Lucinda Bell

Dietary patterns of Australian children aged 14 and 24 months, and associations with socio-demographic factors and adiposity

Background/Objectives:Previous research has shown, in predominantly European populations, that dietary patterns are evident early in life. However, little is known about early-life dietary patterns in Australian children. We aimed to describe dietary patterns of Australian toddlers and their associations with socio-demographic characteristics and adiposity.Subjects/Methods:Principal component analysis was applied to 3 days (1 × 24-h recall and 2 × 24-h record) data of 14 (n=552)- and 24 (n=493)-month-old children from two Australian studies, NOURISH and South Australian Infant Dietary Intake (SAIDI). Associations with dietary patterns were investigated using regression analyses.Results:Two patterns were identified at both ages. At 14 months, the first pattern was characterised by fruit, grains, vegetables, cheese and nuts/seeds (‘14-month core foods’) and the second pattern was characterised by white bread, milk, spreads, juice and ice-cream (‘basic combination’). Similarly, at 24 months the ‘24-month core foods’ pattern included fruit, vegetables, dairy, nuts/seeds, meat and water, whereas the ‘non-core foods’ included white bread, spreads, sweetened beverages, snacks, chocolate and processed meat. Lower maternal age and earlier breastfeeding cessation were associated with higher ‘basic combination’ and ‘non-core foods’ pattern scores, whereas earlier and later solid introduction were associated with higher ‘basic combination’ and ‘24-month core foods’ pattern scores, respectively. Patterns were not associated with body mass index (BMI) z-score.Conclusions:Dietary patterns reflecting core and non-core food intake are identifiable in Australian toddlers. These findings support the need to intervene early with parents to promote healthy eating in children and can inform future investigations on the effects of early diet on long-term health.

Short Tools to Assess Young Children's Dietary Intake: A Systematic Review Focusing on Application to Dietary Index Research

Dietary indices evaluate diet quality, usually based on current dietary guidelines. Indices can therefore contribute to our understanding of early-life obesity-risk dietary behaviours. Yet indices are commonly applied to dietary data collected by onerous methods (e.g., recalls or records). Short dietary assessment instruments are an attractive alternative to collect data from which to derive an index score. A systematic review of studies published before April 2013 was conducted to identify short (≤50 items) tools that measure whole-of-diet intake of young children (birth-five years) and are applicable to dietary indices, in particular screening obesogenic dietary behaviours. The search identified 3686 papers of which 16, reporting on 15 tools (n = 7, infants and toddlers birth-24 months; n = 8, preschoolers 2–5 years), met the inclusion criteria. Most tools were food frequency questionnaires (n = 14), with one innovative dietary questionnaire identified. Seven were tested for validity or reliability, and one was tested for both. Six tools (n = 2, infants and toddlers; n = 4, preschoolers) are applicable for use with current dietary indices, five of which screen obesogenic dietary behaviours. Given the limited number of brief, valid and reliable dietary assessment tools for young children to which an index can be applied, future short tool development is warranted, particularly for screening obesogenic dietary behaviours.

The Relationship between Dietary Patterns and Metabolic Health in a Representative Sample of Adult Australians

Studies assessing dietary intake and its relationship to metabolic phenotype are emerging, but limited. The aims of the study are to identify dietary patterns in Australian adults, and to determine whether these dietary patterns are associated with metabolic phenotype and obesity. Cross-sectional data from the Australian Bureau of Statistics 2011 Australian Health Survey was analysed. Subjects included adults aged 45 years and over (n = 2415). Metabolic phenotype was determined according to criteria used to define metabolic syndrome (0–2 abnormalities vs. 3–7 abnormalities), and additionally categorized for obesity (body mass index (BMI) ≥30 kg/m2 vs. BMI <30 kg/m2). Dietary patterns were derived using factor analysis. Multivariable models were used to assess the relationship between dietary patterns and metabolic phenotype, with adjustment for age, sex, smoking status, socio-economic indexes for areas, physical activity and daily energy intake. Twenty percent of the population was metabolically unhealthy and obese. In the fully adjusted model, for every one standard deviation increase in the Healthy dietary pattern, the odds of having a more metabolically healthy profile increased by 16% (odds ratio (OR) 1.16; 95% confidence interval (CI): 1.04, 1.29). Poor metabolic profile and obesity are prevalent in Australian adults and a healthier dietary pattern plays a role in a metabolic and BMI phenotypes. Nutritional strategies addressing metabolic syndrome criteria and targeting obesity are recommended in order to improve metabolic phenotype and potential disease burden.

