Abstract
Background
Household air pollution is the third largest risk factor for global disease burden, but direct links with cardiovascular disease mortality are limited. This study aimed to evaluate the relationship between household fuel use and cardiovascular disease mortality.
Methods and Results
The Golestan Cohort Study in northeastern Iran enrolled 50045 individuals aged 40 to 75 years between 2004 and 2008, and collected data on lifetime household fuel use and other baseline exposures. Participants were followed through 2012 with a 99% successful follow-up rate. Cox proportional hazards models were fitted to calculate hazard ratios (HRs) for associations between pehen (local dung), wood, kerosene/diesel, or natural gas burning for cooking and heating and all-cause and cause-specific mortality, adjusting for lifetime exposure to each of these fuels and potential confounders. 3073 participants (6%) died during follow-up, 78% of which were attributable to non-communicable diseases, including cardiovascular, oncologic and respiratory illnesses. Adjusted 10-year HRs from kerosene/diesel burning were 1.06 (95% CI 1.02-1.10), and 1.11 (1.06-1.17), respectively, for all-cause and cardiovascular mortality. Subtype-specific analyses revealed a significant increase in ischemic heart disease (10-year HR 1.14 (1.06-1.21)) and a trend toward cerebrovascular accident (10-year HR 1.08 (0.99-1.17)) mortality. Stratification by sex revealed a potential signal for increased risk for all-cause and cardiovascular disease mortality among women versus men, with similar risk for ischemic heart disease mortality.
Conclusions
Household exposure to high-pollution fuels was associated with increased risk for all-cause and cardiovascular disease mortality. Replicating these results worldwide would support efforts to reduce such exposures.
Keywords: air pollution, cerebrovascular accident/stroke, epidemiology, ischemic heart disease, mortality
Introduction
According to the WHO, one-half of the world's population lives in poverty and burns solid fuels for lighting, cooking and heating purposes, and the resultant household air pollution (HAP) is the third largest risk factor for global disease burden, contributing to an estimated 3.5 million deaths in 2010.1,2 While these deaths are traditionally attributed to pneumonia in children and chronic obstructive pulmonary disease and lung cancer in adults,3 data attempting to show a direct link between HAP and cardiovascular disease (CVD) mortality are limited.4 Inferences regarding the relationship between HAP and CVD are largely based on a growing body of evidence showing that HAP is associated with CVD risk factors and morbidity,5-13 as well as known data associating ambient air pollution and direct tobacco inhalation with CVD mortality.14-16
Given the increasing prevalence of CVD worldwide, it is prudent to identify novel modifiable risk factors for CVD burden that may be endemic to low- and middle-income countries (LMICs).17,18 Therefore, the objective of this study is to evaluate the association between household fuel use and CVD mortality.
Methods
The Golestan Cohort Study, originally a prospective study of esophageal cancer, enrolled 50,045 individuals aged 40-75 years in Gonbad City and 326 villages in the Golestan province of northeastern Iran, between 2004-2008. Inclusion and exclusion criteria, and methods of data collection, have been discussed in detail elsewhere.19 The ethical review committees of the Tehran University of Medical Sciences, the International Agency for Research on Cancer, and the National Cancer Institute approved the study methods and all participants signed a written informed consent at enrollment. A lifestyle questionnaire was administered to each study participant to collect information on age at enrollment, sex, ethnicity, socioeconomic status, urban or rural residence, physical activity, diet, tobacco consumption, alcohol consumption, opium consumption, and personal medical history. Anthropometric data and blood pressure were measured and recorded for each participant. Economic status was derived from a composite score of wealth, in accord with earlier publications,20 by applying multiple correspondence analyses to appliance ownership data (including bath in residence, personal car, motorbike, black and white television, color television, refrigerator, freezer, vacuum, and washing machine).
