Maternal depression and child feeding practices in Malaysia: A driver of malnutrition?

Based on emerging evidence from other settings, the authors hypothesise that there is a relationship between maternal depression and child feeding practices in malnourished and well-nourished children. To test this, a case-control study was conducted in eight health clinics in Kuala Langat, peninsular Malaysia.

”œSome research has connected maternal depression to potentially harmful feeding behaviours, such as being less likely to establish boundaries or restrict child intake.”

Participants (≥18 years) were sampled from existing district health clinic lists, with 62 mothers of malnourished children selected at random (case) and 62 well-nourished children (control) being assigned. Child age, gender, and residential area were matched between case and control. The case group was defined as any child with weight-for-age (underweight), height-for-age (stunted), and/or body mass index-for-age (wasted) -2SD below World Health Organization growth chart means. Controls were >-2SD to <+2SD for each of these indices. Children were aged 6–59 months and mothers had no co-morbidities. A translated and locally validated questionnaire was used to determine maternal depression status in mothers.

Adjusted odds ratios indicate that children from depressed mothers had more than twice the odds of malnutrition (AOR 2.5, p=0.03). However, the 95% confidence interval was broad (1.08–6.09), so we should also consider other variables to interpret this study.

The authors used the Krejcie and Morgan (1970) table to determine their sample size, a well-used and validated approach. That said, only a 10% dropout rate was accounted for and there was a 17.7% dropout in this study – possibly leaving it underpowered. Body mass index-for-age was also a criterion, which is not considered to be wholly accurate for children aged under 24 months. In this study, children were aged 6–59 months.

Comorbidities, including diagnosed depression, were excluded in this study. By expanding the criteria to include mothers with diagnosed depression as cases, the study may have yielded a greater effect size, increasing confidence in the findings. Although malnutrition was determined by growth charts, this study also measured mid-upper-arm circumference (MUAC). As expected, there is a significant difference (p<0.001) between malnourished cases and well-nourished controls (14.0cm ±1.4 and 15.3cm ±1.4); however, these average MUAC values are higher than standard wasting cut-offs¹. This may indicate that both groups were closer together in ‘malnourished’ status, especially wasting, which may well reduce the effect size observed.

The link between maternal mental health and child nutrition status remains underexplored, so this study addresses a valuable question. Unfortunately, the broad confidence interval coupled with methodological shortfalls make it difficult to place too much confidence in these results. The central hypothesis is indeed logical, but this study alone does not provide conclusive evidence of an association and will need to be complemented by other studies, ideally ones that address causality in their design.

References

Krejcie R & Morgan D (1970) Determining sample size for research activities. Educational and Psychological Measurement, 30, 607–610.