Impact of a nutrition award scheme on the food and nutrient intakes of 2- to 4-year-olds attending long day care.

OBJECTIVE Early childhood settings are promising avenues to intervene to improve childrens nutrition. Previous research has shown that a nutrition award scheme, Start Right - Eat Right (SRER), improves long day care centre policies, menus and eating environments. Whether this translates into improvements in childrens dietary intake is unknown. The present study aimed to determine whether SRER improves childrens food and nutrient intakes. DESIGN Pre-post cohort study. SETTING Twenty long day care centres in metropolitan Adelaide, South Australia, Australia. SUBJECTS Children aged 2-4 years (n 236 at baseline, n 232 at follow-up). METHODS Dietary intake (morning tea, lunch, afternoon tea) was assessed pre- and post-SRER implementation using the plate wastage method. Centre nutrition policies, menus and environments were evaluated as measures of intervention fidelity. Comparisons between baseline and follow-up were made using t tests. RESULTS At follow-up, 80 % of centres were fully compliant with the SRER award criteria, indicating high scheme implementation and adoption. Intake increased for all core food groups (range: 0·2-0·4 servings/d, P<0·001) except for vegetable intake. Energy intake increased and improvements in intakes of eleven out of the nineteen nutrients evaluated were observed. CONCLUSIONS SRER is effective in improving childrens food and nutrient intakes at a critical time point when dietary habits and preferences are established and can inform future public health nutrition interventions in this setting.

Dietary risk scores of toddlers are associated with nutrient intakes and socio-demographic factors, but not weight status

Aim To determine the convergent validity of the toddler dietary questionnaire (TDQ), previously shown to have good reliability and comparative validity, by assessing whether TDQ-derived dietary risk scores are associated with nutrient intakes, socio-demographic characteristics and weight status. Methods Parents of 117 Australian children aged 12–36 months completed a demographic questionnaire and the 19-item TDQ that assesses intake over the last 7 days of ‘core’ and ‘non-core’ food-groups, which is scored against a dietary risk criteria (0–100; higher score = higher risk). Nutrient intakes were established by applying TDQ food-group nutrient profiles to the amount consumed per week and tallying for all items. Energy and energy-adjusted nutrient intakes across quartiles (Q1–Q4) of dietary risk scores were investigated using linear trend analysis for parametric data. Regression analysis was employed to determine the association between risk score, socio-demographic characteristics and toddler body mass index (BMI) z-score. Results Toddler risk scores (Q1 20.0 ± 2.7 vs Q4 41.6 ± 5.6) were positively associated with energy, total fat, saturated fat, sugar and sodium intakes and negatively associated with protein, fibre, iron, magnesium and phosphorus intakes. Higher risk scores were associated with higher number of people per household (n = 115; β = 2.26 [0.13, 4.38], P = 0.037) and toddler age (n = 115, β = 4.03 [1.10, 6.96], P = 0.008) but not BMI z-scores (n = 114; β = −0.00 [−0.04, 0.03], P = 0.845). Conclusions Dietary risk scores are associated with selected nutrient intakes and socio-demographic characteristics in expected directions, but not BMI z-scores, providing support for use of the TDQ as a dietary screening tool. Longitudinal investigation of the influence of dietary risk on weight status is warranted.

Facilitators and Barriers to the Achievement of Healthy Lifestyle Goals: Qualitative Findings From Australian Parents Enrolled in the PEACH Child Weight Management Program

Objectives: To describe the qualitative research methods used in the Parenting Eating and Activity for Child Health (PEACH) randomized controlled trial and to examine parent‐reported facilitators and barriers to the achievement of program goals. Design: Qualitative study using semistructured interviews. Setting: Parents enrolled in the Australian PEACH randomized controlled trial, a family‐focused child weight management program conducted blinded for review. Participants: A total of 95 parents of overweight children aged 5–10 years participated in face‐to‐face semistructured interviews. Phenomenon of Interest: Factors external to the PEACH intervention that facilitated or inhibited their success. Analysis: Interviews were recorded, transcribed, and analyzed using thematic analysis techniques. Results: Three facilitator themes were identified: (1) internal locus of parental control, (2) external locus of parental control, and (3) child factors. Six barrier themes were identified: (1) internal locus of parental control, (2) external locus of parental control, and (3) child factors; (4) maintenance and managing special occasions; (5) time management challenges; and (6) inconsistencies or lack of support. Conclusions and Implications: The social and environmental barriers and time pressures identified by parents are a common feature of Western family life such that many parents are raising families in stressful situations. Insights provided by this qualitative research provide an important understanding of these experiences. Consideration of such issues in the design and implementation of future child weight management interventions may help to increase the acceptability and effectiveness of such programs.

