Risk Factors for Childhood Stunting in 137 Developing CountriesA Comparative Risk Assessment Analysis at Global, Regional, and Country Levels

A two-year-old child in a rural village is not visibly sick and not acutely starving — just small. She is shorter than she should be, in a way that won't reverse. One in three children under five in the developing world shares that fate. For decades, the field recognized stunting as a massive problem but lacked a clear, ranked answer to the most basic policy question: which causes, in which countries, are doing the most damage? Danaei and colleagues set out to create that accounting. Stunting is defined as a height-for-age z-score more than two standard deviations below the global median based on the World Health Organization's two thousand six Child Growth Standards. It is not a temporary setback. The damage to physical development, cognitive capacity, and long-term economic potential is largely irreversible. In two thousand eleven, the study estimated that 44.1 million two-year-olds across 137 developing countries were stunted, accounting for 36 percent of all two-year-olds in those countries. Furthermore, fourteen percent of childhood deaths are attributable to stunting. The focus on ages 24 to 35 months is deliberate. That timeframe marks the end of the "first 1,000 days" — the period from conception to a child's second birthday — when growth is most sensitive to environmental and nutritional insults. Studying stunting during this period captures the cumulative damage of everything that happened in the womb and during the first two years of life. The question is: what drove that damage, and how much did each factor contribute? To answer this question, Danaei and colleagues used a comparative risk assessment across all 137 countries, targeting 18 modifiable risk factors grouped into five clusters: maternal nutrition and infection; teenage motherhood and short birth intervals; fetal growth restriction and preterm birth; child nutrition and infection; and environmental factors, which included unimproved water, unimproved sanitation, and use of biomass fuel. Country-level exposure was estimated for around two thousand ten using population-based surveys, and effect sizes were obtained from the most recent meta-analyses or, where none existed, from reviews the team conducted themselves. The core metric is the population attributable fraction, or PAF — the share of stunting cases that would disappear if everyone were exposed to the optimal level of a given risk factor. Conceptually, you multiply the prevalence in each exposure category by that category's excess relative risk, sum those products across all categories, and divide by the total. It's a counterfactual engine: what fraction of today's stunting burden is traceable to this one cause? Where a risk factor worked through an intermediate, the team adjusted the calculations accordingly. Maternal underweight, anemia, malaria, and biomass fuel all had published effects on low birth weight rather than directly on stunting, so the authors multiplied through: the PAF of low birth weight attributable to the risk times the PAF of stunting attributable to low birth weight. Breastfeeding practices worked through diarrhea. Zinc deficiency was partly mediated by diarrhea, and lacking direct estimates, the team assumed that fifty percent of zinc's excess risk ran through that pathway, while sensitivity analyses tested zero and one hundred percent. To avoid double-counting when risks overlapped within a cluster, individual PAFs were combined using a multiplicative formula. Uncertainty was propagated through one thousand simulations, with ninety-five percent confidence intervals drawn from the 2.5th and 97.5th percentiles. The headline result is that fetal growth restriction, defined as being born at term but small for gestational age, is the single largest contributor to childhood stunting globally. The study attributes 10.8 million stunting cases to it — out of 44.1 million total — with a ninety-five percent confidence interval of 9.1 to 12.6 million. Sit with that number. The logic is direct: when a child's growth is already constrained in the womb, that prenatal shortfall carries forward. Danaei and colleagues stated it plainly — prenatal restricted growth is logically strongly related to postnatal restricted growth. You don't easily catch up from a start like that. The second largest individual risk is unimproved sanitation, at 7.2 million attributable cases, with a confidence interval of 6.3 to 8.2 million. That number reframes stunting entirely. This is not only a story about food and feeding; it is a story about toilets, water pipes, and the chronic low-grade infection that follows children who live in environments saturated with fecal contamination. That infection diverts calories from growth, disrupts gut function, and compounds over months and years. Third is childhood diarrhea, with 5.8 million cases — a wide confidence interval of 2.4 to 9.2 million, reflecting