Breakfast skipping and overweight/obesity in first grade primary school children: A nationwide register‐based study in Iceland

Although several studies have revealed an association between skipping breakfast and overweight (OW) or obesity (OB) in older children and adolescents, less is known about that association in younger children. The purpose of our study was to assess the association between skipping breakfast and OW/OB in children in the first grade. The sample included 4360 children (51.5% boys) aged 5.6 to 7.4 years who participated in the annual health examination in Iceland during 2016 and 2017, completed by 91% of all first graders in Iceland. Binary logistic regression analysis with odds ratios (ORs) and 95% confidence intervals (CIs) was used to assess the association between skipping breakfast and OW/OB. Skipping breakfast was assessed as not eating breakfast on the day of the assessment, whereas OW or OB was based on measured height and weight relative to the International Obesity Task Force reference. The final analyses were adjusted for bedtime, well‐being in school, commuting to school and physical activity. 7.2% of the boys (n = 162) and 7.5% of the girls (n = 158) had not eaten breakfast. After multivariable adjustment, a statistically significant association emerged between skipping breakfast and OW/OB in girls (OR 1.66, 95% CI 1.17‐2.36) but not in boys (OR 1.02, 95% CI 0.63‐1.63). Because the study's results suggest an association between skipping breakfast and OW/OB only in first‐grade girls in Iceland, sex‐based differences should be further investigated to inform future strategies for preventing OW and OB in young children.


| INTRODUCTION
Worldwide, the number of children with overweight (OW) or obesity (OB) increased from 32 to 42 million between 2000 and 2013, and similarly rising rates in more recent years have ranked OW and OB in children as one of the greatest challenges for today's healthcare systems. 1 Estimated costs related to OW and OB total EUR 10 billion in Europe, 2 and in the United States, medical costs associated with common diseases caused by OB are projected to skyrocket by USD 48 to 66 billion per year by 2030. 3 Furthermore, OB occurring in childhood which passes through to adulthood increases the risk of OB-related complications. 4 In a recent prospective study involving 17-to 23-year follow-up of a nationally representative cohort of US adults, skipping breakfast was associated with a significantly increased risk of mortality from cardiovascular disease. 5 Both cross-sectional 6,7 and prospective studies 8,9 additionally suggest an association between breakfast habits and OW or OB in children and adolescents. Such findings have been partly supported by a recent meta-analysis of 14 cross-sectional studies and two prospective studies on skipping breakfast and OW or OB in children and adolescents. 10 For example, in a cross-sectional study of 6941 children aged 9 to 11 years from 12 countries representing a wide range of economic development, Zakrzewski et al. 11 found that regularly eating breakfast was associated with lower body mass index (BMI) than occasional and rare consumption were. In another meta-analysis of 16 cross-sectional studies on skipping breakfast and OW and OB amongst 59 328 children or adolescents aged 6 to 20 years from Europe, the authors concluded that eating breakfast was associated with a reduced risk of OW as well as OB. 12 However, in this meta-analysis, only one study had included relatively young children, with one group aged 6 to 10 years vs another aged 11 to 14 years, 13 and it remains unclear whether 6-year-olds are truly comparable to 10-yearolds. After all, children less than 7 years old have different milestones in both physical and cognitive development compared to 10-year-olds, 14 and they are far more dependent upon their parents than older children are. Thus, OB prevention efforts may need to begin earlier in life, when children remain dependent upon parental feeding practices. To support such efforts, it is important to gain insight into the breakfast habits of younger children from more homogeneous age groups.
From another angle, studies on the relationship between breakfast habits and OW or OB in young schoolchildren have been few, and knowledge about sex-based differences in that relationship remains scarce. Recent numbers indicate that more girls than boys have OW in Nordic countries such as Norway and Iceland, 15,16 while cross-sectional studies have revealed that more girls than boys are skipping breakfast, 17 as well as that skipping breakfast increases in frequency as children age. 18 Therefore, the chief aim of our study was to explore whether an association exists between skipping breakfast and OW or OB amongst children in the first grade in Iceland. A secondary aim was to assess whether the association between skipping breakfast and OW or OB was modified by sex.    What is already known about this subject • Overweight and obesity in children is one of the greatest challenges for today's healthcare systems.

| Anthropometric measurements
• Skipping breakfast is associated with overweight and obesity in school-aged children and adolescents.

