multifaceted nature of dietary habits in this young age group.

Several studies conducted on school-aged children have also identified two major dietary patterns in this age group, such as those conducted in China [16] and Lebanon [35]. In contrast, other studies have identified three dietary patterns in this population group, such as the “Healthy”, “Transitive” and “Western” dietary patterns in China [36], the “Healthy,” “Western,” and “Sweet-Dairy” in Iran [17] and the “Sweet & Fat”, “Refined Cereals”, and “Animal Products” in Europe (Hebestreit 2017). It is important to highlight that factor analysis is a data-driven approach and discrepancies in the dietary assessment methodologies, the types and number of food groupings and the statistical analysis techniques may explain the observed variability in the identified dietary patterns across the literature [37]. It is noteworthy that the dietary patterns described in our study have been also reported by a previous study conducted amongst Emirati adults [38], while noting that the latter had also identified a third pattern, namely the “Traditional Emirati Pattern”. This discrepancy in relation to the traditional pattern is not surprising given that adherence to traditional diets has been repeatedly shown to be higher in older age groups compared to children and adolescents, a phenomenon that may be linked to the nutrition transition [39]. In fact, amidst the globalization and Westernization of diets and lifestyles, children are always more prone to the adoption of new convenience food consumption habits that tend to increase their adherence to Western dietary patterns [39,40], while decreasing their adoption of traditional diets.

One of the main criticisms of the dietary pattern approach is the labelling of the factors (i.e., patterns), which may be based on the subjective interpretation of the investigators. In our study, the correlations of the patterns’ scores with energy and nutrients’ intakes allowed us to further characterize the factors and justify the labelling. Like most studies examining dietary patterns, the pattern associated with higher fiber intakes and lower saturated and trans-fat intakes was labeled, in our study, as the “Healthy” pattern, whereas the one associated with lower fiber and calcium intakes and higher saturated/trans-fat intakes, was depicted as the Western pattern [35,41,42].

In agreement with findings reported by studies conducted elsewhere [35,43], the Western Dietary Pattern was characterized by higher intakes of animal proteins, refined grains, and starchy vegetables, while being negatively associated with mixed traditional dishes, legumes, and nuts. On the other hand, the Healthy Dietary Pattern was characterized by higher consumption of fruits and vegetables, olive and vegetable oils, whole grains, and milk and dairy products, and negatively associated with fast food, salty snacks, sweets and sugar-sweetened beverages, a finding that is also in line with the literature [36,41]. The study findings showed that higher adherence to the Healthy dietary pattern was associated with lower BAZ, and that children with obesity had a significantly lower adherence to the Healthy pattern. These results are consistent with those reported from China by Shang et al (2012) who showed that, compared to children (6−13 years old) with the highest adherence to the healthy dietary pattern, the odds of obesity were significantly higher for children adhering to the transitive pattern (Odds ratio (OR): 1.11; 95%CI: 0.89–1.38) and the Western pattern (OR: 1.80; 95%CI: 1.15–2.81) [36]. Similarly, in a study conducted amongst 7−11 year old children in Iran, Bahreynian et al (2013) showed that compared to girls in the highest quartile of the healthy dietary pattern, those in the second quartile were more likely to have higher BMI (OR:2.23,95% CI: 1.003, 4.96) [17]. These observations may be explained by the combinations of foods and nutrients that characterize the healthy dietary pattern and which may play synergistic roles in preventing excessive weight gain. For instance, fruits, vegetables and whole grains contain dietary fiber, phytochemicals and other bioactive compounds that may modulate appetite hormones such as glucagon-like peptide (GLP-1) and peptide YY (PYY) [44], improve the gut microbiome [45], and enhance insulin sensitivity [46], all of which may decrease the risk of excessive weight gain [44]. Higher intakes of dairy products, another hallmark of the identified healthy dietary pattern, have been also previously linked to reduced obesity risk in children, an observation that may be explained by the independent or synergistic effects of calcium and dairy protein on the regulation of body fat, lipolysis, lipogenesis and thermogenesis [47,48]. Surprisingly, the healthy dietary pattern has loaded positively on “olive and vegetable oils” intake. Studies have suggested that, as compared to saturated fat, unsaturated fatty acids from olive oil and other vegetable oils can play a role in decreasing the risk of obesity in childhood [49]. In fact, unsaturated fatty acids, and particularly monounsaturated fats, may have an impact on thermogenesis, the lipolytic activity of the adipocyte and the modulation of appetite, which may contribute to the regulation of body weight in childhood [49]. In addition, acknowledging that higher intakes of dietary energy and total fat may contribute to creating a positive energy balance and increasing the risk of obesity [50], the fact that the Healthy dietary pattern was characterized by lower consumption of energy-dense foods (fast food, salty snacks, sweets and sugar-sweetened beverages), and lower energy and fat intakes may also explain the observed inverse association between adherence to the Healthy pattern and BAZ.

