Purpose: Despite evidence that childhood obesity is influenced by early life events and environmental exposures, the contribution of diet in infancy to the development of overweight/obesity remains relatively unexplored. Most studies investigating the effect of infant feeding on overweight/obesity focus on discrete practices (e.g. breastfeeding) and do not consider dietary quality more broadly. We examined associations between adherence to the New Zealand (NZ) national infant feeding guidelines and development of child overweight/obesity and central obesity at age 54-months.Methods: Data were collected on 6,435 infants born in 2009-10, enrolled during pregnancy into the Growing Up in NZ cohort study. Adherence to the guidelines was measured by an Infant Feeding Index (IFI) which was created based on the Ministry of Health’s Food and Nutrition Guidelines (2008). The IFI score ranged from 0 (no adherence) to 12 (complete adherence). Anthropometric measurements and screen time were collected at the 54-month interview. Body-mass-index/age (BMI/A) was calculated and classified according to WHO 2006 growth standards (overweight/obesity: >+2z scores). Waist-to-height ratio (WtHR) >90^th percentile identified children with the highest values for central obesity. Maternal sociodemographic and health behaviour characteristics were described from the antenatal face-to-face interview. Adjusted Poisson regressions with robust variance (risk ratios-RR and 95% confidence intervals-95%CIs) were applied to examine the effect of the IFI scoring (in tertiles) on BMI/A and WtHR. Models were adjusted for maternal antenatal characteristics, screen time and exact age at the 54-month interview (p<0.05).Results: In the fully adjusted model and compared to children in the highest tertile of the IFI, girls who scored in the lowest and middle tertiles were more likely to have BMI/A>+2z at 54-months (RR=1.46, 95%CI 1.03-2.06 and RR=1.55, 95%CI 1.08-2.21 respectively), and boys who scored in the lowest tertile of the IFI were more likely to have WtHR >p90^th at 54-months (RR=1.59, 95%CI 1.06-2.39).Conclusions: In this NZ nationally-generalizable birth cohort, lower adherence to national infant feeding guidelines was associated with the development of overweight/obesity and central obesity in early childhood, with gender differences evident. Improvement in infant feeding practices is a potential strategy for prevention of childhood obesity.

Purpose: Very few interventions targeting obesity-related behaviours investigate effectiveness in different population groups and whether differences persist over time. After the intervention in the INFANT trial, intervention effects were observed for vegetable (positive) and sweet snack (negative) consumption in more educated mothers and the intervention was more effective in increasing vegetable and water intake in younger mothers. Thus, the aim of this study is to assess the effectiveness of the INFANT program according to maternal education and age, two-years post-intervention completion.

Methods: INFANT was a cluster-randomized controlled trial focusing on the prevention of obesity-related behaviours in first-time Australian parents with infants aged four months between 2008-2010 (n=542). All families completing the program (n=492, child age 20 months) were invited to participate in the follow-up when their child was 3.5 years old. Outcomes included are: BMI z-scores, physical activity (ActiGraph), television viewing (parent report), and dietary intake (3 unscheduled 24hr dietary recalls). The moderator variables maternal education and age were dichotomized (university degree vs. no university degree and <32 years vs. ≥32 years, respectively).

Results: Interaction effects were observed for both maternal education and age in the treatment group. Intervention effects were greater for vegetable consumption in infants whose mother had a university degree (mean difference = 29.8g/day [CI₉₅:11.3, 48.3]), and sweet snack consumption was lower (mean difference = -9.3g/day [CI₉₅:-15.0, -3.6]). Conversely, a higher consumption of water was observed among infants whose mothers did not have a university degree (mean difference = 179.4g/day [CI₉₅:41.0, 317.7]). In relation to maternal age, a lower sweet snack consumption was observed in mothers aged 32 years or older in the treatment group (mean difference = -9.9g/day [CI₉₅:-15.3, -4.5]). However, water consumption was greater in mothers younger than 32 years in the treatment group (mean difference = 177.4g/day [CI₉₅:79.8, 274.9]).

Conclusions: Interventions aiming to prevent obesity-related behaviours in young children may have varying effectiveness in different groups of mothers depending on age and education level, even when there has not been an effect for the overall sample. Identifying such differences within populations will allow us to create more effective and tailored interventions in the future.

Purpose: This study used focus group discussions (FGDs) and in-depth interviews (IDIs) with mothers of infants aged 0-24 months to: (i) describe the infant feeding practices of South African women living in Soweto; and (ii) understand from the mothers’ perspective what influences these feeding practices.

