Discussion

This research has explored food and beverage consumption and oral health-related factors in a cohort of Australian children aged 3 and 4 years. It has shown associations between the types of childcare these children experience and the frequency with which they consume particular discretionary foods and beverages. No associations were observed between childcare type and core food consumption, tooth-brushing, dental visiting or parent-reported child oral health status. However, childcare type was significantly associated with the consumption of discretionary items, particularly sweet beverages – fruit juice/fruit drink and soft drink, and ice cream across these ages, which are associated with increased risk of dental disease.

In the current study, the intake of core foods and beverages such as fruits, vegetables and milk was similar across the different types of childcare. Across all childcare groups, many children consumed fruits and vegetables less than once a day (30–51 % of children aged 3 years and 19–33 % of children aged 4 years). The dietary guidelines recommend children of this age consume 2½ to 4½ servings of vegetables and 1 to 1½ servings of fruit each day. It is unlikely that children who do not consume these items daily are meeting their recommended intake⁽³³⁾. A substantial proportion of children were consuming discretionary items twice a week or more, and the results suggest that many children are consuming multiple discretionary items more than twice a week. The dietary guidelines recommend children aged under 8 years avoid discretionary items or restrict consumption to ½ a serving per day⁽³⁴⁾. The cumulative intake of multiple discretionary items suggests that many children in this study are exceeding these recommendations. Discretionary food consumption can displace core foods, which contain the nutrients and energy children need to develop^{(34,Reference Johnson, Bell and Zarnowiecki35)} . Discretionary foods also tend to be high in salt and sugar, and regular provision may encourage children to develop a preference for these flavours^{(Reference Birch36)}. There is evidence that children’s dietary intake may track across time^{(Reference Mikkila, Rasanen and Raitakari14,Reference Spence, Campbell and Lioret37)} and high consumption of discretionary items can lead to excessive energy intake. In the long term, this can cause weight gain and increase the risk of diabetes, CVD and tooth decay^{(Reference Maffeis and Tato38,Reference Tinanoff and Palmer39)} .

At 3 years of age, compared with children receiving PCO, children attending FCO were 59 % less likely to be consuming fruit juice/drink and 71 % less likely to be consuming ice cream. Children attending a mixture of formal and informal childcare were 68 % less likely to be consuming fruit juice/drink. At 4 years of age, associations were only seen with sweetened beverages – fruit juice/drink and soft drink – with children attending either FCO or ICO, 58–86 % less likely to be consuming these beverages compared with children in PCO. Although these results were statistically significant, the wide CI indicate the magnitude of association may be much smaller (as low as 4 %) or higher (88 %) than observed. Fruit juices and fruit drinks tend to be perceived as healthier than soft drinks^{(Reference Bucher and Siegrist40)}, despite being similar in terms of energy density, sugar content and the detrimental effect on teeth^{(Reference Birkhed41)}. Although 100 % fruit juice may have some beneficial nutrients and are counted as one serve of fruit in the Australian Dietary Guidelines⁽¹⁾, they lack the fibre and are less satiating than a piece of whole fruit^{(Reference Bolton, Heaton and Burroughs42,Reference Flood-Obbagy and Rolls43)} . Additionally, young children innately preference sweet flavours^{(Reference Birch36)}, so providing fruit juice and fruit drinks at an early age can encourage them to preference these beverages over water and milk.

There are few published studies comparing dietary intake by type of childcare and those that exist focus on formal childcare settings rather than informal childcare. In Finland and Canada, children in formal childcare were less likely to consume soft drink compared with those not in childcare, but no difference in fruit juice consumption was observed^{(Reference Lehtisalo, Erkkola and Tapanainen44,Reference Pabayo, Spence and Cutumisu45)} . This is interesting considering both fruit juice and soft drink were less likely to be consumed by children attending care in the current study. The timing of data collection may provide some explanation for these differences, with data from Canada and Finland collected in the mid-2000s, whilst data for VicGen were collected from 2011 onwards. The increased focus on childhood obesity rates in the early to mid-2000s^{(Reference Hesketh and Campbell46,Reference Swinburn and Wood47)} encouraged the introduction of nutrition policies in many early childhood services^{(Reference Montague48)}. Policies tended to promote water and milk as the only beverages offered to children in these settings whilst discouraging sweet beverages, including fruit juice^{(Reference de Silva-Sanigorski, Bell and Kremer21,Reference Mazarello Paes, Hesketh and O’Malley49,Reference Ritchie, Sharma and Gildengorin50)} . Compared with the Canadian and Finnish children in these studies, children in the VicGen study may have been more likely to be attending childcare where beverage policy may have been implemented for many years.

