Weight status and obesity-related dietary behaviours among culturally and linguistically diverse (CALD) children in Victoria, Australia

This study utilised baseline (2014) data of children participating in Healthy Together Victoria and Childhood Obesity Study (HTVCO). HTVCO is a repeat cross-sectional study which aims to examine the effect of Healthy Together Victoria (HTV), a community-based lifestyle initiative of the State Government, on the prevalence of overweight and obesity and associated risk factors among school-aged children [23].

Data collection comprised objective measures of height and weight (3–5 min per participant) and a self-reported questionnaire administered via an electronic tablet (20–30 min to complete). All measurements were conducted during the school day by trained research staff.

Of the 2407 participants, 50.2% were female, and the mean age was 11 years (SD = 1.1). More than half (54.4%) of children lived in major cities, and 56.7% lived in areas categorised as being in the lower two SEIFA quintiles. Just under a third (29.2%) of participants were classified as CALD: 7.9% European, 6.1% Middle-Eastern, 4.9% Southern Asian, 4.1% Southeast Asian, 3.4% African and 2.9% Eastern Asian backgrounds.

Over a third (37.3%) of all participants were classified as overweight or obese (Table 1). Except for Middle-Eastern children, no significant differences in overweight/obesity prevalence were observed between CALD subgroups and the non-CALD group. More than half (53%) of Middle-Eastern children were categorised as overweight/obese, compared to 38.5% of non-CALD children (p < 0.001).

Across CALD subgroups, several variations in dietary behaviours were found (Table 1). African, Southeast Asian, Southern Asian and Middle-Eastern children had a significantly higher weekly consumption of takeaway foods compared to non-CALD children (p < 0.05). Frequent takeaway food consumption was highest among the Southeast Asian group, with 58.2% of children. Except the Middle-Eastern group, there were no significant differences in the proportion of children consuming two or more sugary or salty snacks per day among CALD subgroups compared to the non-CALD group. Snack consumption was almost 10% higher among Middle-Eastern compared to non-CALD children (36.6 and 27.0%, respectively; p = 0.14). SSB consumption frequency was significantly higher among Middle-Eastern (35.4%, p = 0.007) and African (36.6%, p = 0.019) groups compared to the non-CALD group (25.0%).

Model 1 (unadjusted) shows children of Middle-Eastern backgrounds as 1.94 times more likely than non-CALD children to be overweight or obese (CI = [1.36, 2.76]; p < 0.001) (Table 2). This association remained statistically significant after adjustment (OR = 1.99; 95% CI = [1.31, 3.04]; p = 0.001) (Model 2). No other significant associations in the odds of overweight and obesity among the other CALD sub-group compared to the non-CALD group were observed.

Associations with dietary behaviours differed across CALD subgroups. Apart from takeaway food and SSB consumption in the Middle-Eastern group, all other associations between CALD subgroups and dietary behaviours remained statistically significant after adjustment. In the adjusted model, children of Southeast Asian and African background were 1.58 (95% CI = [1.06, 2.35]; p = 0.024) and 1.61 (95% CI = [1.17, 2.21]; p = 0.003) times more likely, respectively, to consume takeaway one or more times per week than non-CALD children. Compared to the non-CALD group, children of Middle-Eastern backgrounds had 1.5 times the odds of consuming two or more EDNP snacks per day (95% CI = [1.05, 2.15]; p = 0.026). The odds of consuming one or more SSBs per day were significantly higher among children of African backgrounds (OR = 1.76; 95% CI = [1.30, 2.38]; p < 0.001).

In the present study, over half (53%) of children of Middle-Eastern backgrounds were classified as overweight or obese, compared to 36.7% of non-CALD children. This finding is supported by previous Australian research, which has consistently identified high rates of overweight and obesity among Middle-Eastern background children over the past two decades [9, 29]. The proportion of overweight or obese Middle-Eastern background and non-CALD children in the present study is considerably higher than that reported in the 2015 SPANS (37.5 and 13.3%, respectively) [19]. This is likely due to methodological differences in SPANS, such as a opt-in recruitment resulting in a response rate of 67.9% and the use of International Obesity Taskforce cut-points to classify weight status, which typically has more conservative estimates of overweight and obesity than the WHO system [30, 31]. Nevertheless, findings from the present study add to the large body of evidence identifying a need to address the disproportionate prevalence of overweight and obesity among children of Middle-Eastern backgrounds.

