BackgroundThree billion people use solid cooking fuels, and 4 million people die from household air pollution annually. Shifting households to clean fuels, like liquefied petroleum gas (LPG), may protect health only if stoves are consistently used. Few studies have used an implementation science framework to systematically assess “de-implementation” of traditional stoves, and none have done so with pregnant women who are more likely to adopt new behaviors. We evaluated an introduced LPG stove coupled with a phased behavioral intervention to encourage exclusive gas stove use among pregnant women in rural Guatemala.MethodsWe enrolled 50 women at < 20 weeks gestation in this prospective cohort study. All women received a free 3-burner LPG stove and ten tank refills. We conducted formative research using COM-B Model and Theoretical Domains Framework (TDF). This included thematic analysis of focus group findings and classes delivered to 25 pregnant women (Phase 1). In Phase 2, we complemented classes with a home-based tailored behavioral intervention with a different group of 25 pregnant women. We mapped 35 TDF constructs onto survey questions. To evaluate stove use, we placed temperature sensors on wood and gas stoves and estimated fraction of stove use three times during pregnancy and twice during the first month after infant birth.ResultsClass attendance rates were above 92%. We discussed feasible ways to reduce HAP exposure, proper stove use, maintenance and safety. We addressed food preferences, ease of cooking and time savings through cooking demonstrations. In Phase 2, the COM-B framework revealed that other household members needed to be involved if the gas stove was to be consistently used. Social identity and empowerment were key in decisions about stove repairs and LPG tank refills. The seven intervention functions included training, education, persuasion, incentivization, modelling, enablement and environmental restructuring. Wood stove use dropped upon introduction of the gas stove from 6.4 h to 1.9 h.ConclusionsThis is the first study using the COM-B Model to develop a behavioral intervention that promotes household-level sustained use of LPG stoves. This study lays the groundwork for a future LPG stove intervention trial coupled with a behavioral change intervention.Trial registrationNCT02812914, registered 3 June 2016, retrospectively registered.

Background: Exposure to arising from solid fuel combustion is estimated to result in million premature deaths and 91 million lost disability-adjusted life years annually. Interventions attempting to mitigate this burden have had limited success in reducing exposures to levels thought to provide substantive health benefits. Objectives: This paper reports exposure reductions achieved by a liquified petroleum gas (LPG) stove and fuel intervention for pregnant mothers in the Household Air Pollution Intervention Network (HAPIN) randomized controlled trial. Methods: The HAPIN trial included 3,195 households primarily using biomass for cooking in Guatemala, India, Peru, and Rwanda. Twenty-four-hour exposures to , carbon monoxide (CO), and black carbon (BC) were measured for pregnant women once before randomization into control ( ) and LPG ( ) arms and twice thereafter (aligned with trimester). Changes in exposure were estimated by directly comparing exposures between intervention and control arms and by using linear mixed-effect models to estimate the impact of the intervention on exposure levels. Results: Median postrandomization exposures of particulate matter (PM) with aerodynamic diameter ( ) in the intervention arm were lower by 66% at the first (71.5 vs. ), and second follow-up visits (69.5 vs. ) compared to controls. BC exposures were lower in the intervention arm by 72% (9.7 vs. ) and 70% (9.6 vs. ) at the first and second follow-up visits, respectively, and carbon monoxide exposure was 82% lower at both visits (1.1 vs. ) in comparison with controls. Exposure reductions were consistent over time and were similar across research locations. Discussion: Postintervention exposures in the intervention arm were at the lower end of what has been reported for LPG and other clean fuel interventions, with 69% of samples falling below the World Health Organization Annual Interim Target 1 of . This study indicates that an LPG intervention can reduce exposures to levels at or below WHO targets. https://doi.org/10.1289/EHP10295

IntroductionIncreasing use of cleaner fuels, such as liquefied petroleum gas (LPG), and abandonment of solid fuels is key to reducing household air pollution and realising potential health improvements in low-income countries. However, achieving exclusive LPG use in households unaccustomed to this type of fuel, used in combination with a new stove technology, requires substantial behaviour change. We conducted theory-grounded formative research to identify contextual factors influencing cooking fuel choice to guide the development of behavioural strategies for the Household Air Pollution Intervention Network (HAPIN) trial. The HAPIN trial will assess the impact of exclusive LPG use on air pollution exposure and health of pregnant women, older adult women, and infants under 1 year of age in Guatemala, India, Peru, and Rwanda.MethodsUsing the Capability, Opportunity, Motivation–Behaviour (COM–B) framework and Behaviour Change Wheel (BCW) to guide formative research, we conducted in-depth interviews, focus group discussions, observations, key informant interviews and pilot studies to identify key influencers of cooking behaviours in the four countries. We used these findings to develop behavioural strategies likely to achieve exclusive LPG use in the HAPIN trial.ResultsWe identified nine potential influencers of exclusive LPG use, including perceived disadvantages of solid fuels, family preferences, cookware, traditional foods, non-food-related cooking, heating needs, LPG awareness, safety and cost and availability of fuel. Mapping formative findings onto the theoretical frameworks, behavioural strategies for achieving exclusive LPG use in each research site included free fuel deliveries, locally acceptable stoves and equipment, hands-on training and printed materials and videos emphasising relevant messages. In the HAPIN trial, we will monitor and reinforce exclusive LPG use through temperature data loggers, LPG fuel delivery tracking, in-home observations and behavioural reinforcement visits.ConclusionOur formative research and behavioural strategies can inform the development, implementation, monitoring and evaluation of theory-informed strategies to promote exclusive LPG use in future stove programmes and research studies.Trial registration numberNCT02944682, Pre-results.