The lifestyle questionnaire also tabulated years of exposure to household fuels for cooking and heating, separately, over each study participant's lifetime. Included fuels were natural gas (liquefied petroleum gas), kerosene/diesel, wood, and pehen (local term for cow dung). Lifetime exposure to each fuel type was calculated for cooking and heating separately, and then combined to obtain an aggregate lifetime exposure (in years) to each fuel type, while accounting for overlap periods to avoid overestimation of the years of exposure. Each year of overlap was counted as one year of fuel exposure in the final analysis. A sensitivity analysis was performed by counting each year of overlap as two years of exposure, and yielded no significant change in the final results. Fuel exposure was treated as both a continuous and categorical variable in order to obtain risk of death for each year of fuel burning and assess mortality estimates due to increasing levels of fuel exposure, respectively.
Participants were followed through the end of December 2012 or until death. Annual phone calls were made to collect information on outpatient procedures, illnesses, hospitalizations, and death. Inability to contact a study participant after seven attempts over two weeks prompted investigators to contact friends or local health workers to help collect information on study participants. This method yielded successful follow-up of 99% of all study participants.21
Notification of participant death was followed-up by a visit from a general practitioner to the home of the deceased to complete a validated verbal autopsy questionnaire via interview of relatives.22 Other ancillary information, including relevant medical documents detailing lab, electrocardiographic and radiographic tests, pathology reports, physician notes, and hospital discharge information were collected and reviewed by two independent internists to determine cause of death. In the event of disagreement, a third experienced internist also reviewed all known information surrounding a participant's death. All results were recorded in a computerized database according to ICD-10 codes.
Statistical analysis
Cox proportional hazard models were fitted to estimate unadjusted and adjusted 10-year hazard ratios and 95% confidence intervals for all-cause and cause-specific mortalities with relation to each specific fuel type. Given that participants used multiple fuels over their lifetimes, often concurrently, the models included lifetime exposure to all fuels. The final model therefore included all fuel types, as well as those covariates for which unadjusted 10-year hazard ratios (HRs) were significant and those which otherwise would be considered to impact all-cause and cause-specific mortality. These included age at enrollment, sex, urban or rural residence, Turkmen ethnicity, socioeconomic category (low, low-medium, medium-high, and high based on the economic score detailed above), body mass index (BMI) (≤ 17.9 kg/m2, 18.0-24.9 kg/m2, 25.0-29.9 kg/m2, and ≥ 30.0 kg/m2 implying underweight, normal, overweight and obese respectively), physical activity (irregular non-intense, regular non-intense and irregular/regular intense), any history of tobacco, alcohol, and opium consumption, as well personal history of diabetes mellitus, hypertension, and ischemic heart disease. Mediation and effect modification analyses were performed for many of these variables to better understand potential causes pathways, as well as populations most vulnerable to the effects of fuel exposure. P values ≤ 0.05 were considered statistically significant. All statistical analyses were performed using STATA version 11.1 (StataCorp, College Station, TX, USA).
Results
Of the 50,045 study participants, 28,824 were female (58%) and 21,221 were male (42%) with a mean age of 52.1 years at study enrollment. Most study participants were of Turkmen ancestry (74%) and lived in rural areas (80%). Whereas 86% of women had no formal education, 49% of men had no formal education. The population was fairly evenly distributed across all socioeconomic levels. Over half of the population was overweight or obese. History of smoking, opium consumption and alcohol use were reported by 22%, 17% and 3% of individuals, respectively. Self-reported prevalence of diabetes and ischemic heart disease were 7% and 6%, respectively. Hypertension was reported in 42% of the population.