Contrary to prior research, overweight and obesity prevalence among all other CALD subgroups did not differ significantly to that of non-CALD children. Australian children of African [9], European [9‐11] and Pacific Islander [9, 10] backgrounds have previously been found to have an increased risk of excess weight. Due to the small sample sizes of CALD subgroups in the present study, there may not have been sufficient power to detect differences in weight status compared to the non-CALD group. Differences in identified at-risk groups between the present and previous studies may also be somewhat explained by alternate methods of classifying CALD background. For instance, Waters et al. [11] determined CALD background according to maternal region of birth. An additional 24% of children were classified as CALD using this method compared to language spoken at home, thus resulting in a higher potential of identifying overweight and obesity among those classified as CALD [11]. Additionally, children of Pacific Islander backgrounds were excluded from analysis in the present study due to insufficient numbers. As such, future research should endeavour to recruit a sample large enough to include and be representative of all CALD subgroups so that at-risk groups can be accurately identified.

Excess takeaway consumption appeared to be a significant issue among several CALD subgroups in the present study. Excluding children of European and Eastern Asian backgrounds, more than half of children among all other CALD subgroups consumed takeaway one or more times per week. This is concerning, as takeaway foods are typically high in energy, saturated fat, salt and sugar, and have been associated with excess weight in children [32‐34]. Furthermore, higher consumption of takeaway foods has been shown to not only be associated with an increased risk of obesity; but also other metabolic disorders such as insulin resistance and type 2 diabetes; along with cardiovascular disorders [35]. There are many aspects that may influence takeaway consumption, including its relative affordability, accessibility, convenience and media advertising [36, 37]. However, the persisting relationship between takeaway consumption and Southeast Asian or African background in the adjusted regression model is indicative of an independent cultural influence. For example, high cultural regard for Western foods and dietary acculturation have been identified as factors influencing takeaway consumption among African migrants in Australia [38, 39].

Overall, the proportion of children consuming one or more SSBs per day within CALD subgroups was similar to that of non-CALD group in the present study. However, a significantly higher prevalence of SSB consumption was identified among Middle-Eastern and African background children. Furthermore, compared to non-CALD children, African children had almost two times the odds of excess SSB consumption. Similar results among children of Middle-Eastern backgrounds were reported in SPANS, who were found to have a significantly higher prevalence of soft drink consumption and availability of soft drinks in the home compared with non-CALD children [19]. To our knowledge, this is the first study to investigate SSB consumption among children of African backgrounds in Australia. Considering the association between SSB consumption and increased weight status [40], this behaviour among children of African backgrounds warrants further investigation. Additionally, this finding highlights the importance of investigating obesity-related dietary behaviours among minority CALD subgroups.

Children of Middle-Eastern backgrounds were also found to have a significantly higher prevalence of consuming two or more sugary or salty snacks per day compared to non-CALD children. This finding is consistent with previous research [14, 19, 41]. In an analysis of the cumulative intake of ‘junk’ foods (e.g. sweet or savoury biscuits, ice-cream) among children in NSW, Boylan et al. [41] found that over half of children from Middle-Eastern backgrounds were high junk food consumers. It has also been shown that children of Middle-Eastern backgrounds are often rewarded for good behaviour with sweets [14, 19], which may partially explain the higher odds of excess snack consumption compared to non-CALD children in the present study.

Data for this study were drawn from a large sample of Victorian primary school children spanning broad demographic and geographic areas. At a school-level, non-response bias was minimised due to the opt-out recruitment process. Therefore, the present study may have more accurately estimated overweight/obesity prevalence compared to research with opt-in procedure [42]. A further strength is the use of standardised protocols for the collection of height and weight measurements.

There are also some limitations of this research to consider. Firstly, as these data are cross-sectional, causality cannot be inferred. In addition, BMI is not able to distinguish between fat and fat-free mass and therefore has limitations as a measure of overweight and obesity at an individual level. However, at a population-level, BMI is an accurate, cost and time efficient measure for population surveillance of childhood overweight and obesity and has very high specificity [43]. Dietary behaviours were self-reported by children and were therefore subject to recall and measurement bias. Food frequency questionnaires are commonly completed by parent proxy in children less than 12 years old due to cognitive development among young children potentially influencing food recall and estimated portion sizes [44]. However, parents of children may not be aware of foods consumed out of the home, and there is research to suggest that children as young as 8 years old are more accurate reporters of food frequency consumption than their parents [45]. The classification of CALD background according to language spoken at home is a potential limitation of this study, as it is possible that some CALD children speak predominantly English at home. Finally, while the sample was large, there was not sufficient power among all CALD subgroups to detect differences in weight status or dietary behaviours compared to non-CALD children. Future research examining differences in CALD groups speaking predominantly English compared to those speaking predominantly their native language would shed light on whether disparities between groups are influenced by lifestyle, genetic or a combination of factors. Examining the effects of acculturation on behaviours that influence weight status in migrant children living in Australia would also be beneficial [14]. In depth qualitative research to understand dietary choices in these CALD groups is also required.