It is important to recognize that, in contrast to our results, other studies did not find any association between dietary patterns and pediatric overweight or obesity, highlighting the controversy in the literature [43]. For instance, while Choi et al (2011) [43] have identified three dietary patterns amongst 7–8 year old children in Korea (Korean, modified Western, and Western patterns), they showed that the odds of obesity did not differ between the three dietary patterns. In addition, unlike the findings reported by some other investigations [16], the association between the Western pattern scores and BAZ did not reach statistical significance in our study. These results may be explained by some of the characteristics of the identified Western pattern. In fact, although the Western pattern loaded positively on refined grains, starchy vegetables and animal-based products, it was not associated with higher intakes of sweets, sweetened beverages, fast food or total fat. This may have diluted the observed association between this pattern and BMI. It is important to acknowledge that the specific foods contributing to each pattern may vary between studies given the data-driven nature of the factor analysis [40]. In addition, the discrepancies between the various study findings may arise from differences in age-groups, selection of food groupings, number of retained patterns and their labels as well as the type of confounders the studies have adjusted for [51,52].

The findings of this study should be interpreted considering the following limitations. First, the three selected Emirates for this study collectively house 85% of the UAE’s population, and therefore the exclusion of the remaining 15% residing in the other four Emirates may constitute a limitation. Second, and importantly, the cross-sectional design of this study constitutes a methodological limitation with respect to causal inference. Because dietary intake and anthropometric outcomes were measured simultaneously at a single time point, it is not possible to establish the direction of any observed associations. Specifically, it cannot be determined whether adherence to the Healthy dietary pattern preceded and contributed to lower BMI-for-age Z-scores, or whether children’s weight status influenced their dietary choices. This is particularly relevant in the context of pediatric obesity research, where reverse causation cannot be excluded since parents of children with obesity may modify their child’s diet in response to weight concerns. Accordingly, the findings of this study should be interpreted as cross-sectional associations, and causal conclusions should be avoided. Prospective longitudinal studies, or ideally intervention trials, are needed to confirm the direction and causality of the associations observed in this study. The third limitation is possibly related to the factor analysis methodology, a data-driven approach which tends to characterize population-specific patterns [40]. Therefore, our results possibly represent patterns that are, in some features, specific to the pediatric population in the UAE. In addition, the use of factor analysis entails several arbitrary assumptions such as those related to the choice of food groupings, the number of identified factors as well as their labels [51,52]. However, to decrease subjectivity in these approaches, the food groupings adopted in this study were comparable to those described by others [16,17,43] and the selection of the retained factors was conducted after the evaluation of the screen plots and eigenvalues. Another limitation stems from the adoption of the 24-HR for dietary assessment, a method that may be prone to reporting bias by the caregivers [53]. Nevertheless, despite its reliance on memory and the potential day-to-day variation in consumption patterns, the 24-HR approach was found to accurately estimate dietary energy intakes at the population level [54]. Moreover, the adoption of the multiple pass approach when implementing the recalls may have helped in attenuating the potential limitations of the 24-HR [55]. Additionally, to reduce interviewer errors, the 24-HRs were administered by research nutritionists who underwent extensive training prior to the initiation of data collection. The nutritionists who performed the interviews were also trained to avoid verbal and non-verbal judgmental cues in order to reduce social desirability bias.

Conclusions

In conclusion, this study identified two major dietary patterns among school-aged children in the UAE: the “Western Pattern” and the “Healthy Pattern,” with the latter being associated with lower risk of obesity in the study population. These findings should foster the development of culture-specific programs and interventions aimed at curbing the burden of pediatric overweight/obesity, which has been shown to be on the rise in the UAE [56]. Interventions promoting the adoption of heathier dietary patterns, characterized by higher intakes of fruits, vegetables, whole grains and dairy products, while being low in fast food, sweets, sweetened beverages and other energy-dense foods should be developed. In fact, it has been shown that the provision of dietary guidance using a dietary pattern approach tends to be clearer and easier to comprehend and implement than approaches focused on solely nutrients, and this may be specifically valid for children and their caregivers [51]. Recognizing that healthy eating patterns adopted in childhood may exert immediate nutritional benefits, while also potentially tracking into the adult years, interventions that aim at fostering and establishing healthier food consumption habits and dietary skills in childhood may modulate the individual’s lifetime risk for diet-related morbidity.

Supporting information

Acknowledgments

The authors gratefully acknowledge all the children and parents who participated in the study.