Table 1 shows level of exposure to each fuel type, in 25-year increments, across select baseline demographics and medical history. While only 12% of participants burned pehen, 99% burned natural gas, 99% burned kerosene/diesel and 90% burned wood. Exposure distributions are shown in Supplemental Table 1, highlighting a greater amount of the population being exposed to natural gas and kerosene/diesel compared to wood and pehen. Generally, natural gas and kerosene/diesel users were of similar socioeconomic status wherein exposure times were higher with increasing socioeconomic status. Among individuals with medium-high and high socioeconomic status, 59, 57, 17 and 2.4% had natural gas, kerosene/diesel, wood and pehen burning exposure durations of 30 years or more, respectively. Pehen exposure was low overall, but highest among Turkmen people.
Table 1. Years of fuel exposure stratified by demographics and health characteristics (Data available for 50029 individuals).
| Fuel | Gas | Kerosene/Diesel | Wood | Pehen | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Years of Exposure | 0 | 1 to 25 | 26 to 50 | ≥51 | 0 | 1 to 25 | 26 to 50 | ≥51 | 0 | 1 to 25 | 26 to 50 | ≥51 | 0 | 1 to 25 | 26 to 50 | ≥51 |
| Sex, n (%) | ||||||||||||||||
| Male | 138 (0.7) | 6784 (32.0) | 13800 (65.1) | 491 (2.3) | 142 (0.7) | 6562 (30.9) | 13709 (64.6) | 800 (3.8) | 1477 (7.0) | 12133 (57.2) | 6626 (31.2) | 977 (4.6) | 18243 (86.0) | 1910 (9.0) | 975 (4.6) | 85 (0.4) |
| Female | 187 (0.7) | 8626 (29.9) | 19329 (67.1) | 674 (2.3) | 192 (0.7) | 8524 (29.6) | 19100 (66.3) | 1000 (3.5) | 3697 (12.8) | 16792 (58.3) | 7352 (25.5) | 975 (3.4) | 25694 (89.2) | 2025 (7.0) | 1006 (3.5) | 91 (0.3) |
| BMI, n (%) | ||||||||||||||||
| Underweight (≤ 17.9 kg/m2) | 53 (2.2) | 880 (36.5) | 1434 (59.4) | 46 (1.9) | 21 (0.9) | 717 (29.7) | 1554 (64.4) | 121 (5.0) | 152 (6.3) | 1251 (51.8) | 847 (35.1) | 163 (6.8) | 2088 (86.5) | 202 (8.4) | 114 (4.7) | 9 (0.4) |
| Normal (18.0-24.9 kg/m2) | 174 (1.0) | 6342 (35.3) | 11101 (61.8) | 346 (1.9) | 173 (1.0) | 5666 (31.5) | 11387 (63.4) | 737 (4.1) | 1401 (7.8) | 9645 (53.7) | 5923 (33.0) | 994 (5.5) | 15456 (86.0) | 1556 (8.7) | 859 (4.8) | 92 (0.5) |
| Overweight (25.0-29.9 kg/m2) | 64 (0.4) | 4946 (29.2) | 11489 (67.9) | 426 (2.5) | 73 (0.4) | 5011 (29.6) | 11281 (66.7) | 560 (3.3) | 1869 (11.0) | 10067 (59.5) | 4467 (26.4) | 522 (3.1) | 14954 (88.3) | 1288 (7.6) | 630 (3.7) | 53 (0.3) |