Methods: Semi-structured FGDs were conducted with 19 mothers.  The FGDs were stratified according to the age of the mother’s baby as follows: one each for mothers of 0-6-month olds, 7-14-month olds and 15-24-month olds respectively.  Four mothers from each FGD then attended an IDI.     

Results/findings: While mothers understood that breastfeeding was beneficial, babies were exclusively breastfed for a very short time.  The dietary diversity of weaning foods was low (mostly grain-based porridge) and the introduction of high-quality novel foods was limited as babies grew older.  In addition, ‘junk’ food was commonly given as a snack.  Infants were fed using a bottle or spoon and feeding mostly occurred separately to family meal times.  Mothers feeding practices were influenced by the belief that how and what babies eat is important for their health; with an unwillingness to eat being a sign of ill health.  It was therefore common for mothers to force feed their babies – potentially ignoring satiety cues.  In addition, mothers believed that solid food was necessary for babies before six months of age.  Family matriarchs (the participants own mothers and grandmothers) were highly influential to the feeding practices adopted by mothers; with their advice often contradicting that of health professionals.  As such, mothers chose to lie about how they were feeding their babies when at health facilities due to fears of being judged and shouted at.         

Conclusions: Our study shows that, in urban African settings such as Soweto, shifting the norms around mixed feeding and the early introduction of solid foods, as well as promoting more responsive feeding approaches, must play a central role in the establishment of healthy appetites and eating behaviours in the early years.  In addition, engaging family members – particularly grandmothers and great-grandmothers, as well as fathers where possible – is important in promoting more supportive household and community structures around infant feeding.

Purpose

Snacking is a major contributor to children’s daily energy intake. However, research on snacking patterns is mostly limited to self-report and has offered little information on the social context of snacking. Recently, 168 children aged 11-13 in the Wellington region recorded their world for four days, using wearable cameras that automatically took pictures every 7s and GPS devices, providing 1.3M images linked to 2.9M GPS points. Using this data, we aimed to assess the frequency, nature and social context of children’s snacking patterns. 

Methods

Two days’ of image data (Thursday and Saturday) were analysed for snacking episodes and their context (e.g setting, source and screen use). Snacking was defined as any individual food item or meal consumed in between the three main meals. Data were entered in Excel and foods categorised as core or non-core, according to a nutrient profiling model. Data were analysed to determine differences in snacking patterns by setting, ethnicity, gender and deprivation.

Results/ findings

Children consumed more than twice as many non-core snack foods as core snack foods. Confectionary, cookies/cakes and sugary drinks and juices were the most commonly consumed non-core snacks. Fresh fruit was the most commonly consumed core snack. School cafeterias and convenience stores near schools were a significant source of non-core snack foods and drinks. Snacking was commonly associated with screen time.

Conclusions

This study provides objective evidence on children’s snacking patterns and social context in children’s everyday lives. The extent of non-core snacks sourced from school cafeterias and convenience stores near schools is concerning, raising support for healthy school food programmes and broader interventions targeting the obesogenic environment in which children live. Given the attenuated obesity epidemic, it is paramount that healthy snack options are made available to children. 

Introduction: Malaysia is experiencing a nutrition transition because of positive economic growth and rapid urbanisation. Therefore, public health attention needs to be drawn towards both undernutrition and overnutrition.

Purpose: We examined secular trends and sociodemographic determinants of underweight (UW) and overweight including obesity (OW) in 6-17-year-old Malaysians from 2006 to 2015.

Methods: We analysed data from the National Health and Morbidity survey conducted in 2006, 2011 and 2015. These cross-sectional surveys involved objective measures of height and weight. We used the International Obesity Task Force BMI criteria to define UW and OW. We performed trend analysis using log-binomial regression separately by sex and age groups, and stratified by ethnicity, residence area, household (HH) size and HH income, taking account the complex survey design. Individuals were stratified into three age groups: pre- (6-9 years), early (10-13 years) and late (14-17 years) adolescence (total n=28,094).

Results: Overall, the UW prevalence decreased among Malaysian youths between 2006 and 2015, while the OW prevalence increased. The prevalence of UW decreased from 22% (95%CI 21.0, 23.1) to 18% (95%CI 16.2, 20.3) in boys, and 23% (95%CI 22.2, 24.2) to 19% (95%CI 17.4, 21.0) in girls. Significant decreasing trends in UW prevalence were observed in all sex-age groups, except preadolescent boys. During the same time period, prevalence of OW increased from 20% (95%CI 19.2, 21.3) to 26% (95%CI 23.8, 28.5) in boys, and 19% (95%CI 17.9, 19.9) to 24% (95%CI 22.4, 26.5) in girls. The increase in OW prevalence over time was significant in all sex-age groups, except late adolescent boys. Prevalence of UW and OW varied across sociodemographic subgroups with differences observed in few subgroups over time.