Interventions for Improving Young Children’s Dietary Intake through Early Childhood Settings: A Systematic Review

Early childhood settings (ECS) offer a unique opportunity to intervene to improve children’s nutrition. This paper reviews the literature on early childhood setting interventions that aim to improve children’s dietary intake. Environmental and individual determinants of children’s dietary intakes were also investigated. Prospective intervention studies targeting centres, staff, parents/caregivers or children, were reviewed. Methodological quality was assessed. Twenty six studies (14 weak, 12 moderate quality) were included. Interventions were delivered primarily via training workshops and/or written materials. Study findings favoured intervention effectiveness in 23 studies. Improvements were seen in children’s intake for 8 out of 11 studies assessing dietary intake outcomes. Small increases in fruit and vegetable consumption were observed in five studies. Most studies measuring parental or centre food provision observed post-intervention improvements across a number of food groups, including fruit, vegetables, whole grains and sweetened beverages. Significant improvements in child, parent and/or staff knowledge, attitudes or behaviours were observed consistently across studies. For those studies that included a comparison group, these improvements were observed only in the intervention group. ECS interventions can achieve changes in children’s dietary intake and associated socio- environmental- determinants, although the quality of current research limits confidence in study findings. Future intervention development needs to carefully consider the behavioural targets, modifiable determinants and utilise age-appropriate and effective behaviour change theory, in addition to inclusion of dietary intake outcomes.

Validation testing of a short food‐group‐based questionnaire to assess dietary risk in preschoolers aged 3–5 years

To the Editor, Accurate measurement of dietary intake is crucial for understanding the relationship between diet and chronic disease, monitoring trends in predictors of health and determining intervention effectiveness. Traditional dietary assessment methods such as food records and recalls are subject to substantial error whereas short questionnairestyle methods enable rapid food intake reporting, increased cooperation and completion, and derivation of food-based data that are useful for monitoring dietary guideline compliance. Dietary assessment in young children is particularly challenging due to variation in diet across brief time periods. Thus age-specific assessment tools are required. The reliable and valid short food-based Toddler Dietary Questionnaire (TDQ) was adapted for use with Australian preschoolers. This study aimed to determine the test–retest reliability, relative validity and convergent validity of this new Preschooler Dietary Questionnaire (PDQ). Primary caregivers of preschoolers aged 3–5 years, recruited via Flinders University newsletter advertisements and a study-specific Facebook page, completed a two-stage online survey. Stage 1 comprised a demographic questionnaire and the 19-item PDQ (PDQ1). Stage 2 (completed 2.1 1.0 weeks later) comprised a second PDQ (PDQ2) and a validated 54-item Food Frequency Questionnaire (FFQ). The PDQ and TDQ are the same apart from the use of age-appropriate portion size categories. Intake is scored against a dietary risk criterion (0–100; higher score = higher risk) (Table 1) and scores categorised into: low (0–24); moderate (25–49); high (50–74) and very high (75–100) dietary risk. Data were analysed using SPSS version 22.0 (IBM SPSS Statistics, IBM Corporation, Armonk, New York, USA). Dietary risk scores (sections 1–3; total) were examined for test–retest reliability (PDQ1 vs PDQ2) and relative validity (PDQave ((PDQ1 + PDQ2)/2) vs FFQ) at the individual (intraclass correlations (ICC), Pearson’s correlations; low ≤0.50; moderate 0.51–0.69; high ≥0.70) and group level (paired t-tests). To assess the strength of agreement between the two methods (PDQave and FFQ), Bland–Altman plots were constructed, assessed visually and linear regression analysis performed to test for systematic bias. Cross classification of subjects into dietary-risk categories was determined and standard linear regression employed to determine convergent validity by assessing the relationship of dietary risk scores with socio-demographic characteristics and Body Mass Index (BMI) z-score, adjusting for covariates. Seventy-four parents (35.5 4.1 years, 81% universityeducated) of preschoolers (54% female, 3.7 0.6 years, BMIz 0.31 1.02, average Index of Relative Socioeconomic Advantage and Disadvantage (IRSAD) score 1003.8 58.66) completed all study questionnaires. Mean total dietary risk scores ranged from 32.7 9.2 (PDQ2) to 35.1 9.7 (FFQ) (Table 1; i.e. ‘moderate’ risk). Total and section risk scores from each PDQ administration were highly correlated (ICC 0.83–0.92) yet statistically different for section 1 (mean bias 4.1, 95% CI 0.5, 7.7, P = 0.027) and total (mean bias 1.5, 95%CI 0.1, 3.0, P = 0.040) risk scores (Table 1). Most (82%) participants were correctly classified (18%, adjacent category) upon each PDQ administration. PDQave and FFQ dietary risk scores were highly correlated for sections 1 and 3 and total risk scores (all r ≥ 0.80), but not for section 2 (r = 0.67) (Table 1). Despite no statistically significant difference between the PDQave and FFQ for section scores (mean bias range: −1.0 section 3 to −2.8 section 1), there was for total risk scores (mean bias −1.6, 95% CI −2.9, −0.4, P = 0.009). Bland– Altman plots (Figure 1) show that the PDQave provides a higher estimate of risk than the FFQ for section and total risk scores (i.e. positive mean differences). Most measurements fell within wide 95% limits of agreement (total risk scores, LOA −11.9, 8.6) and there was no significant linear trend for the fitted regression line (Table 1). Participants were classified into the same (80%) or adjacent (20%) category upon administration of each tool. PDQave scores were significantly negatively associated with the number of people per household (β −0.32, 95% CI −6.69, −0.59, P = 0.020) but not with preschoolers’ BMIz score (β −0.09, 95% CI −0.02, −0.04, P = 0.512). The PDQ is the first tool of its kind for use in Australian preschoolers as it assesses whole of diet intake (core/non-core food groups), allowing comprehensive evaluation of intake against dietary guidelines and thus assessment of dietary risk. Despite significant differences between PDQ-derived total dietary risk scores upon two administrations and on comparison with scores derived from a FFQ, the differences were small (1.5 and −1.6 respectively, out of 100 points). Bland–Altman plots revealed arguably wide 95% LOA but no systematic bias between the two tools, indicating good group-level agreement. The high proportion of participants classified into the same risk category highlights the usefulness of the PDQ as a screening instrument to identify those at highest risk requiring intervention. However, the lack of association between PDQ-derived dietary risk scores and preschoolers’ BMIz score, consistent with findings from Accepted January 2018