| Obese (≥ 30.0 kg/m2) | 34 (0.3) | 3239 (25.5) | 9100 (71.6) | 346 (2.7) | 67 (0.5) | 3688 (29.0) | 8583 (67.5) | 381 (3.0) | 1751 (13.8) | 7957 (62.6) | 2740 (21.5) | 271 (2.1) | 11431 (90.0) | 889 (7.0) | 377 (3.0) | 22 (0.2) |
| Ethnicity, n (%) | ||||||||||||||||
| Turkmen | 224 (0.6) | 11927 (32.0) | 24286 (65.2) | 800 (2.2) | 254 (0.7) | 11242 (30.2) | 24352 (65.4) | 1389 (3.7) | 3153 (8.5) | 21253 (57.1) | 11128 (29.9) | 1703 (4.6) | 31431 (84.4) | 3700 (9.9) | 1934 (5.2) | 172 (0.5) |
| Non-turkmen | 101 (0.8) | 3483 (27.2) | 8843 (69.1) | 365 (2.9) | 80 (0.6) | 3844 (30.1) | 8457 (66.1) | 411 (3.2) | 2021 (15.8) | 7672 (60.0) | 2850 (22.8) | 249 (2.0) | 12506 (97.8) | 235 (1.8) | 47 (0.4) | 4 (0.0) |
| Residence, n (%) | ||||||||||||||||
| Rural | 319 (0.8) | 14017 (35.0) | 25160 (62.9) | 502 (1.3) | 280 (0.7) | 11272 (28.2) | 26838 (67.1) | 1608 (4.0) | 2805 (7.0) | 22278 (55.7) | 13013 (32.5) | 1902 (4.8) | 34126 (85.3) | 3737 (9.3) | 1961 (4.9) | 174 (0.4) |
| Urban | 6 (0.1) | 1393 (13.9) | 7969 (79.4) | 663 (6.6) | 54 (0.5) | 3814 (38.0) | 5871 (59.5) | 192 (1.9) | 2369 (23.6) | 6647 (66.3) | 965 (9.6) | 50 (0.5) | 9811 (97.8) | 198 (2.0) | 20 (0.2) | 2 (0.0) |
| Education, n (%) | ||||||||||||||||
| Nil | 304 (0.9) | 12221 (34.8) | 21831 (62.2) | 768 (2.2) | 292 (0.8) | 10863 (30.9) | 22391 (63.8) | 1578 (4.5) | 2419 (6.9) | 18803 (53.5) | 12034 (34.3) | 1868 (5.3) | 29956 (85.3) | 3180 (9.1) | 1818 (5.2) | 170 (0.5) |
| ≤5th standard | 17 (0.2) | 2179 (25.8) | 6099 (72.2) | 158 (1.9) | 25 (0.3) | 2516 (29.8) | 5774 (68.3) | 138 (1.6) | 1230 (14.6) | 5733 (67.8) | 1431 (16.9) | 59 (0.7) | 7844 (92.8) | 477 (5.6) | 130 (1.5) | 2 (0.0) |
| 6th-8th standard | 4 (0.2) | 408 (18.2) | 1743 (77.7) | 88 (3.9) | 5 (0.2) | 584 (26.0) | 1612 (71.9) | 42 (1.9) | 443 (19.8) | 1553 (69.2) | 236 (10.5) | 11 (0.5) | 2122 (94.6) | 105 (4.7) | 14 (0.6) | 2 (0.1) |
| 9th-12th standard | 0 (0.0) | 485 (15.4) | 2559 (81.4) | 101 (3.2) | 8 (0.3) | 817 (26.0) | 2287 (72.7) | 33 (1.1) | 838 (26.7) | 2064 (65.6) | 229 (7.3) | 14 (0.5) | 2981 (94.8) | 145 (4.6) | 17 (0.5) | 2 (0.1) |
| College/Graduate | 0 (0.0) | 117 (11.0) | 897 (84.3) | 50 (4.7) | 4 (0.4) | 306 (28.8) | 745 (70.0) | 9 (0.9) | 244 (22.9) | 772 (72.6) | 48 (4.5) | 0 (0.0) | 1034 (97.2) | 28 (2.6) | 2 (0.2) | 0 (0.0) |
| Socioeconomic Status, n (%) | ||||||||||||||||
| Low | 268 (1.8) | 6564 (45.0) | 7511 (51.5) | 242 (1.7) | 201 (1.4) | 5335 (36.6) | 8334 (57.1) | 715 (4.9) | 734 (5.0) | 6790 (46.6) | 5917 (40.6) | 1144 (7.8) | 12044 (82.6) | 1574 (10.8) | 887 (6.1) | 80 (0.6) |
