Cluster-randomized trial on complementary and responsive feeding education
Matern Child Nutr. 2012 May 24. doi: 10.1111/j.1740-8709.2012.00413.x.
Cluster-randomized trial on complementary and responsive feeding education to caregivers found improved dietary intake, growth and development among rural Indian toddlers.
Vazir S, Engle P, Balakrishna N, Griffiths PL, Johnson SL, Creed-Kanashiro H, Fernandez Rao S, Shroff MR, Bentley ME.
- Source – Summary written by Alive&Thrive to highlight key infant and young child feeding findings as well as program implications.
Background: In India around half of all children under 5 years of age are malnourished. Inappropriate infant feeding and care practices contribute to this problem. In food secure areas there is evidence that interventions that educate mothers/caregivers on complementary and responsive feeding can improve the nutrition, growth and development of infants and toddlers.
Objectives: The study aimed to determine whether an educational intervention focusing on responsive feeding and mother-child interaction in combination with appropriate breastfeeding and complementary feeding from 3 to 15 months of age could improve infant dietary intake, hemoglobin, growth and development.
Methods: A cluster-randomized study of 600 mother-infant pairs in rural India compared outcomes between 3 groups: 1) a control group (CG) that received standard care: routine services provided by Integrated Child Development Services (ICDS), the major national program in India that provides supplementary nutrition and health care; 2) a complementary feeding group (CFG) that received ICDS services and 11 nutrition education messages on sustained breastfeeding and complementary feeding; and 3) a responsive complementary feeding and play group (RCF&PG) that received ICDS services, the 11 nutrition education messages, 8 messages and skills on responsive feeding, 8 developmental stimulation messages, and 5 toys with instructions on how to use them to engage and play with their children. Intervention group messages were delivered through home visits conducted twice or four times a month over 12 months by trained village women.
Results: At 9 months and 15 months, intervention children were more likely to be fed rice, goat/chicken liver, banana, egg, spinach, and added fat than CG children. However, the median number of days per week that children were fed animal-source foods (other than buffalo milk) and greens was very low in all groups.
The intervention groups had higher intakes of micronutrients than the CG, but their intakes were still well below the Recommended Dietary Intakes (RDIs) for all micronutrients except calcium at 9 months. Energy and protein intakes met recommendations across all study groups.
Both intervention groups showed improvements in hemoglobin from 3 months to 15 months compared to the CG. Final values at 15 months were 9.3 g/dL (CFG + RCF&PG) vs. 9.0 g/dL (CG).
The group that received only the complementary feeding messages had slightly higher average length than the control group, while the group that received both feeding and development messages did not differ from the control group. At 15 months, the prevalence of stunting was 28% in the CFG, 36% in the RCF&PG and 37% in the CG. Weight gain did not differ between groups.
There were no differences in average length or weight between groups when infants were 3, 6, 9, 12, or 15 months of age. After adjusting for initial length, maternal characteristics, and child age, children in the CFG had greater length gain than those in the CG, but there was no difference between the RCF&PG and CG. Weight gain did not differ between groups. At 15 months, the prevalence of stunting was 28% in the CFG, 36% in the RCF&PG and 37% in the CG.
At 15 months, children in the RCF&PG had higher mental development scores than children in the CG, while CFG children did not (mean Bayley Mental Development Scores: RCF&PG: 107, CFG: 106, CG: 104); there were no differences in motor development scores among the 3 groups.
Overall, there was no significant benefit of the RCF&P intervention compared to the CF intervention. The RCF&P intervention provided a large number of recommendations to mothers (27 vs. 11 messages for the CF intervention), possibly making it difficult for mothers to put them all into practice due to time and resource constraints.
The researchers suggested that some of the messages may have been unnecessary. For example, their data indicated that maternal knowledge and beliefs about responsive feeding showed little variability and were already very positive at baseline for 5 out of 8 items in all 3 groups, and were not affected by the intervention. This suggests that some of these messages could have been omitted without compromising the intervention.
The reasons why the RCF&P intervention showed effects on mental development but not on child growth deserve further exploration. One possibility is that mothers found it easier to respond to messages on child stimulation than child feeding because the desired behaviors for the former were more “actionable” and less constrained by resources than some of the child feeding messages. Resources to increase child consumption of animal-source foods may be particularly limiting as suggested by the finding that infants rarely consumed such foods (other than buffalo milk) even after the intervention.
Program implications: Educational interventions can improve infant and young child feeding as well as child development. More operational research is needed to demonstrate how to select and integrate key messages in ways that best promote behavior change.