The World Health Organization claims that household air pollution (HAP) kills 4.3 million people per year. Is this true?
Reader Note: This is a work in progress. It is a pretty big job. I will try to add to it regularly and produce a summary report in the end. But I don’t want to just sit on the results as parts are completed. When updates are made, a notice will be posted on JunkScience.
JunkScience.com has already debunked the notion that airborne fine particulate matter (PM2.5) kills either on a short-term or long-term basis. That debunking also applies to HAP. But if you need more evidence specifically on HAP, read on.
This is the 2016 WHO report that makes the claim at issue, “Burning Opportunity: Clean Household Energy for Health, Sustainable Development, and Wellbeing of Women and Children.
The WHO’s claimed “key facts” are:
- Around 3 billion people cook and heat their homes using open fires and simple stoves burning biomass (wood, animal dung and crop waste) and coal.
- Over 4 million people die prematurely from illness attributable to the household air pollution from cooking with solid fuels.
- More than 50% of premature deaths due to pneumonia among children under 5 are caused by the particulate matter (soot) inhaled from household air pollution.
- 3.8 million premature deaths annually from noncommunicable diseases including stroke, ischaemic heart disease, chronic obstructive pulmonary disease (COPD) and lung cancer are attributed to exposure to household air pollution.
Let’s first look at the WHO’s own references purporting to lend credibility to its claims about HAP and death. These references are listed on pages 110-113 of the WHO report.
Study #1. Armstrong JR, Campbell H. Indoor air pollution exposure and lower respiratory infections in young Gambian children. Int J Epidemiol. 1991 Jun;20(2):424-9.
Evaluation. I have marked up the study here. The study does not address death, just respiratory infections — which in any event are not the “noncommunicable diseases” that make up 95% of the WHO’s claim. Exposure to cooking smoke is conflated with exposure to parental smoking. The flaky statistical results are inconsistent between boys and girls. All data is self-reported. There are no measured or even estimated exposures to cooking smoke. This study does not support the WHO claim that cooking smoke causes death.
Study #2. Bates MN et al. Acute lower respiratory infection in childhood and household fuel use in Bhaktapur, Nepal. Environ Health Perspect. 2013 May;121(5):637-42. doi: 10.1289/ehp.1205491. Epub 2013 Mar 1
Evaluation. I have marked up the study here. The study does not address death from noncommunicable disease, merely acute lower respiratory infections. The study is notable for its nonsensical results reporting kerosene was a greater risk factor for illness than biomass burning.
Study #3. Bruce NG et al. Control of household air pollution for child survival: estimates for intervention impacts. BMC Public Health. 2013;13 Suppl 3:S8. doi: 10.1186/1471-2458-13-S3-S8. Epub 2013 Sep 17.
Evaluation. I have marked up this study, here. The child survival estimates are based on results from five studies on indoor smoke and child mortality. The results are all weak associations, mostly not statistically significant. The combination of these studies into a meta-analysis is inappropriate — and still only produces a weak association. Importantly, the results are based on all-cause mortality, an overly broad health endpoint. If indoor air kills, it would logically cause death from heart- and/or lung-related causes. If deaths were limited to just those heart- and/or lung-related, the associations would be even weaker if not nonexistent. We will evaluate the component studies individually later.
Study #4. Chafe Za et al. Household Cooking with Solid Fuels Contributes to Ambient PM2.5 Air Pollution and the Burden of Disease. Environ Health Perspect. 2014 Dec; 122(12): 1314–1320. Published online 2014 Sep 5. doi: 10.1289/ehp.1206340.
Evaluation. This paper merely assumes that fine particulate matter (PM2.5) causes death. It provides no information on whether HAP cause death. It just assumes it does.
Study #5. Chen Y et al. Evidence on the impact of sustained exposure to air pollution on life expectancy from China’s Huai River policy. Proc Natl Acad Sci U S A. 2013 Aug 6;110(32):12936-41. doi: 10.1073/pnas.1300018110. Epub 2013 Jul 8.
Evaluation. This study addresses outdoor air pollution, not HAP. Even so, it does not credibly link particulate matter in outdoor air with death because of its methodological limitations (i.e., it is ecologic epidemiology).
Study #6. Guttikunda SK et al. Particulate pollution in Ulaanbaatar, Mongolia. Air Quality, Atmosphere & Health September 2013, Volume 6, Issue 3, pp 589-601
Evaluation. The study abstract is here. The paper uses models to estimate deaths caused by PM2.5 in outdoor air. The authors assume (vs. try to show) that PM2.5 kills. This assumption is faulty. The study sheds no light on any the key issue deaths from HAP.
