Purpose
Good nutrition in early childhood is essential for brain development, growth, and the establishment of healthy eating behaviours. Until now, food hardship (ie, any type of food deprivation) and its relationship with nutrition of children under five has not been broadly investigated in Aotearoa New Zealand. In this study, data from the Growing Up in New Zealand longitudinal study was analysed to explore food hardship as experienced by under five-year olds.
This study sought to understand the relationship between household food hardship and early childhood nutrition, specifically whether a mother’s report of being forced to buy cheaper food or having to go without fresh fruit and vegetables to pay for other things they need, plus the use of special food grants and food banks to obtain sufficient food, had an impact on their child’s breastfeeding, fruit and vegetable intake, and unhealthy food and drink intake in the preschool period. Data from the contemporary Growing Up in New Zealand longitudinal cohort study of more than 6,000 children, followed from before their birth in 2009/10, was used to paint a detailed picture of food hardships experienced by young children.
Methodology
The Growing Up in New Zealand study Growing Up in New Zealand is a contemporary longitudinal study, tracking the development of approximately 6,800 children from before birth until they are young adults (Morton et. al., 2010). Growing Up in New Zealand recruited pregnant women whose babies were due between 25th April 2009 and 25th March 2010 and the cohort has been demonstrated to be generalisable to all national births in the 2007-2010 period (Morton et al., 2015). The scale and diversity of the cohort allows for robust analyses by ethnic group and socioeconomic position.
Growing Up in New Zealand collected information at face-toface interviews with mothers or primary caregivers when the cohort children were 9-months and 54-months old; data collection wave one (DCW1) and data collection wave one (DCW5). Information was gathered about three aspects of food hardship: buying cheaper food in order to pay for other things needed, accessing food assistance (special food grants or food banks), and going without fresh fruit and vegetables to pay for other things needed.
The objectives of the research were addressed in sequence. To answer research questions 1a, b and c, descriptive statistics such as counts and percentages (proportions) were reported for the total cohort, and by ethnic group, with chi square tests for difference used to determine statistical significance. Derived variables were created to explore the coexistence (overlap) of food hardships at the same time point, and also the proportion of families moving in and out of different food hardship (transitions) between the two time points. To enable a health equity focus and in response to the higher proportions of Māori and Pacific children living in households experiencing food hardship, the remaining
analyses considered these two ethnic groups separately for all analyses, in addition to the total cohort.
For research question 2a, associations were explored between the food hardship indicators and maternal and household characteristics available. Univariate logistic regression was used to determine odds ratios to show the burden of food hardship among particular population groups. To answer research question 3, associations were explored between food hardship and child nutrition-related variables at each time point respectively (for total cohort and then stratified by ethnic group). Logistic regression was used to determine odds ratios to show the burden of poor nutrition for particular population groups. Univariate logistic regressions were carried out for each nutrition indicators and list of predictors separately. Final multivariate regression models were produced for each of the nutrition indicators associated with the three food hardships at the two time points, adjusted for potential confounders of maternal age and education, child ethnic group (in the total cohort models), equivalised household income and neighbourhood deprivation (recognising that neighbourhood food environments that adversely affect child nutrition may be socially patterned).
All missing data for indicators, predictor and sociodemographic variables were examined (see Appendix 1). No imputations were carried out for missing data. Three variables in the dataset had more than 10% missing data for food hardship indicators (Appendix 1 Table 5): household income at nine months and 54-months, and breastfeeding to 12 months.
Those who declined to answer, or did not know their household income, were more likely to experience any of the three food hardship variables (Appendix 1 Table 6). Those who were missing breastfeeding data were only slightly more likely to experience any of the three food hardship variables (Appendix 1 Table 7).
A significance level of 0.1 was used for inclusion in the final models based on the univariate regressions, and p<0.05 was considered statistically significant for the final models. Statistical analyses were performed in Stata/SE 15.0 on the external Growing Up in New Zealand data access platform (Guacamole v.0.8.4).
Key Results
- The research found that food hardships were prevalent among families of infants and pre-schoolers.
- All three food hardships were more common in the first year of life compared to the preschool years.
- Ethnic inequities in food hardship were evident from infancy, and especially marked for infants experiencing all three forms of food hardship:
- One in four Māori infants and almost one in three Pacific infants lived in households that reported use of a special food grant or food bank in the previous year compared to one in fifteen Pakeha infants.
- While only two percent of Pakeha infants experienced all three food hardships at age nine months, one in six Pacific infants (15.5 percent) and one in eleven Māori (9.1 percent) experienced these.
- All measures of food hardship in the study were separately associated with poorer nutrition in children (regardless of household income, ethnicity, mother’s age and education).Children living in families that experienced food hardship were more likely to:
- have stopped breastfeeding before their first birthday
- tried unhealthy food and drinks early in infancy
- had fewer servings per day of fruit or vegetables at nine-months
- a less varied intake of fruit and vegetables at four-years of age
- consumed three or more soft drinks a week at four-years of age.