The Association Between Maternal Infant Feeding Practices and Child Weight at 11-17 Months in First-time Australian MothersJansen E1, Daniels L1*, Byrne R1, Mihrshahi S1, Magarey A1,2, Nicholson J 3,4

 1 Institute of Health and Biomedical Innovation, School of Public Health, Queensland University of Technology, Brisbane, Australia2 Nutrition and Dietetics, School of Medicine, Flinders University, Adelaide, Australia3 Queensland University of Technology, Brisbane, Australia4 Murdoch Children’s Research Institute, Melbourne, Australia* t:61 (0) 7 3138 6139 | e: l2.daniels@qut.edu.au

AimExamine the cross-sectional associations between maternal infant feeding practices/beliefs and child weight in Australian children aged 11-17 months, taking into account maternal and child-related characteristics

MethodsParticipantsThe sample for this study included 293 mother-child dyads from the control group enrolled in the NOURISH RCT (in 2008) [5]

Inclusion criteria Mothers:

• First-time mothers; facility with English• > 18 years• No self report of eating disorder or mental health problems

Children: • Healthy term (gestational age >35 weeks, birth weight

>2500g)• 11-17 months old

Outcome Measures Anthropometric data were measured by trained staff at 2-7 months and 11-17months of age The following variables were self-reported using a written questionnaire:

Feeding practices and beliefs• Based on Infant Feeding Questionnaire (IFQ) [6]• Factor structure reanalysed for NOURISH sample due to:

- Concurrent rather than retrospective use of IFQ - High prevalence of breastfeeding 3 items related to

bottle feeding excluded- Infant sample (Baughcum sample mean age 16.2±3.5

months)- Australian sample

New, validated 4-factor-solution used here

Data Analyses Multiple regression analysis

• Two models: one for each ‘weight’ outcome as the dependent variable 1. weight-for-age z-scores at 14 months 2. change in weight-for-age z-scores from 4 to 14 months

• Z-scores were calculated from WHO Child Growth Standards (2006 reference scores)

• Both models were adjusted for covariates (see Table 2)

Child weight (DV)

Variable R2

Feeding practices & beliefs (IV)

Factors β p-value

Weight-for-age z-score at 14 months

.283 Concern for underweight -.224 .000

Concern for overweight .197 .001

Awareness of cues .041 .468

Responsive infant-feeding -.104 .060

Change in weight-for-age z-score from

4 to 14 months

.128 Concern for underweight -.072 .241

Concern for overweight .104 .104

Awareness of cues .026 .675

Responsive infant-feeding -.147 .016

Note: Cases with missing data on any independent variable or covariate were excluded; adjusted R 2 and Standardized β’s for models including all covariates are reportedDV= dependent variables, IV= independent variables

Conclusions In the adjusted models, maternal feeding practices/beliefs explained more

variance of child weight at 14 months than weight gain until that age. Responsive infant-feeding was inversely related to both weight indicators,

providing evidence for the existence of an association at an early age. This is in line with speculative, mostly cross-sectional findings that

responsive feeding plays a role in accelerated growth and overweight; assumingly because inappropriate responses to child’s hunger and satiety cues impede the development of the child’s self-regulation of energy intake, leading to overeating and overweight eventually [4].

Encouraging responsive feeding (e.g. baby feeds whenever she wants, feeding to stop baby being unsettled) may be an important strategy to promote healthy child weight during the first 1.5 years of a child’s life.

Longitudinal studies are needed to determine direction of relationships.

Strengths and Limitations+ Concurrent use of revalidated Infant Feeding Questionnaire [6] + Adjustment for a range of maternal & child-related characteristics - Maternal feeding practices and beliefs were self-reported - Cross-sectional data analysis no information about direction of effects± Applied less frequently used measurement tool of child feeding as other

commonly used instruments to assess feeding practices have not been validated in very young children

References[1] Department of Health and Ageing. Canberra, 2008; [2] Ventura & Birch. Int J Behav Nutr Phys Act 2008;5:15; [3] Hurley et al. J Nutr 2011, 141(3):495-501; [4] DiSantis et al. Int J Obes 2011, 35:480-492; [5] Daniels et al. Bmc Public Health 2009;9:387; [6] Baughcum et al. J Dev Behav Pediatr 2001;22(6):391-408; [7] Killen et al. Int J Eat Disorder 1994;16(3):227-38