| Low-medium | 28 (0.3) | 3522 (34.9) | 6393 (63.4) | 148 (1.5) | 61 (0.6) | 3096 (30.7) | 6576 (65.1) | 358 (3.6) | 751 (7.4) | 5774 (57.2) | 3181 (31.5) | 385 (3.8) | 8815 (87.4) | 800 (7.9) | 439 (4.4) | 37 (0.4) |
| Medium-high | 13 (0.1) | 3342 (27.1) | 8744 (70.8) | 248 (2.0) | 40 (0.3) | 3408 (27.6) | 8512 (68.9) | 387 (3.1) | 1227 (9.9) | 7724 (62.6) | 3087 (25.0) | 309 (2.5) | 10882 (88.1) | 908 (7.4) | 501 (4.1) | 56 (0.5) |
| High | 16 (0.1) | 1982 (15.2) | 10479 (80.6) | 527 (4.1) | 32 (0.3) | 3247 (25.0) | 9385 (72.2) | 340 (2.6) | 2462 (18.9) | 8636 (66.4) | 1792 (13.8) | 114 (0.9) | 12914 (93.8) | 653 (5.0) | 154 (1.2) | 3 (0.0) |
| Physical Activity, n (%) | ||||||||||||||||
| Irregular, non-intense | 220 (0.7) | 9791 (31.9) | 20035 (65.3) | 627 (2.0) | 196 (0.6) | 9021 (29.4) | 20177 (65.8) | 1279 (4.2) | 2966 (9.7) | 16696 (54.4) | 9553 (31.1) | 1458 (4.8) | 26809 (87.4) | 2453 (8.0) | 1284 (4.2) | 127 (0.4) |
| Regular, non-intense | 64 (0.5) | 3516 (26.0) | 9486 (70.0) | 478 (3.5) | 87 (0.6) | 4294 (31.7) | 8798 (65.0) | 365 (2.7) | 1793 (13.2) | 8471 (62.5) | 2929 (21.6) | 351 (2.6) | 11814 (87.2) | 1093 (8.1) | 589 (4.4) | 48 (0.4) |
| Irregular or regular intense | 42 (0.7) | 2077 (36.5) | 3524 (61.9) | 50 (0.9) | 50 (0.9) | 1733 (30.4) | 3757 (66.0) | 153 (2.7) | 396 (7.0) | 3691 (64.8) | 1465 (25.7) | 141 (2.5) | 5241 (92.1) | 349 (6.1) | 102 (1.8) | 1 (0.0) |
| Ever Smoker, n (%) | ||||||||||||||||
| Yes | 107 (1.0) | 3512 (32.5) | 6945 (64.2) | 917 (2.3) | 79 (0.7) | 3297 (30.5) | 6986 (64.6) | 450 (4.2) | 787 (7.3) | 6026 (55.7) | 3434 (31.8) | 565 (5.2) | 9291 (85.9) | 981 (9.1) | 497 (4.6) | 43 (0.4) |
| No | 218 (0.6) | 11898 (30.3) | 26184 (66.8) | 248 (2.3) | 255 (0.7) | 11789 (30.1) | 25823 (65.9) | 1350 (3.4) | 4387 (11.2) | 22899 (58.4) | 10544 (26.9) | 1387 (3.5) | 34646 (88.3) | 2954 (7.5) | 1484 (3.8) | 133 (0.3) |
| Opiates Ever Consumed, n (%) | ||||||||||||||||
| Yes | 105 (1.2) | 2912 (34.3) | 5329 (62.7) | 148 (1.7) | 65 (0.8) | 2557 (30.1) | 5494 (64.7) | 378 (4.5) | 507 (6.0) | 4628 (54.5) | 2883 (33.9) | 476 (5.6) | 7199 (84.8) | 829 (9.8) | 434 (5.1) | 32 (0.4) |
| No | 220 (0.5) | 12494 (30.1) | 27797 (66.9) | 1017 (2.5) | 269 (0.7) | 12526 (30.2) | 27313 (65.8) | 1422 (3.4) | 4667 (11.2) | 24294 (58.5) | 11093 (26.7) | 1476 (3.6) | 36733 (88.5) | 3106 (7.5) | 1547 (3.7) | 144 (0.4) |
| Alcohol Ever Consumed, n (%) | ||||||||||||||||
