Drivers of stunting reduction in Peru: a country case study

Abstract

Background: Peru reduced its under-5 child stunting prevalence notably from 31.3% in 2000 to 13.1% in 2016.

Objectives: We aimed to study factors and key enablers of child stunting reduction in Peru from 2000-2016.

Methods: Demographic and Health Surveys were used to conduct descriptive analyses [height-for-age z scores (HAZ) means and distributions, equity analysis, predicted child growth curves through polynomial regressions] and advanced regression analyses. An ecological (at department level) multilevel regression analysis was conducted to identify the major predictors of stunting decline from 2000 to 2016, and Oaxaca-Blinder decomposition was conducted to identify the relative contribution of each factor to child HAZ change. A systematic literature review, policy and program analysis, and interviews with relevant stakeholders were conducted to understand key drivers of stunting decline in Peru.

Results: The distribution of HAZ scores showed a slight rightward shift from 2000 to 2007/2008, and a greater shift from 2007/2008 to 2016. Stunting reduction was higher in the lowest wealth quintile, in rural areas, and among children with the least educated mothers. Decomposing predicted changes showed that the most important factors were increased maternal BMI and maternal height, improved maternal and newborn health care, increased parental education, migration to urban areas, and reduced fertility. Key drivers included the advocacy role of civil society and political leadership around poverty and stunting reduction since the early 2000s. Key enablers included the economic growth and the consolidation of democracy since the early 2000s, and the acknowledgement that stunting reduction needs much more than food supplementation.

Conclusions: Peru reduced child stunting owing to improved socioeconomic determinants, sustained implementation of out-of-health-sector and within-health-sector changes, and implementation of health interventions. These efforts were driven through a multisectoral approach, strong civil society advocacy, and keen political leadership. Peru's experience offers useful lessons on how to tackle the problem of stunting under differing scenarios, with the participation of multiple sectors.

Keywords: Latin America; Peru; children; exemplar; linear growth; mixed methods; nutrition; stunting.

FIGURE 1

Map of Peru and regional under-5 stunting prevalence. (A) Map of Peru. (B) The prevalence of under-5 stunting in selected South American countries, 1990–2018. Source: Joint Malnutrition Estimates (4).

FIGURE 2

Conceptual framework showing distal, intermediate, and proximal determinants of stunting. Framework represents all variables that were identified and available for quantitative analysis.

FIGURE 3

HAZ distribution and growth faltering among children <5 in Peru (A) Kernel density plot for HAZ distribution in children <5 y from 2000 to 2016. (B) Victora curve using data from the 2000, 2007/2008, and 2016 surveys among children <5 y, including PAHO mean HAZ curve (103). (C) Victora curve using data from the 2000, 2007/2008, and 2016 surveys among children <5 y with linear splines. DHS, Demographic and Health Survey; HAZ, height-for-age z score; PAHO, Latin America and the Caribbean.

FIGURE 4

Under-5 stunting estimates by department in Peru. (A) Stunting estimates for children under-5 in Peru, 2000. (B) Stunting estimates for children under-5: compound annual growth rate from 2000 to 2016.

FIGURE 5

Stunting prevalence disaggregated by socioeconomic and geographic dimensions. (A) Stunting prevalence by wealth quintile, 2000–2016. (B) Stunting prevalence by maternal education, 2000–2016. (C) Stunting prevalence by residential area, 2000–2016. Estimates are based on original Peru Demographic and Health Survey rounds from 2000 to 2016.

FIGURE 6

Decomposing predicted changes in HAZ (i.e., percentage contribution of determinant domains) for children under-5 from 2000 to 2016. The <6 mo age category results are not presented owing to small sample sizes and minimal HAZ changes in this population causing unstable parameters. Indicators included maternal and newborn health care (skilled birth attendant and antenatal care ≥4), low birth weight, parental education (maternal and paternal education), mountainous population migration (altitude), maternal nutrition (maternal BMI and height), fertility (parity and interpregnancy interval), reduction in household crowding (number of household members), and conditional cash transfer program (Juntos). “Other” category includes child age, sex, and region for 6- to 23-mo and 24- to 59-mo age groups in addition to economic improvement (wealth index) (0.3%) and childhood vaccines (measles) (0.4%) for the 6- to 23-mo age group, and maternal age (older age pregnancy) (0.5%) and reduction in diarrhea incidence (0.6%) for children under-5. Parental education breakdown: for children 6–23 mo: maternal: 17.5%, paternal: 2.2%; children 24–59 mo: maternal: 13.1%, paternal: 1.9%; and children under-5: maternal: 17.8%, paternal: 2.7%. HAZ, height-for-age z score.

FIGURE 7

Overview of laws, policies, programs, and enablers between 1995 and 2016 in Peru. CAGR, compound annual growth rate; DHS, Demographic and Health Survey; UHC, Universal Health Coverage.