Invited Perspective: A Critical Part of a Real-World Environmental Health Trial Is to Demonstrate That the Intervention Reduced Exposure

Observational studies have associated a wide variety of health problems with biomass-derived household air pollution. However, previous high-profile clinical trials of improved cookstoves did not demonstrate significant reductions in adverse health outcomes in the primary prespecified intent-to-treat analyses.^1–3 A major concern about these trials is that the interventions used did not reduce exposures to cookstove-derived pollutants sufficiently to result in health improvements.⁴ In this issue of Environmental Health Perspectives, Johnson et al. report that the Household Air Pollution Intervention Network (HAPIN) trial appears to have overcome the challenges involved in reducing exposures.⁵ The next step is to demonstrate health improvements.

Large clinical trials of environmental interventions in resource-poor countries are unusual for many reasons: They are difficult to conduct, expensive, and fraught with logistical challenges. They can also raise ethical concerns, including concerns about withholding interventions from the control group.⁶ In the case of cooking-related exposures, prior clinical trials resulted in uneven results. The ensuing conversations focused on two issues. First, many participants did not exclusively or continually use novel cooking approaches, for a variety of structural, logistical, economic, or social reasons. Second, changing stove technologies without also changing the biomass fuel sources yielded only modest improvement in pollution, such that the intervention pollutant concentrations were not reduced deeply into the postulated “steep” part of the concentration–response relationship.⁷

In 2015, the National Institutes of Health solicited applications for a large multisite clinical trial to assess whether a cookstove and fuel intervention to reduce household air pollutants in low- and middle-income countries would achieve improvements in health outcomes.⁸ With additional support from the Bill & Melinda Gates Foundation, HAPIN was born. HAPIN investigators, with centers in India, Rwanda, Guatemala, and Peru, randomly assigned 3,200 pregnant women who used biomass cookstoves to either receive liquified petroleum gas stoves and fuel supplies or continue using their current stoves.^9–11 As primary outcomes, the trial sought to determine whether the intervention reduced low birth weight, prevented severe pneumonia incidence in infancy, prevented childhood stunting, and reduced blood pressure in older household members.

The HAPIN trial began recruitment in May 2018, with data collection continuing through August 2021. The investigators encountered all the challenges that are typically expected for a large clinical trial of its type. However, the study was further complicated by the COVID-19 pandemic through a major portion of the period for collecting health outcome information.

Nevertheless, the research team persevered, and the striking results reported by Johnson et al. demonstrate the effectiveness of their intervention in lowering pollutant concentrations. This study reflects not only the use of a much cleaner fuel than in prior large trials but also the efforts of the team to ensure adherence to the protocol and prevent “stacking” of stoves—the practice of continuing to use polluting biomass stoves in addition to the cleaner intervention device.¹² Average pollutant concentrations fell in the intervention group from $>100\mathrm{\mu }\mathrm{g}/{\mathrm{m}}^3$ at baseline to $<35\mathrm{\mu }\mathrm{g}/{\mathrm{m}}^3$. Concentrations remained unchanged in the control homes. Reductions were consistent across the four sites and persisted through follow-up. Although these dramatic improvements did not achieve the concentrations set by the U.S. Environmental Protection Agency¹³ and the World Health Organization¹⁴ for ambient exposures, there is reason to believe they have moved exposures meaningfully down the hypothesized concentration–response relationship.

Ambitious trials such as HAPIN, if successful, can accomplish many things. By using a robust experimental design, a well-conducted randomized clinical trial may convince those who question the importance of findings seen in observational research. It can tell us whether an intervention reduces exposures. It can answer questions about the health effects of the exposures of interest, including exposure–response relationships. It can inform us about the feasibility of interventions. It can investigate the barriers (technical, social, and behavioral) to interventions. Ultimately, and depending on the results, it can motivate investment of resources in public health interventions, setting the stage for implementation research and action based on the findings. Even as some have argued that randomized trials may not be necessary to understand environmental health effects,¹⁵ a carefully designed and ethically conducted trial certainly can be an influential component of the evidence base.¹⁶

The extraordinary efforts of the HAPIN investigative team have paid off with regard to identifying several opportunities. It clearly is possible to implement a clean cooking technology intervention in highly varied resource-poor environments on several continents—if you have the funds and tenacity to do so. We also now know that doing so will substantially reduce people’s exposure to pollutants over some period of time. Now the stage is set for the health analyses that will follow.

Refers to https://doi.org/10.1289/EHP10295