In Peruvian shantytowns, an intervention delivered through health facilities to educate mothers on how to improve nutrition in their young children was associated with better dietary intake and a reduced likelihood of stunted growth early in life, according to results of a cluster-randomized controlled trial.¹ The proportions of mothers correctly answering questions on age-specific infant feeding practices and the proportions who reported following these practices were significantly higher in the intervention group than in the control group. Moreover, the proportions of infants born in the control population who had inadequate dietary intake of iron and zinc were significantly higher than the proportions among infants born in the intervention population. At 18 months of age, only 5% of children in the intervention group had stunted growth, compared with 16% of those in the control group.

The intervention was tested in Trujillo, a city in which stunting in childhood is common; although nearly all families have access to nutritious foods, the cultural preference is for foods with a low energy density. Twelve health facilities serving shantytowns were randomized to the intervention or the control condition. The intervention, implemented as a program of the regional health authority, aimed to integrate nutrition services into established child-oriented national programs, such as immunization, monitoring of growth and development, and management of acute respiratory infections and diarrhea. To improve the quality and provision of nutritional counseling, the program provided simple, standardized messages for intervention facility staff to deliver whenever they had contact with mothers of young children. The intervention also introduced a facility accreditation program to promote institutional compliance with the intervention.

The researchers enrolled a birth cohort of healthy infants born between August 1999 and February 2000 in shantytowns served by the facilities. Fieldworkers visited families shortly after their infant's birth and at eight intervals thereafter, until the child was 18 months old; during the visits, the workers collected social, demographic, dietary and health information; measured children's weight and length; and assessed mothers'knowledge and practices regarding feeding.

Analyses were based on 187 and 190 infants born in shantytowns served by intervention and control health facilities, respectively. The groups were generally well balanced with respect to infant characteristics. Roughly half of infants were male, and two in five were first-born children. The average birth weight was about 3.4 kg. Almost all mothers began breastfeeding their infants at birth.

The proportion of mothers reporting shortly after their infant's birth that they had received advice on nutrition from their health facility was significantly greater in the intervention group than in the control group (52% vs. 24%). In addition, when mothers were stratified by the age of their child, the proportions correctly answering questions about age-specific feeding practices were significantly greater in the intervention group. For example, among mothers whose child was eight months old, a greater proportion of those in the intervention group than of those in the control group could name the three most important supplementary foods for a 7-8-month-old infant—chicken liver, eggs and fish (64% vs. 48%).

Analyses of actual feeding practices at various ages likewise showed significant differences in favor of the intervention group. For example, the proportion who said they fed their infants nutrient-dense thick foods first at the main meal was greater in the intervention group than in the control group for children aged six months (31% vs. 20%), nine months (35% vs. 17%) and 12 months (42% vs. 26%). In addition, a higher proportion of children in the intervention group than in the control group received chicken liver, fish or egg at ages six months (65% vs. 51%) and eight months (61% vs. 49%).

In analyses of energy and nutrient intake, children in the intervention group were significantly less likely to fall short of recommended daily intakes from foods given as breast milk supplements at various ages. For example, smaller proportions of children in the intervention group than of those in the control group failed to meet dietary requirements for iron at eight months (91% vs. 96%), nine months of age (93% vs. 99%) and 18 months of age (37% vs. 48%); and dietary requirements for zinc at nine months of age (77% vs. 87%).

In analyses of growth, the proportions of children in the intervention and control groups who had stunted growth (a length-for-age more than two standard deviations below the median for the reference population) were similar for about the first six months of life, but diverged sharply thereafter. By the age of 18 months, only 5% of children in the intervention group had stunted growth, compared with 16% of children in the control group. After adjustment for other factors possibly related to stunting, children in the control group had significantly higher odds of stunted growth than did their counterparts in the intervention group (odds ratio, 3.0). Also at this age, on average, children in the intervention group were 0.7 cm taller and weighed about 200 g more than those in the control group after adjustment for potentially confounding factors.

"The results," the investigators assert, "add to evidence that nutrition education without the provision of food supplements can improve the dietary intake of young children and improve growth." However, they caution, the sustainability of the intervention and its generalizability to other populations (including those in which access to nutritious foods is limited) are yet to be determined. Noting that the difference in growth between groups was greater than expected given the differences in feeding practices and dietary intake, the investigators speculate that the early age at intervention may have been crucial.

The new study's findings are "encouraging," the author of an accompanying commentary notes, because previous studies of supplementary feeding interventions in childhood have often had poor designs or negative results. "The positive results of the intervention give hope to policymakers that this strategy will be effective elsewhere,"² she writes, but adds that wide coverage by health services of the population—particularly the most vulnerable groups—is required for the success of an intervention delivered through these services. —S. London

REFERENCES

1. Penny ME et al., Effectiveness of an educational intervention delivered through the health services to improve nutrition in young children: a cluster-randomised controlled trial, Lancet, 2005, 365(9474):1863-1872.

2. Dewey KG, Infant nutrition in developing countries: what works? commentary, Lancet, 2005, 365(9474): 1832-1834.