Study #7. Janssen N et al. Black Carbon as an Additional Indicator of the Adverse Health Effects of Airborne Particles Compared with PM10 and PM2.5.
Evaluation. The study address vehicle exhaust. It has nothing to do with HAP.
Study #8. Lam NL et al. Kerosene: a review of household uses and their hazards in low- and middle-income countries. J Toxicol Environ Health B Crit Rev. 2012;15(6):396-432. doi: 10.1080/10937404.2012.710134.
Evaluation. This article is a review of other published literature, including the parroting of EPA-funded claims that PM2.5 kills. It contains no original research or data.
Study #9. Lim SS et al. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012 Dec 15; 380(9859): 2224–2260. doi: 10.1016/S0140-6736(12)61766-8.
Evaluation.This is part of the so-called Global Burden of Disease study from which the WHO derives the body count attributable to HAP. This study, however, merely assumes that PM2.5 kills. Thew assumption is based on EPA-funded junk science concerning outdoor air pollution that we have already debunked elsewhere. So it’s not helpful in the determination of whether HAP actually does kill. It is worth observing the following: the notion that one can take the results of epidemiologic studies, throw in some imaginary exposure data across the global population and then calculate global body counts is entirely inappropriate and bankrupt. Epidemiology studies are intended to help determine whether a causal relationship exists, not how large risk may be if a causal relationship is found. This study is statistical malpractice, not science.
Study #10. Smith KR et al. Effect of reduction in household air pollution on childhood pneumonia in Guatemala (RESPIRE): a randomised controlled trial. Lancet Volume 378, No. 9804, p1717–1726, 12 November 2011.
Evaluation. This study reports that reducing indoor wood smoke emissions had no effect on clinically-diagnosed pneumonia in children. This study does not support the WHO’s claims for HAP killing people.
So the above-summarized 10 studies are the published research the WHO cites to support its claims that HAP kills millions per year. In fact, none of these studies support the WHO’s claims on any sort of empirical basis. Now we will turn to examining studies referenced, to the extent this has been done, by the 10 studies — e.g., the component studies of the meta-analysis of Study #3 (Bruce NG) and the studies cited in Study #9 (Global Burden of Disease study).
Study #11. Kashima S et al. Effects of traffic-related outdoor air pollution on respiratory illness and mortality in children, taking into account indoor air pollution, in Indonesia. J Occup Environ Med. 2010 Mar;52(3):340-5. doi: 10.1097/JOM.0b013e3181d44e3f.
Evalulation. This study is included in the meta-analysis by Bruce. Moving past the important fact that the study was designed to study the effects of traffic-related outdoor air quality (not the effects of HAP), the study results, as presented by Bruce, were statistically nonsignificant (i.e., the relative risk confidence intervals all included the no-effect level of 1.0). So this study does not help the WHO’s case.
Study #12. Tielsch JM et al. Exposure to indoor biomass fuel and tobacco smoke and risk of adverse reproductive outcomes, mortality, respiratory morbidity and growth among newborn infants in south India. Int J Epidemiol. 2009 Oct;38(5):1351-63. doi: 10.1093/ije/dyp286. Epub 2009 Sep 16.
Evaluation. This study is included in the meta-analysis by Bruce. The study results are statistically nonsignificant (i.e., the relative risk confidence intervals all included the no-effect level of 1.0). So this study does not help the WHO’s case.
Study #13. Bassani DG et al. Child mortality from solid-fuel use in India: a nationally-representative case-control study. BMC Public Health. 2010; 10: 491. Published online 2010 Aug 17. doi: 10.1186/1471-2458-10-491.
Evaluation. This study is included in the meta-analysis by Bruce. Though the study reports significant associations between indoor solid fuel use and child mortality, the statistical results are weak (i.e., prevalence ratios below 2.0). The insurmountable problem with this study is that all-cause mortality is the health endpoint versus more biologically plausible heart- and lung-related deaths. Limiting the study to biologically plausible deaths would likely dramatically change the results. The authors acknowledge this as follows: “One of the major limitations of the study is the fact that we did not have information on cause-specific mortality in this study.” So this study does not establish that HAPs kill children.
Study #14. Mtango FD. Risk factors for deaths in children under 5 years old in Bagamoyo district, Tanzania.Trop Med Parasitol. 1992 Dec;43(4):229-33.