What this study adds
• In this nationwide register-based study of Icelandic first grade primary school children, prevalence of breakfast skipping was similar in boys and girls (7.2% vs 7.5%).
• Icelandic girls in first grade primary school who skipped breakfast had 66% higher odds for being overweight/ obese compared to girls who had breakfast.
• The association between breakfast skipping and overweight/obesity was not found in boys. measurements were taken while the children wore light clothing but no shoes. BMI was calculated as weight in kilograms divided by the squared value of height in metres (kg/m 2 ). All anthropometric measurements were performed according to standardized techniques by trained school nurses. Age-and sex-specific international cut-off values for OW and OB proposed by the International Obesity Task Force (IOTF) were used to define OW and OB. 19

| Confounders
Studies have suggested that factors such as physical inactivity and sedentary behaviours, 20,21 reduced sleep length 22,23 and discomfort or stress at school are associated with OW and OB amongst schoolchildren, 24,25 and that those factors are also associated with skipping breakfast. [26][27][28] In our study, physical activity was assessed by asking two questions: 'Do you participate in sport activities, or do you often play outdoors?' with the response categories 'Yes', 'No' or 'Do not know'; and 'How did you get to school today?' with the response categories 'Walking or cycling' or 'Was driven'. Sleep routine was captured by asking 'What time should you go to bed when you have school the next day?' with the response categories 'Before 9 PM', 'After 9 PM' and 'Do not know'. Well-being in school was assessed by asking 'How do you enjoy school?' with the following response categories: 'Enjoying school', 'Not enjoying school' and 'Do not know'.
The Scientific Committee for Health Care in Reykjavik and the University of Iceland (Vís-HH-UI), the National Bioethics Committee in Iceland, the Data Protection Authority in Iceland and the Regional Norwegian Committee for Medical and Health Research Ethics approved our study.

