Clean Cooking Fuels may Improve Health during Pregnancy: NEJM

In the article by Clasen et al., the HAPIN investigators report their findings on the effects of LPG stoves on birth weight, the first of four primary outcomes of the trial. A total of 3200 pregnant women across four countries (Guatemala, India, Peru, and Rwanda) were randomly assigned to continue cooking with a biomass stove or to switch to an LPG stove. The women in the intervention households were provided with a free LPG stove and fuel during pregnancy and through the first year of the infant's life. Contrary to expectations, the mean birth weight did not differ significantly between 1593 infants born to women in intervention households and 1607 infants born to women in control households (2921 g and 2898 g, respectively). In addition, no demonstrable benefit of cooking with LPG stoves over biomass stoves was observed with regard to the prevalence of low birth weight or the incidence of preterm birth or stillbirth.

The HAPIN trial builds on almost two decades of trials of interventional stoves designed to improve the health of pregnant women and young children. In the first of these trials, which was conducted in Guatemala during the years 2002 through 2004, a total of 534 households with a pregnant woman or young infant were randomly assigned to continue cooking with a traditional unvented wood-burning stove or to switch to one outfitted with a chimney. In the households that used stoves with venting, the incidence of physician-diagnosed pneumonia before a child reached 18 months of age did not differ significantly from that in the households that used unvented stoves, although the incidence of severe pneumonia was lower by approximately a third. In a post hoc analysis that included women who were pregnant during the trial, birth weight appeared to be higher in the households that used vented stoves than in those that used unvented stoves, although the differences were not significant.

The lack of substantive effects of the intervention was attributed to an inadequate reduction in exposure, since pollutants were relocated outdoors rather than minimized at the source. Several additional randomized trials of improved stoves that were designed to burn biomass more efficiently with less smoke were subsequently undertaken; yet again, these trials showed little to no benefit with respect to pregnancy outcomes or respiratory health in children. Subsequent randomized trials assessing the use of "cleaner" liquid fuels such as LPG (vs. biomass) in pregnancy likewise yielded null results; only a trial in Nigeria that investigated cooking with ethanol showed better outcomes with the interventional stoves than with the control stoves, and the difference in birth weight was significant only in the adjusted results.

To overcome some shortcomings that were identified in the earlier trials, the HAPIN trial was designed to include a larger sample from multiple countries and to pair the intervention with intensive behavior-based messaging about the benefits of clean cooking to minimize stove stacking (i.e., concurrent use of traditional biomass fuels). The trial had high intervention fidelity and showed that measured personal exposures to fine particulate matter, black carbon, and carbon monoxide in the intervention households were lower than in the control households by 66%, 71%, and 83%, respectively, reductions that were greater than those achieved in previous trials.

A potential explanation of the null findings is that the intervention may not have been introduced early enough to have an influence on birth weight. The results of exploratory analyses, in which the benefits appeared to be greater with earlier introduction of the intervention, support this hypothesis. Furthermore, despite impressive reductions in exposures, they may not have been sufficient to lead to an increase in birth weight. A nonlinear exposure–response curve between air pollution and other health outcomes such as cardiovascular disease has been shown, with the steepest part of the curve at the lowest exposures. It is plausible that improvements in birth outcomes will be attainable only with even greater reductions in exposures that may require electrification or other innovations.

Although Clasen et al. did not find that replacing biomass stoves with LPG stoves led to an improvement in birth weight, the investigators reported only one of four primary outcomes; it remains to be seen whether the LPG intervention can reduce the incidences of severe childhood pneumonia and stunting or lower the blood pressures in older women residing in the household. Whatever the results, efforts to transition to clean and affordable energy for the global population living in poverty should not be abandoned. Reducing household air pollution is just one component of a broader strategy to improve air quality; improving the health of pregnant women and their infants will also require reductions in pollution from sources beyond the hearth such as traffic, trash burning, or commercial cooking. Clean air is a priority for communities and cannot be achieved by relying solely on changes in individual or household behaviors.

Source: Blair J. Wylie, M.D., M.P.H., and Kwaku P. Asante, M.B., Ch.B., M.P.H., Ph.D.; n engl j med 387;19