Exploring Grandparents' Roles in Young Children's Lifestyle Behaviors and the Prevention of Childhood Obesity: an Australian Perspective

Childhood obesity remains a significant public health issue. Because lifestyle behaviors and weight are established early and track through life stages, prevention strategies must commence in the first years of life. Traditionally, such strategies target parents or formal child care providers. Yet grandparents are increasingly providing care to grandchildren and therefore have an important role in their eating and activity behaviors, which creates a major research gap. This commentary piece, focusing on the Australian context, argues that it is imperative and timely for obesity prevention research to include investigations regarding the role of grandparents in the prevention of obesity-related behaviors in young children.

Poor dietary patterns at 1–5 years of age are related to food neophobia and breastfeeding duration but not age of introduction to solids in a relatively advantaged sample

Previous studies have investigated associations between individual foods or food group intake, and breastfeeding duration, age of solid introduction and food neophobia. This study aimed to investigate associations between whole dietary patterns in young children, and breastfeeding duration, age of solid introduction and food neophobia. Parents of children (N = 234) aged 1-5 years completed an online questionnaire. Dietary risk scores were calculated using the Toddler (1-3 years) or Preschool (>3-<5 years) Dietary Questionnaires which evaluates the previous weeks food-group intake (scored 0-100; higher score = higher risk of poor dietary quality). Neophobia was measured using the Child Food Neophobia scale (1.0-4.0; higher score = more neophobic). Associations were investigated using multivariable linear regression, adjusting for covariates. Children (54% female, 3.0 ± 1.4 years) were from advantaged families and were breastfed until 11.8 (5.0-16.0) months, started solids at 5.6 ± 1.4 months of age, moderately neophobic (2.1 ± 0.7) and at moderate dietary risk (29.2 ± 9.2). Shorter breastfeeding duration (β = -0.21; p = 0.001) and poorer child food neophobia scores (β = 0.36; p < 0.001) were associated with higher dietary risk scores. Age of introduction to solids showed no association with dietary risk (p = 0.744). These findings suggest that in addition to breastfeeding promotion, supporting parents to manage neophobic behaviour may be important in promoting healthy eating patterns in early childhood.