In addition to childcare centre policies, in early 2012, the National Quality Framework introduced new quality standards across formal childcare settings in Australia, which included guidelines in relation to the provision of nutritious foods in formal childcare settings⁽¹⁹⁾. Childcare services are assessed on seven quality areas and given a rating, which they must display at their centre. Services must meet a certain standard or be working towards the standard to continue operating. The implementation of this framework commenced around the same time as 3-year-old data collection in this study, and it is likely that the potential effects of the framework were not immediately seen as services worked towards meeting the benchmarks. The nutrition component of the National Quality Framework is relatively small, and the degree to which services adhere to these guidelines is unclear, with recent research showing discretionary foods are regularly offered at childcare centres^{(Reference Wallace, Costello and Devine24)}. This is additionally complicated across family day care services, where carers provide childcare out of their own home to a small number of children. The ratio of food provided by the childcare provider to food provided by parents may vary across different family day care services.

The reasons for consumption differences between children attending ICO and a mixture of formal and informal childcare are more difficult to ascertain given the lack of existing evidence about what occurs in informal childcare settings. We theorised that formal childcare policy and guidelines may account for the associations seen with children in a mixture of childcare types. However, a sensitivity analysis was conducted to explore this at age 4 years and there were no associations between the proportion of time children spent in each type of childcare and their consumption of soft drink or fruit juice/drink. This suggests that factors across the formal and informal childcare environment may influence beverage intake for these children. Additionally, the differences seen between ICO and PCO warrant further exploration. Grandparent care is the most common type of informal childcare used in Australia⁽¹⁷⁾; however, there is little published evidence about food and beverage consumption in this environment. Traditionally, grandparents have been considered treat givers and parents tend to report their children are likely to receive unhealthy foods and beverages from grandparents^{(Reference Roberts and Pettigrew51)}. However, emerging research has challenged this with grandparents who have reported the healthiness of an item as the main influence on the foods and beverages they provide to their grandchildren^{(Reference Jongenelis, Talati and Morley52)}. Additional evidence suggests that as the amount of time a grandparent cares for their grandchild each week increases, they tend to provide more parent-like care, where unhealthy foods and beverages are limited^{(Reference Farrow53)}. Further research into grandparent childcare is warranted to explore the provision of foods and beverages in this environment.

The oral health variables used in this study were selected to capture parent reports of overall oral health, dental visiting patterns and oral hygiene behaviours. While no significant relationship was observed between childcare type and any of these variables, this analysis has identified oral health practices that fall short of recommendations. To maintain good oral hygiene, teeth should be brushed twice a day⁽⁵⁴⁾; however, at age 3 years, only 30–40 % of children across all childcare types were brushing their teeth more than once a day, and this increased only slightly by age 4 years. These results suggest oral health promotion strategies targeted at families during this time were not eliciting the desired behaviours in this group and reasons for this need to be explored. It is also recommended that children visit a dental professional by the time they are 2 years old⁽⁵⁴⁾. In 2011, all children under 12 year of age were eligible for general dental services through the public system and had priority access⁽⁵⁵⁾; however, only 18–36 % of 3 year olds and just over 50 % of 4 year olds in this sample had visited a dental professional. Some families considered the oral health check they received as part of the VicGen study to be sufficient, as reported anecdotally during study visits. Geographical barriers may have also played a role, given that a large proportion of this cohort resided in rural areas and only two-thirds of those eligible for public dental care in rural areas live within a 20 km radius of a public dental clinic⁽⁵⁶⁾. The reasons behind not visiting a dental professional are likely to be multi-factored and will be explored in further publications from the VicGen study.

In Victoria, most parents will receive oral health information from the Maternal and Child Health Service at their child’s 8-, 12- and 18-month health checks. However, there is limited published evidence around the implementation and effectiveness of other oral health promotion strategies in different childcare settings in Victoria. There is evidence around the effectiveness of an outreach programme where dental professionals visit preschools and reduce barriers to attending dental services^{(Reference Mason, Mayze and Pawlak57)}, which could theoretically be applied to childcare services. Currently, Smiles 4 Miles⁽⁵⁸⁾ and the Achievement Program⁽⁵⁹⁾ are two ongoing programmes run statewide that have a focus on oral health promotion. At the time of data collection, however, very few childcare services were participating in Smiles 4 Miles, with the focus on participation of kindergartens. The Achievement Program, whilst targeting childcare services, was in its infancy at this stage, with oral health messaging starting to be incorporated into the programme. The potential influences of these programmes are unlikely to have had an effect by the time of data collection for the VicGen cohort, but families may be influenced by the structure and expansion of these programmes over time.

The results presented must be interpreted in the context of the limitations of this research. The sample includes participants from metropolitan, regional and rural Victoria but may not necessarily be representative of the broader population. The FFQ captures a broad range of core and discretionary foods and beverages; however, as it was designed for use in relation to oral health, it only measures the frequency with which items are consumed, not the actual amount as no serving size was specified. Also, parents may not always know what their child is consuming if they are not in their care. The care-type measure is also relatively crude. Parents reported childcare during a typical week over the last month, which may not represent the childcare their child usually receives. Additionally, both measures were parent reported and thus open to bias compared with more objective measures. There may also be potential differences in the types of informal childcare (e.g., grandparents v. nannies v. friends) that make a difference to these outcomes that are not captured in these data.