Background Household air pollution is a major contributor to death and disability worldwide. Over 95% of rural Guatemalan households use woodstoves for cooking or heating. Woodsmoke contains carcinogenic or fetotoxic polycyclic aromatic hydrocarbons (PAHs) and volatile organic compounds (VOCs). Increased PAHs and VOCs have been shown to increase levels of oxidative stress. Objective We examined PAH and VOC exposures among recently pregnant rural Guatemalan women exposed to woodsmoke and compared exposures to levels seen occupationally or among smokers. Methods Urine was collected from 23 women who were 3 months post-partum 3 times over 72-hours: morning (fasting), after lunch, and following dinner or use of wood-fired traditional sauna baths (samples=68). Creatinine-adjusted urinary concentrations of metabolites of 4 PAHs and 8 VOCs were analyzed by liquid chromatography—mass spectrometry. Creatinine-adjusted urinary biomarkers of oxidative stress, 8-isoprostane and 8-OHdG, were analyzed using enzyme-linked immunosorbent assays (ELISA). Long-term (pregnancy through 3 months prenatal) exposure to particulate matter and airborne PAHs were measured. Results Women using wood-fueled chimney stoves are exposed to high levels of particulate matter (median 48-hour PM2.5 105.7 μg/m3; inter-quartile range (IQR): 77.6–130.4). Urinary PAH and VOC metabolites were significantly associated with woodsmoke exposures: 2-naphthol (median (IQR) in ng/mg creatinine: 295.9 (74.4–430.9) after sauna versus 23.9 (17.1–49.5) fasting; and acrolein: 571.7 (429.3–1040.7) after sauna versus 268.0 (178.3–398.6) fasting. Urinary PAH (total PAH: ρ = 0.89, p < 0.001) and VOC metabolites of benzene (ρ=0.80, p < 0.001) and acrylonitrile (ρ=0.59, p < 0.05) were strongly correlated with long-term exposure to particulate matter. However urinary biomarkers of oxidative stress were not correlated with particulate matter (ρ = 0.01 to 0.05, p > 0.85) or PAH and VOC biomarkers (ρ =−0.20 to 0.38, p > 0.07). Urinary metabolite concentrations were significantly greater than those of heavy smokers (mean cigarettes/day = 18) across all PAHs. In 15 (65%) women, maximum 1-hydroxypyrene concentrations exceeded the occupational exposure limit of coke-oven workers. Conclusions The high concentrations of urinary PAH and VOC metabolites among recently pregnant women is alarming given the detrimental fetal and neonatal effects of prenatal PAH exposure. As most women used chimney woodstoves, cleaner fuels are critically needed to reduce smoke exposure.

Household air pollution (HAP) generated from solid fuel combustion is a major health risk. Direct measurement of exposure to HAP is burdensome and challenging, particularly for children. In a pilot study of the Household Air Pollution Intervention Network (HAPIN) trial in rural Guatemala, we evaluated an indirect exposure assessment method that employs fixed continuous PM 2.5 monitors, Bluetooth signal receivers in multiple microenvironments (kitchen, sleeping area and outdoor patio), and a wearable signal emitter to track an individual's time within those microenvironments. Over a four-month period, we measured microenvironmental locations and reconstructed indirect PM 2.5 exposures for women and children during two 24-h periods before and two periods after a liquefied petroleum gas (LPG) stove and fuel intervention delivered to 20 households cooking with woodstoves. Women wore personal PM 2.5 monitors to compare direct with indirect exposure measurements. Indirect exposure measurements had high correlation with direct measurements (n = 62, Spearman ρ = 0.83, PM 2.5 concentration range: 5-528 µg/m 3). Indirect exposure had better agreement with direct exposure measurements (bias: −17 µg/m 3) than did kitchen area measurements (bias: −89 µg/m 3). Our findings demonstrate that indirect exposure reconstruction is a feasible approach to estimate personal exposure when direct assessment is not possible.