A copy of this study is not readily available online. The abstract is here. Although the reports a weak association between “child sleeping in the room where cooking is done” and death (relative risk = 2.78), the study was designed to look at risk factors for death and not specifically designed to study the hypothesis that HAP causes death. This poor design also produced an association between “the child eating with others ” and death that was more than three times larger than that for “child sleeping int the room where the cooking is done.” More importantly, the study only considers all-cause mortality, rather than heart- and/or lung-related mortality.
Study #15. Wichmann JL. Influence of cooking and heating fuel use on 1-59 month old mortality in South Africa. Matern Child Health J. 2006 Nov;10(6):553-61.
The abstract for the study is here. The statistical association is weak (relative risk = 1.95) and the confidence interval verges on the no-effect level (1.04 – 3.68). Since the study relies on the all-cause mortality endpoint, these marginal results would likely evaporate if the health endpoint was narrowed to mortality from heart- or lung-related causes. The authors acknowledge “the potential for residual confounding despite adjustment.” Moreover, the study relies on outdoor air pollution junk science to bolster its reported correlation (“the better documented evidence on outdoor air pollution and mortality suggest this association may be real”). But outdoor kills no one and so this study establishes nothing.
So the component studies of the Bruce meta-analysis fail to support the WHO’s claim about HAPs. What about the studies referenced by Study #9, the Global Burden of Disease Study? The study is here. We will be considering these references from Study #9:
GBOD Referenced Study #132.Pope CA 3rd. Lung cancer and cardiovascular disease mortality associated with ambient air pollution and cigarette smoke: shape of the exposure-response relationships. Environ Health Perspect. 2011 Nov;119(11):1616-21. doi: 10.1289/ehp.1103639. Epub 2011 Jul 19.
Evaluation. This study does not address HAP. The study is an effort by EPA-funded researchers to explain a paradox they created: Low levels of PM2.5 in outdoor cause cause death but much higher levels of PM2.5 from secondhand smoke and direct smoking do not. See GBOD Referenced Study #133, below, for further discussion.
GBOD Referenced Study #133. Pope CA 3rd et al. Cardiovascular mortality and exposure to airborne fine particulate matter and cigarette smoke: shape of the exposure-response relationship. Circulation. 2009 Sep 15;120(11):941-8. doi: 10.1161/CIRCULATIONAHA.109.857888. Epub 2009 Aug 31.
Evaluation. The study does not address HAP. The study is an effort by EPA-funded researchers to explain a paradox they created: Low levels of PM2.5 in outdoor cause cause death but much higher levels of PM2.5 from secondhand smoke and direct smoking do not. Their crazy explanation is that, “The exposure-response relationship between cardiovascular disease mortality and fine particulate matter is relatively steep at low levels of exposure and flattens out at higher exposures.” This phenomenon, of course, directly contradicts the time-honored toxicological principle of “the dose makes the poison.” To do reach this conclusion, basic principles of epidemiology were entirely and wantonly ignored in this paper. The study authors, for example, treated relative risks (i.e, correlations) as probabilistic estimates. This is fundamental malpractice as a relative risk is mere indicator of the strength of correlation in a particular study population; it is not portable measure of risk that may be applied to other populations.
GBOD Referenced Study #134. Krewski D et al. Extended follow-up and spatial analysis of the American Cancer Society study linking particulate air pollution and mortality. Res Rep Health Eff Inst. 2009 May;(140):5-114; discussion 115-36.
Evaluation. This study does not address HAP. It addresses PM2.5 in U.S. outdoor air. You can read the synopsis of the study here. But it is irrelevant to the issue at hand — whether and to what extent HAP kills.
GBOD Referenced Study #135. Brook RD et al. Particulate matter air pollution and cardiovascular disease: An update to the scientific statement from the American Heart Association. Circulation. 2010 Jun 1;121(21):2331-78. doi: 10.1161/CIR.0b013e3181dbece1. Epub 2010 May 10.
Evaluation. This is not a study of HAP. It purports to be a review of PM2.5 research. It is authored by the usual cabal of EPA-funded researchers. No more need for discussion of it here.
GBOD Referenced Study #136. Committee on the Medical Effects of Air Pollutants. The Mortality Effects of Long-Term Exposure to Particulate Air Pollution in the United Kingdom. London. Health Protection Agency. 2009.