| Yes | 6 (0.4) | 287 (16.7) | 1326 (77.0) | 103 (6.0) | 9 (0.5) | 432 (25.1) | 1227 (71.3) | 54 (3.1) | 234 (13.6) | 1181 (68.6) | 278 (16.1) | 29 (1.7) | 1613 (93.7) | 94 (5.5) | 14 (0.8) | 1 (0.1) |
| No | 319 (0.7) | 15123 (31.3) | 31803 (65.8) | 1062 (2.2) | 325 (0.7) | 14654 (30.3) | 31582(65.4′) | 1746 (3.6) | 4940 (10.2) | 27744 (57.4) | 13700 (28.4) | 1923 (4.0) | 42324 (87.6) | 3841 (8.0) | 1967 (4.1) | 175 (0.4) |
| Hypertension, n (%) | ||||||||||||||||
| Yes | 133 (0.6) | 6114 (28.8) | 14332 (67.5) | 644 (3.0) | 148 (0.7) | 6057 (28.5) | 14083 (66.4) | 935 (4.4) | 1730 (8.2) | 11584 (54.6) | 6809 (32.1) | 1100 (5.2) | 148 (0.7) | 6057 (28.6) | 14083 (66.4) | 935 (4.4) |
| No | 190 (0.7) | 9196 (32.1) | 18730 (65.4) | 517 (1.8) | 184 (0.6) | 8953 (31.3) | 18646 (65.1) | 850 (3.0) | 3436 (12.0) | 17264 (60.3) | 7093 (24.8) | 840 (2.9) | 184 (0.6) | 8953 (31.3) | 18646 (65.1) | 850 (3.0) |
| Diabetes Mellitus, n (%) | ||||||||||||||||
| Yes | 12 (0.4) | 828 (24.0) | 2495 (72.2) | 122 (3.5) | 19 (0.6) | 882 (25.5) | 2392 (69.2) | 164 (4.7) | 330 (9.6) | 2085 (60.3) | 926 (26.8) | 116 (3.4) | 3092 (89.4) | 247 (7.1) | 107 (3.1) | 11 (0.3) |
| No | 313 (0.7) | 14580 (31.3) | 30631 (65.8) | 1043 (2.2) | 315 (0.7) | 14201 (30.5) | 30415 (65.3) | 1636 (3.5) | 4844 (10.4) | 26837 (57.6) | 13050 (28.0) | 1836 (3.9) | 40840 (87.7) | 3688 (7.9) | 1874 (4.0) | 165 (0.4) |
| Ischemic Heart Disease, n (%) | ||||||||||||||||
| Yes | 17 (0.6) | 744 (24.4) | 2169 (71.1) | 121 (4.0) | 21 (0.7) | 800 (26.2) | 2086 (68.4) | 144 (4.7) | 194 (6.4) | 1711 (56.1) | 994 (32.6) | 152 (5.0) | 2719 (89.1) | 198 (6.5) | 114 (3.7) | 20 (0.7) |
| No | 308 (0.7) | 14666 (31.2) | 30960 (65.9) | 1044 (2.2) | 313 (0.7) | 14286 (30.4) | 30723 (65.4) | 1656 (3.5) | 4980 (10.6) | 27214 (57.9) | 12984 (27.6) | 1800 (3.8) | 41218 (87.7) | 3737 (8.0) | 1867 (4.0) | 156 (0.3) |
As of December 2012, there were 3073 deaths for which complete demographic data were available, of which 78% were attributable to non-communicable cardiovascular, cancer, and respiratory diseases and showed little difference in incidence across socioeconomic strata (Supplemental Table 2). Among deaths attributed to CVD (n = 1578), 928 (59%) were due to ischemic heart disease and 519 (33%) were due to cerebrovascular accident, with the remaining 8% attributable to other forms of CVD including sudden cardiac death, congestive heart failure, hypertension, chronic rheumatic heart disease, pulmonary embolism, and other cardiovascular system diseases not otherwise specified.
