Effects of cooking with liquefied petroleum gas versus biomass on hemoglobin concentrations in pregnant women: a pre-specified exploratory analysis of the HAPIN trial
- Sheela S. Sinharoy
- Wenlu Ye
- Ajay Pillarisetti
- Sant-Rayn Pasricha
- Lisa M. Thompson
- Anaite Diaz-Artiga
- Usha Ramakrishnan
- Ghislaine Rosa
- Maggie L. Clark
- Dana Boyd Barr
- Vigneswari Aravindalochanan
- Kyle Steenland
- Shirin Jabbarzadeh
- Lindsay J. Underhill
- Miles A. Kirby
- Amy E. Lovvorn
- William Checkley
- Jennifer L. Peel
- Thomas F. Clasen
- Kirk R. Smith
- Eduardo Canuz
- Adly Castañaza
- Carmen Lucia Contreras
- Oscar De León
- Irma Sayury Pineda Fuentes
- Mayari Hengstermann
- John P. McCracken
- Erick Mollinedo
- Libny Monroy
- Alexander Ramirez
- Elisa Puzzolo
- Sarah Rajkumar
- Bonnie N. Young
- Priya D’Souza
- Savannah Gupton
- Ian Hennessee
- Grace Lee
- Jiawen Liao
- Julia N. McPeek
- Parinya Panuwet
- P. Barry Ryan
- Kalpana Balakrishnan
- Sarada Satyamoorthy Garg
- Krishnendu Mukhopadhyay
- Durairaj Natesan
- Naveen Puttaswamy
- Karthikeyan Dharmapuri Rajamani
- Rengaraj Ramasami
- Sudhakar Saidam
- Sankar Sambandam
2026-06-10
Evidence linking household air pollution exposure and blood hemoglobin concentration is lacking. We examine the effect of a liquefied petroleum gas cookstove and fuel intervention on hemoglobin concentration, along with associations between household air pollution exposures and hemoglobin concentration, among pregnant women. We enroll 800 pregnant women each in Guatemala, Peru, India, and Rwanda in an open-label randomized controlled trial (NCT02944682). In 3178 women (intervention=1585; control=1593), we measure hemoglobin concentration and 24-hour personal exposure to particulate matter with an aerodynamic diameter ≤2.5μm (PM 2.5 ), black carbon (BC), and carbon monoxide (CO) at three timepoints (9-20, 24-28, and 32-36 weeks gestation). We evaluate the effects of the intervention on hemoglobin concentration and conduct exposure-response analyses to examine associations between 24-hour personal exposure to measured pollutants and hemoglobin concentration. We identify a significant increase in hemoglobin in the intervention group (0.074 g/dL, 95% CI: 0.002, 0.145) compared to the control group. In exposure-response analyses, each 1ppm increase in CO exposure is associated with a 0.015 g/dL (95% CI: 0.008, 0.023) increase in hemoglobin. In our analyses, neither PM 2.5 nor BC are associated with hemoglobin concentration. Further research may be needed to examine the biological mechanisms underlying our findings.