This report is here. It does not address HAP. It merely assumes that PM2.5 in outdoor air kills and then estimates population body counts and years of life lost.
GBOD Referenced Study #137. Cooke RM et al. A probabilistic characterization of the relationship between fine particulate matter and mortality: elicitation of European experts. Environ Sci Technol. 2007 Sep 15;41(18):6598-605.
Evaluation. This study addresses particulate matter from the Kuwait Oil Fires — not HAP. This study suggests that the risk of death from PM2.5 may be greater than estimated by other outdoor air studies — which not only is not justified by any facts but directly contradicts the GBOD studies authored by Pope mentioned above.
GBOD Referenced Study #138. Burnett RT et al. An integrated risk function for estimating the global burden of disease attributable to ambient fine particulate matter exposure. Environ Health Perspect. 2014 Apr;122(4):397-403. doi: 10.1289/ehp.1307049. Epub 2014 Feb 11.
Evaluation. The study is here. It does not address the key issue of whether HAP causes death; it merely assumes that it does.
GBOD Referenced Study #139. Baumgartner J et al. Indoor air pollution and blood pressure in adult women living in rural China. GBOD Referenced Study #138.
Evaluation. The study reports that HAP “may” be associated with slight increases in blood pressure. Not only are the estimated increases slight, clinically insignificant and of speculative long-term health impact, they are not death. So this study does not support the WHO’s claim of HAP causing death.
GBOD Referenced Study #140. Smith KR et al. Effect of reduction in household air pollution on childhood pneumonia in Guatemala (RESPIRE): a randomised controlled trial. Lancet Volume 378, No. 9804, p1717–1726, 12 November 2011.
Evaluation. This study has already been evaluated (Study #10, above) as unhelpful to the WHO.s HAP claim.
We will now examine other published studies on HAP to see if any support the WHO’s claims. If you run across a study that is not evaluated here, please feel free to submit it to JunkScience.com.
First, we will evaluate studies found through PubMed with the search term “household air pollution and mortality.”
PubMed Study #1. Mitter SS et al. Household Fuel Use and Cardiovascular Disease Mortality: Golestan Cohort Study. Circulation. 2016 Jun 14;133(24):2360-9. doi: 10.1161/CIRCULATIONAHA.115.020288.
Evaluation. The study is here ($$$). The table with the results is below.
As you can see, all the associations are “no effect”, weak and/or statistically insignificant — al of which means this study reports nothing of interest. Though the authors tout the kerosene/diesel results, kerosene and diesel burn cleaner than wood or animal dung. This study does not help WHO’s case that HAP cause death.
PubMed Study #2. Naz S et al. Household air pollution and under-five mortality in India (1992-2006). Environ Health. 2016 Apr 26;15:54. doi: 10.1186/s12940-016-0138-8.
Evaluation. The study is here. More weak statistical associations in a study where the health endpoint was all-cause mortality, which over-includes deaths potentially attributable to HAP. Much study “data” is self-reported and not subject to verification. The authors rely on the WHO’s GBOD study to support their conclusions. But as pointed out earlier, the GBOD study relies on debunked EPA junk science.
PubMed Study #3. Kim C et al. Cooking Coal Use and All-Cause and Cause-Specific Mortality in a Prospective Cohort Study of Women in Shanghai, China. Environ Health Perspect. 2016 Apr 19.
Evaluation. The study is here. The result of interest is highlighted in the table, below.
While the study is easily dismissed because of its weak and/or nonsignificant associations, for the sake of argument, let’s not do that for the highlighted result. The group of women at issue (coal use for 30+ years) seems to contain somewhere between 10,000 to 15,000 women. Yet only 115 are in the group have supposedly died from heart disease. That is a very small group. Heart disease is very common. The exposures to PM2.5 are not known by the researchers. Not persuasive.
PubMed Study #4. Patel AB et al. Impact of exposure to cooking fuels on stillbirths, perinatal, very early and late neonatal mortality – a multicenter prospective cohort study in rural communities in India, Pakistan, Kenya, Zambia and Guatemala. Matern Health Neonatol Perinatol. 2015 Jul 21;1:18. doi: 10.1186/s40748-015-0019-0. eCollection 2015.
Evaluation. The study is here. Its weak associations are further attenuated/eviscerated by failure determine whether cooking was indoor or outdoor, failure to account for socio-economic status and failure to ascertain medical cause of reported miscarriages. Also note the weak association become nonsignificant in babies that survive the first three days.