Conclusions: Malaysia is facing a double burden of malnutrition with increasing trends of OW and persisting UW among 6-17 years old. Further research is needed to explore if  double burden also exists at other levels than the country;  such as state, household and individual levels.

Purpose: To examine parent meal-related behaviors and perceptions and their associations with home-served evening meals where half the plate is fruits and vegetables (1/2 plate FV) or FV intake among school-age children in a rural United States community. Past research has found protective associations between family meal frequency, parent meal-related behaviors and perceptions, and child FV intake; however, this research focused primarily on suburban or urban areas, leaving a gap in understanding for rural communities.

Methods: Baseline data (collected 2017/2018) were drawn from the New Ulm at Home (NU-HOME) randomized controlled trial, conducted in a rural community by academic researchers in collaboration with community leaders and affiliated health systems. The trial recruited 114 children ages 7-10 [59% girls; M=8.9, SD=1.0 years] and parents [97% women; M=37.9, SD=5.4 years]. Parents and children completed surveys; children were assisted by research staff. Research staff completed 2 dietary recalls with parents to measure children’s dietary intake. Logistic regression and general linear models (GLM) analyzed associations of family meal frequency and parent mealtime behaviors (meal planning) and perceptions (cooking self-efficacy, parent/child cooking skills) with ½ plate FV, and parent role modeling with child FV intake, respectively.

Results/findings: On average, parents reported eating evening meals with all/most family members 5 times per week and children were served ½ plate FV at the evening meal 3 times per week. Child FV intake averaged 2.36 servings/day; only 10.5% met the recommendation of at least 4 FV servings/day. Unadjusted multivariable logistic regression results indicated children had significantly higher odds of having ½ plate FV with higher family meal frequency (OR=1.39, CI=1.07-1.79) and parent meal planning (OR=1.25, CI=1.02-1.53). In unadjusted GLM, higher child FV intake was significantly associated with frequent (vs. infrequent) parent role modeling of vegetable consumption at evening meals (2.67 vs. 2.12 servings; b=0.55, SE=0.25, p=.027).

Conclusions: In this rural community, parents play significant roles in shaping FV intake of school-age children, whether by planning and serving regular family meals incorporating FV or by role-modeling healthful dietary intake habits at home. However, more work needs to be done to bring FV intake to recommended levels.

Purpose: WHO defines adolescence as age 10-19 years. Physical inactivity during childhood and adolescence increases the risk of obesity, and both are associated with non-communicable diseases. Parents’ socioeconomic attributes may influence their children’s physical activity (PA). We assessed PA among adolescents and examined associations with parental education and socioeconomic status (SES).

Methods: Children born to women, who participated in the Mumbai Maternal Nutrition Project, living in the slums of Mumbai, were recruited. Antropometry and percentage body fat (fat%) were measured and parental SES was documented. In a subgroup, PA was measured over 7 days, using accelerometers (MTI Actigraph). Actigraph data gave objective measures of PA counts and time spent in sedentary, moderate or vigorous intensity activities (calculated using Evenson cut points). Associations of parental exposures with PA were analysed using correlation and multiple linear regression.

Results: 144 children were recruited (mean (SD) age: 11.3(0.4) years; 77 boys) and 79 children (40 boys) used accelerometers. Median (IQR) BMI of the children was 15.5 (14.3, 17.7) kg/m²; 83% were underweight and 3% overweight. Majority of parents (>70%) had completed secondary education. During school vacations, boys spent more time than girls in vigorous activities (165.4 v 81.1 minutes/day; p<0.001). Girls had higher fat% compared to boys (16.0% v 10.2%, p=0.03). Weight (r=0.27, p=0.001), BMI (r=0.26, p=0.004) and fat% (r=0.24, p=0.003) were positively correlated with SES. Fathers’ education was positively associated with accelerometer counts/minute (B=65.9, p=0.049), and negatively with time spent in sedentary activities (B= -33.6, p=0.005). Total family income was positively associated with time spent in vigorous activities (B=0.002, p=0.03, including age, sex, income, SES, mother and fathers’ education as covariates). There was no association of maternal education with PA.

Conclusion: Girls are less active and have a higher fat% than boys. The fathers’ education, income and socioeconomic status were associated with adolescent’s time spent on sedentary and vigorous activities. As girls are less active than boys, interventions aimed at girls, involving parents, are needed. The sample size of this study is small and results would need replication in larger sample size.