Impact of Improved Biomass and Liquid Petroleum Gas Stoves on Birth Outcomes in Rural Nepal: Results of 2 Randomized Trials
Joanne Katz, James M. Tielsch, Subarna K. Khatry, Laxman Shrestha, Patrick Breysse, Scott L. Zeger, Naoko Kozuki, William Checkley, Steven C. LeClerq, Luke C. Mullany
Global Health: Science and Practice · 2020 · DOI: 10.9745/GHSP-D-20-00011
Countries: Nepal
What Was Studied
Two sequential randomized trials were conducted among rural households in Sarlahi District, in the low-lying Terai plains of southern Nepal bordering Bihar, India — an area with near-universal use of traditional open-burning biomass stoves. Trial 1 (March 2010-August 2012) was a cluster-randomized step-wedge trial in which 51 clusters of households sequentially received a vented 2-burner improved biomass stove (Envirofit G-3300) in place of traditional stoves, with kitchen PM2.5 and carbon monoxide measured before and after installation; 2,379 live-born infants were enrolled. Trial 2 (March 2013-March 2014) was a household-randomized parallel trial comparing continued use of the vented biomass stove (270 live births) against a 2-burner LPG stove with a free year-long gas supply (279 live births). Outcomes were mean birth weight, gestational age, and rates of low birth weight (<2500g), preterm birth (<37 weeks), and small-for-gestational-age (SGA) births, measured by trained local teams within 72 hours of birth.
What They Found
Both trials found no significant protective effect on birth outcomes despite real reductions in air pollution. In Trial 1, the improved biomass stove cut mean kitchen PM2.5 from 1,380 µg/m³ to 936 µg/m³ (and CO from 11.0 to 6.7 ppm), but this was still more than 37 times the WHO indoor air guideline (25 µg/m³). Comparing infants with no in-utero exposure to the improved stove versus full-pregnancy exposure, the authors report only a non-significant 16% reduction in low birth weight and a non-significant 23g increase in mean birth weight, with no consistent dose-response pattern across intermediate exposure levels. In Trial 2, switching households to LPG (with a free year's gas supply) reduced PM2.5 further, to 442 µg/m³ versus 885 µg/m³ for vented biomass (still 18 times the WHO standard), and CO to 1.7 versus 5.5 ppm — yet low birth weight was numerically higher with LPG (31%) than vented biomass (23%), preterm birth was identical (42% vs 42%), and small-for-gestational-age was similar (17% vs 13%), with none of these differences reaching statistical significance. The authors concluded plainly that neither intervention reduced adverse birth outcomes, attributing the null finding to persistently high exposure driven by incomplete adherence (about 50% of monitored Trial 2 households reported using a non-study stove in a given week, versus ~10% in Trial 1) and high ambient outdoor air pollution.
What This Means for Nepal
This trial is Nepal's own evidence base for the "clean cooking transition" now listed among the Ministry of Health's 2026 reform priorities, and it delivers a sobering, policy-relevant null result: roughly 54% of Nepali households still cook with solid biomass, and household air pollution is a major driver of the country's ~15% low-birth-weight rate and the COPD burden that together help make air pollution Nepal's #1 mortality risk factor. Because even the LPG arm left kitchen PM2.5 at 18 times the WHO guideline — due to continued "stove stacking" (parallel use of the old open-burning stove) and high ambient outdoor pollution — the key lesson for the Ministry of Health (MoHP) and its "35 by 35" PM2.5 target is that distributing cleaner stoves or subsidising LPG cylinders is not sufficient on its own to protect birth outcomes. Concrete recommendation: any future MoHP or donor-funded clean-cooking programme (including LPG subsidy schemes) should budget for household exposure monitoring and active stove-stacking mitigation (e.g., verified retirement of the old stove), rather than assuming fuel/stove distribution alone will translate into measurable maternal and newborn health gains — exposure levels achieved, not just the stove or fuel type distributed, are what determine health impact.
Contextualisation
Nepal's most recent World Bank assessment identifies air pollution as the country's #1 mortality risk factor, and roughly 54% of Nepali households still cook with solid biomass fuel (versus ~44% using LPG) — making this Sarlahi District trial a direct empirical test of Nepal's own clean-cooking transition strategy. The finding that neither a chimney-vented improved biomass stove nor a full year of free LPG significantly reduced low birth weight, preterm birth, or small-for-gestational-age births — because kitchen PM2.5 remained 18 to 37 times above the WHO guideline even after the intervention — is a cautionary, Nepal-specific data point for the Ministry of Health's "35 by 35" PM2.5 target and for any programme that distributes cleaner stoves without addressing continued biomass "stove stacking."