| Statistical analysis
To compare the results with the majority of scientific literature on skipping breakfast amongst children, OW and OB were analysed together as the outcome variable OW/OB.
Differences between children who skipped and did not skip breakfast were analysed with chi-square tests (χ 2 ). Odds ratios (ORs) and 95% confidence intervals (CIs) for OW/OB associated with skipping breakfast were estimated by using logistic regression models adjusted for sport activity, commuting to school, bedtime and wellbeing in school. In those models, children who did not skip breakfast served as the reference category, and the covariates were dummy-      Significantly, greater proportions of boys and girls who skipped breakfast were driven to school than ones who had eaten breakfast.
Moreover, significantly more girls who had eaten breakfast also reported that going to bed before 9 PM was preferable on school nights. Amongst boys who skipped breakfast, 6.8% reported enjoying school, compared to 1.7% of the boys who had eaten breakfast.
The results of the multivariable analyses for boys appear in  T A B L E 3 Logistic regression (Enter) with odds ratios (OR) and 95% confidence intervals (CI) on the association between not eating breakfast and overweight/obesity in Icelandic boys in first grade of primary school 2016/2017 (n = 2246) Amongst children in Iceland's primary schools in our study, the prevalence of skipping breakfast, from 7.2% to 7.5%, takes partial support from the findings of a recent systematic review of 37 studies on the association of skipping breakfast with weight and cardiometabolic risk factors in children and adolescents. 33 According to Monzani et al., 33 skipping breakfast amongst adolescents and girls ranged from 10% to 30% and showed an increasing trend. Results from Sweden (ie, with children aged 7-9 years) and from the Netherlands (ie, with children aged 6 years) included in their review included prevalence rates of skipping breakfast of 4.6% and 6.4%, respectively, which are somewhat less than those rates found in our study.
Our study's major findings include a statistically significant relationship between skipping breakfast and OW or OB in girls but not in boys. That finding was supported by the result of the test of additive interaction, which suggests that sex modified the effect of skipping breakfast on the odds of having OW or OB. RERI exceeded 1, which implies a sufficient-cause interaction without requiring the monotonicity assumption. 30 The few studies that have included the youngest children (ie, 2-8 years old) also reported an association between skipping breakfast and OW or OB 13,34 ; however, only one of them involved stratifying the analysis by sex. 34 Mushtaq et al. reported that skipping breakfast was associated with OW in girls only. 35 Another cross-sectional study in Poland concluded that skipping breakfast or eating it irregularly was a risk factor for OB for girls but not for boys. 36 However, results from longitudinal studies, 37,38 as well as a review of 16 cross-sectional studies and a recently meta-analysis, 10 suggest that skipping breakfast is associated with OW and OB in boys and girls, 12 although not all of those studies involved using formal statistical tests for sex-based differences. Furthermore, in most studies involving statistical testing for effect modification, researchers have used a multiplicative scale with an interaction term, which is unsuitable to assess whether an intervention would benefit some subgroups more than others, which has been identified as an important public health issue. 30 The observed sex-based differences may be attributed to different factors. Results suggest that the foetal environment poses consequences for the long-term cardiometabolic health of infants and children, as well as we females seem particularly susceptible to developing disrupted glucose homeostasis and increased adiposity as a result of exposure to environments high in sugar and in utero undernutrition, respectively. 39 Differences in dietary intake and eating behaviours between boys and girls have also been shown, 40 and along those lines, cultural aspects such as thinness as a criterion for beauty amongst girls may also exert influence. In Japan, a recent study on the association between eating breakfast and childhood OB followed 42 663 children aged 1.5 years until they were 12 years old. 41 The results confirmed that 12% to 32% of their mothers and fathers had skipped breakfast when the child was 1.5 years old and that the stron- increased as a result of the intervention, although the intervention did not affect the combined incidence of OW and OB. 43 However, and surprisingly, because the incidence and prevalence of OB were greater in schools hosting the intervention than in control schools after 2.5 years, the authors concluded that further research is needed to identify approaches to increase participation in breakfast programmes that do not increase OB in students. In Iceland, some schools offer school lunch at a cost; however, no information about children who eventually had lunch arranged by their school was used in our study.
Several cross-sectional studies have revealed an unambiguous association between skipping breakfast and OW or OB in children, especially in ones aged 9 to 16 years. 44,45 In a study of 9-to 11-yearolds from 12 countries all over the world, the association varied by site, and non-significant associations were found in Australia, Finland and Kenya. 11 The level of human development, cultural practices, precision, which may explain their relatively small contribution following adjustment. Third, the prevalence of skipping breakfast was difficult to compare across studies due to differences in its definition. In a recent systematic review of 39 studies on skipping breakfast amongst children and adolescents, 33 a wide range of definitions of skipping breakfast were used, and six of those studies used the same definition used in our study-that is, of not having eaten breakfast on the day of the assessment. Nevertheless, that measure probably does not reflect the daily breakfast habits of all children included. The highly heterogeneous assessment of skipping breakfast is well known and suggested as a topic concerning methodology for future studies to examine as a way to produce more significant results. 33 Last, both skipping breakfast and OW or OB in early childhood could be influenced by different mechanisms, including genetics, factors during pregnancy and both contextual and environmental factors.
In a US study, home environment in relation to food explained a great deal about children's diets, including that both social (eg, mealtime structure) and physical aspects (eg, food availability) of that environment were strongly associated with the consumption of healthy vs unhealthy foods. 48 Several studies have revealed that the prevalence of OB is greater amongst children in rural vs urban areas in some countries 49,50 ; therefore, it is likely that geographical factors influenced the associations observed in our study. However, such factors were not considered in the study due to a lack of access to the children's addresses (eg, whether they lived in rural or urban districts of Iceland) and the unavailability of data regarding their parents' socio-economic status and breakfast habits. However, because breakfast consumption is a parental feeding practice for children in the age group studied, future research on skipping breakfast and OW or OB should include family-related factors.
In summary, our study revealed an association between skipping breakfast and OW or OB amongst first-grade girls in Iceland but not boys. Those sex-based differences should be further investigated to inform future health programmes and early interventions directed towards preventing OW and OB during early childhood.