genuine uncertainty in effect sizes, but still a burden that is enormous. Taken together as clusters, fetal growth restriction and preterm birth accounted for 32.5 percent of stunting globally — about 14.4 million cases, with a confidence interval of 12.6 to 16.2 million. Environmental factors ranked second at 21.7 percent, accounting for 9.6 million cases. Maternal nutrition and infection contributed 14.4 percent, child nutrition and infection accounted for 13.5 percent, and teenage motherhood and short birth intervals made up just 1.9 percent. However, those global averages conceal striking regional variation. The fetal growth restriction and preterm cluster led in every single region — with no exceptions. But the second-ranked cluster shifted depending on the region. In South Asia, sub-Saharan Africa, and East Asia and the Pacific, environmental factors ranked second. South Asia's numbers are sobering: the fetal growth and preterm cluster accounted for 40.9 percent of stunting there, and environmental factors added 24.5 percent. In sub-Saharan Africa, the split was 30.6 and 27.0 percent respectively — meaning that in parts of Africa, poor sanitation and contaminated water are nearly as damaging as restricted fetal growth. In Latin America and the Caribbean, Central Asia, and the Middle East and North Africa, child nutrition and infection took the second spot instead — 18.1 percent in Latin America, compared to just 8.9 percent for environmental factors there. The infrastructure gap that dominates South Asia and sub-Saharan Africa is less acute in those regions, shifting the burden toward postnatal feeding and disease. The within-region variation is equally striking. In sub-Saharan Africa, the burden attributable to unimproved sanitation in southern Africa was less than half that in central, east, and west Africa. Within Latin America, diarrhea carried almost three times the stunting burden in Andean and central countries as it did in tropical and southern regions. In East Asia, diarrhea accounted for just 1.9 percentage points of attributable stunting prevalence, versus more than 3.5 points in Central, South, and Southeast Asia. The country level gets even more specific. In China and many sub-Saharan African countries, unimproved sanitation actually surpassed fetal growth restriction as the leading individual risk. Bangladesh ranked first globally for teenage motherhood as a stunting driver. Malawi ranked first for preterm but appropriate-for-gestational-age births. Somalia had the largest prevalence attributable to discontinued breastfeeding. That granularity is the point. A global average buries the local signal. A policymaker in Bangladesh faces a different ranked list than one in Malawi or China, and this analysis — for the first time across 137 countries — provides them that list. The policy implications flow directly from the rankings. If fetal growth restriction is the largest single cause, then the most effective interventions should start before or during pregnancy — such as maternal nutrition supplementation, infection control, and antenatal care that reaches women in the first trimester rather than the third. Danaei and colleagues specifically cite maternal iron supplementation, balanced protein-energy supplementation, and multiple micronutrient supplementation as the most evidence-backed tools for reducing fetal growth restriction. Reaching women early enough remains a logistical challenge in many settings, but the data supports the case for trying harder. If unimproved sanitation is the second largest cause globally, and the environmental cluster as a whole accounts for more than one in five stunting cases, then reducing stunting requires investment in water and sanitation infrastructure alongside nutrition programs. These are not parallel tracks — they converge on the same child. The authors are candid about limitations. Most effect sizes come from observational studies, so causality is inferred rather than proven. Several risk factors — such as maternal smoking, prenatal alcohol, lead, and arsenic exposure — lacked sufficient country-level data and were excluded, meaning the true attributable burden is likely higher than what was measured. Some exposures were approximated using proxies. The population attributable fraction framework assumes the effect sizes from published studies generalize proportionally across very different country contexts. What this analysis delivers, despite those limits, is a ranked, country-specific map of where stunting originates. It is not a monolith to be tackled with a single intervention, but rather a problem with different leading causes in different places — causes that can now be prioritized. The child who is small at two years old was shaped by what happened in the womb, by the water her family used, and by the infections that found her before her immune system was ready. Identifying those causes precisely is the first step toward preventing them. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.