Cox proportional hazards models revealed an increased risk for all-cause mortality associated with kerosene/diesel use (10-year HR 1.06, 95% CI 1.02-1.10) relative to other fuel uses, in the multivariable-adjusted model. Natural gas burning, relative to other fuel uses, was associated with lower risk for death (10-year HR 0.95, 95% CI 0.91-0.98). (Table 2)
Table 2. Cox proportional hazard ratios for mortality risk per 10-years of fuel use for the total population and stratified by sex.
| TOTAL | MEN | WOMEN | ||||
|---|---|---|---|---|---|---|
|
| ||||||
| Fuel | Adjusted 10-year | p | Adjusted 10-year | p | Adjusted 10-year | p |
| Hazard Ratio (95% CI)* | Value | Hazard Ratio (95% CI)† | Value | Hazard Ratio (95% CI)† | Value | |
| All causes | n = 3073 | n = 1723 | n = 1350 | |||
|
| ||||||
| Gas | 0.95 (0.91-0.98) | 0.006 | 0.97 (0.92-1.03) | 0.29 | 0.92 (0.87-0.98) | 0.008 |
| Kerosene/Diesel | 1.06 (1.02-1.10) | 0.001 | 1.04 (0.99-1.09) | 0.13 | 1.09 (1.03-1.15) | 0.003 |
| Wood | 1.01 (0.97-1.05) | 0.70 | 1.00 (0.94-1.05) | 0.86 | 1.02 (0.96-1.08) | 0.54 |
| Pehen | 1.00 (0.97-1.03) | 0.93 | 1.00 (0.96-1.05) | 0.94 | 0.99 (0.94-1.04) | 0.70 |
|
| ||||||
| Cardiovascular Diseases | n = 1578 | n = 866 | n = 712 | |||
|
| ||||||
| Gas | 0.94 (0.89-0.99) | 0.02 | 0.97 (0.90-1.05) | 0.46 | 0.91 (0.84-0.99) | 0.02 |
| Kerosene/Diesel | 1.11 (1.06-1.17) | < 0.001 | 1.08 (1.01-1.16) | 0.04 | 1.14 (1.06-1.22) | < 0.001 |
| Wood | 1.03 (0.97-1.08) | 0.35 | 1.00 (0.93-1.08) | 1.00 | 1.05 (0.97-1.13) | 0.24 |
| Pehen | 1.03 (0.98-1.07) | 0.23 | 1.03 (0.97-1.09) | 0.38 | 1.01 (0.95-1.08) | 0.69 |
|
| ||||||
| Cancer | n = 670 | n = 365 | n = 305 | |||
|
| ||||||
| Gas | 0.97 (0.89-1.06) | 0.48 | 0.97 (0.87-1.09) | 0.62 | 0.97 (0.85-1.10) | 0.60 |
| Kerosene/Diesel | 0.99 (0.91-1.08) | 0.83 | 1.03 (0.92-1.14) | 0.63 | 0.94 (0.82-1.08) | 0.37 |
| Wood | 0.95 (0.87-1.03) | 0.22 | 0.98 (0.88-1.09) | 0.71 | 0.90 (0.78-1.04) | 0.14 |
| Pehen | 1.04 (0.97-1.11) | 0.27 | 1.04 (0.95-1.13) | 0.39 | 1.03 (0.93-1.14) | 0.63 |
|
| ||||||
| Respiratory Disease | n = 155 | n = 96 | n = 59 | |||
|
| ||||||
| Gas | 0.98 (0.83-1.16) | 0.86 | 1.06 (0.86-1.31) | 0.60 | 0.89 (0.68-1.17) | 0.41 |
| Kerosene/Diesel | 0.91 (0.76-1.08) | 0.28 | 0.91 (0.73-1.13) | 0.38 | 0.89 (0.67-1.20) | 0.45 |
| Wood | 0.83 (0.69-1.01) | 0.06 | 0.81 (0.64-1.03) | 0.09 | 0.85 (0.63-1.16) | 0.31 |