BackgroundBy 2-3 years of age approximately 15% of Australian children are overweight and a further 4% are obese [1]. Obesity prevention strategies need to start as early as possible and one potential strategy is promoting positive feeding practices. Early feeding practices can influence child food intake, eating behaviours and weight status [2]. Nonresponsive feeding has been implicated in the development of childhood obesity [3]. Practices such as only allowing an infant to feed at set times, have the potential to override a child’s internal cues of hunger and satiety. This may impair self-regulation of energy intake, leading to overeating and excess energy intake [4]. To date there are very few studies examining feeding practices in children under two years of age and how they relate to child weight status.

Factors and items composing each factor

Cronbach’sα

Factor 1Concern about infant undereating or becoming underweight Worry baby is not feeding enough (Q2) Struggle to get baby to feed (Q6) Baby needs encouragement to eat enough (Q13)

Worried baby will become underweight (Q16)

.79

Factor 2Concern about infant overeating or becoming overweight Worry baby is feeding too much (Q5) Upset if baby feeds too much (Q7) Worried baby will become overweight (Q19)

.70

Factor 3Awareness of infant’s hunger and satiety cues

I know when baby is hungry (Q15) I know when baby is full (Q17) Baby knows when she is hungry (Q18) Baby knows when she is full (Q20)

.76

Factor 4Responsive infant-feeding Allow baby to feed whenever she wants (Q1) Only allow baby to feed at set times (Q3)R

When upset, feeding baby is the first thing you do (Q4) Feeding is the best way to stop baby being unsettled (Q14)

.66

Note: Words in italic represent minor word changes* Two items were excluded because they loaded weakly onto a separate factorR indicates item is reverse coded for scoring. Higher scores on each factor indicate more concern about underweight, more awareness of cues, more concern about overweight, more responsive to feeding on cues

Table 1: New factor structure of IFQ [6] – 15 items*, 4 factors

Results continued Table 2: Characteristics of study sample (N=293)

a Difference between weight-for-age z-score at 4 months and weight-for-age z-score at 14 months of age b N=271 due to missing data on this variablec Mother’s concern about her own weight; Weight Concern Scale by Killen et al. (1994) [7]; range 0-5, higher scores indicate more concern; N=287 due to missing data on this variable d N=270 due to missing data on this variable

Variables Mean (SD) or Frequency

Child gender Girl 149 (51%)

Age at assessmentChild’s age (months)Mother’s age (years)

14 (1)30 (5)

Child weight Birth weight (in kg) Weight-for-age z-score (at 4m) Weight-for-age z-score (at 14m) Change from 4 to 14ma

3.5 (0.4)-0.02 (0.91)0.60 (0.85)0.62 (0.69)

Perception of child’s weight statusb

UnderweightNormal weightOverweight

27 (10%)232 (86%)12 (4%)

Pre-pregnancy weight status (self-reported) Underweight NormalOverweight

4 (1%)239 (82%)50 (17%)

Maternal weight concernc

2.5 (0.8)Education level

Non-University University

112 (38%)181 (62%)

Feeding at 4 monthsd

Fully breastfeedingMixed feeding

148 (55%)122 (45%)

Table 3: Adjusted regression models associating child weight with maternal feeding practices/beliefs (N=262)

Associations between child weight and feeding practices/beliefs, controlled for all child-related and maternal covariates from Table 2 are presented in the table below. While weight at 14 months was related to 3 out of the 4 feeding practices/beliefs (i.e. concern for under- and overweight, and responsive infant-feeding); change in weight from 4 to 14 months was significantly inversely related to responsive infant-feeding. Additionally, 5 covariates in total were independently related to weight and change in weight – birth weight, child’s age, mother’s perception of child’s weight status, and mother’s concern about her own weight.

ResultsCharacteristics of first-time mothers and their children are presented in Table 2.

AcknowledgementsNOURISH is funded by NHMRC ID 426704.S Mihrshahi’s post doctoral fellowship is funded by HJ Heinz Co.