Chimney stove intervention to reduce long-term wood smoke exposure lowers blood pressure among Guatemalan women

Abstract

Background and objective: RESPIRE, a randomized trial of an improved cookstove, was conducted in Guatemala to assess health effects of long-term reductions in wood smoke exposure. Given the evidence that ambient particles increase blood pressure, we hypothesized that the intervention would lower blood pressure.

Methods: TWO STUDY DESIGNS WERE USED: a) between-group comparisons based on randomized stove assignment, and b) before-and-after comparisons within subjects before and after they received improved stoves. From 2003 to 2005, we measured personal fine particle (particulate matter with aerodynamic diameter < 2.5 microm; PM(2.5)) exposures and systolic (SBP) and diastolic blood pressure (DBP) among women > 38 years of age from the chimney woodstove intervention group (49 subjects) and traditional open wood fire control group (71 subjects). Measures were repeated up to three occasions.

Results: Daily average PM(2.5) exposures were 264 and 102 microg/m(3) in the control and intervention groups, respectively. After adjusting for age, body mass index, an asset index, smoking, secondhand tobacco smoke, apparent temperature, season, day of week, time of day, and a random subject intercept, the improved stove intervention was associated with 3.7 mm Hg lower SBP [95% confidence interval (CI), -8.1 to 0.6] and 3.0 mm Hg lower DBP (95% CI, -5.7 to -0.4) compared with controls. In the second study design, among 55 control subjects measured both before and after receiving chimney stoves, similar associations were observed.

Conclusion: The between-group comparisons provide evidence, particularly for DBP, that the chimney stove reduces blood pressure, and the before-and-after comparisons are consistent with this evidence.

Figure 1

Between-group comparisons. Smoothed probability densities by randomized stove group (control, n = 71; intervention, n = 49) during the trial period (June 2003 through December 2004) of (A) 24-hr average PM_2.5; (B) SBP; and (C) DBP. Mean ± SD for each distribution shown in parentheses next to number of repeated measures (n).

Figure 2

Before-and-after comparisons. Smoothed probability densities for control group by study period (trial period June 2003 through December 2004, echo-intervention period March 2004 through March 2005) of (A) 24-hr average PM_2.5; (B) SBP; and (C) DBP. Mean ± SD for each distribution shown in parentheses next to number of repeated measures (n).