| Pehen | 1.02 (0.89-1.18) | 0.75 | 0.98 (0.81-1.18) | 0.38 | 1.03 (0.82-1.30) | 0.77 |
|
| ||||||
| Ischemic Heart Disease | n = 928 | n = 542 | n = 386 | |||
|
| ||||||
| Gas | 0.94 (0.87-1.01) | 0.09 | 0.97 (0.88-1.06) | 0.49 | 0.91 (0.82-1.02) | 0.09 |
| Kerosene/Diesel | 1.14 (1.06-1.21) | < 0.001 | 1.13 (1.04-1.24) | 0.006 | 1.13 (1.02-1.25) | 0.02 |
| Wood | 1.04 (0.97-1.12) | 0.28 | 1.01 (0.92-1.11) | 0.85 | 1.07 (0.96-1.19) | 0.22 |
| Pehen | 1.03 (0.97-1.09) | 0.29 | 1.04 (0.96-1.12) | 0.38 | 1.01 (0.92-1.11) | 0.82 |
|
| ||||||
| Cerebrovascular Accident | n = 519 | n = 256 | n = 263 | |||
|
| ||||||
| Gas | 0.92 (0.83-1.00) | 0.06 | 0.93 (0.81-1.06) | 0.29 | 0.91 (0.80-1.04) | 0.16 |
| Kerosene/Diesel | 1.08 (0.99-1.17) | 0.08 | 1.05 (0.92-1.19) | 0.47 | 1.10 (0.98-1.24) | 0.10 |
| Wood | 1.01 (0.92-1.11) | 0.85 | 1.01 (0.88-1.15) | 0.90 | 1.01 (0.89-1.15) | 0.90 |
| Pehen | 1.01 (0.94-1.09) | 0.73 | 1.02 (0.92-1.13) | 0.71 | 1.00 (0.90-1.11) | 0.98 |
adjusted for age at enrollment, sex, rural living, socioeconomic status, education level, Turkmen ancestry, physical activity, body mass index, tobacco use, alcohol use, opiate use, self-reported diabetes mellitus, hypertension, and ischemic heart disease, and total lifetime use of other fuels
adjusted for age at enrollment, rural living, socioeconomic status, education level, Turkmen ancestry, physical activity, body mass index, tobacco use, alcohol use, opiate use, self-reported diabetes mellitus, hypertension, and ischemic heart disease, and total lifetime use of other fuels
Regarding cardiovascular mortality, kerosene/diesel burning was associated with an increased risk of death (10-year HR 1.11, 95% CI 1.06-1.17) whereas natural gas burning was associated with a reduced risk for death (10-year HR 0.94, 95% CI 0.89-0.99). Subtype-specific analysis of cardiovascular-related deaths yielded similar mortality risks for ischemic heart disease mortality for kerosene/diesel use (10-year HR 1.14, 95% CI 1.06-1.21), while the relative protective role for natural gas (10-year HR 0.94, 95% CI 0.87-1.01) only trended toward significance. For death due to cerebrovascular accident, the HRs for kerosene/diesel burning and natural gas burning trended toward statistical significance (10-year HR 1.08, 95% CI 0.99-1.17; HR 0.92, 95% CI 0.83-1.00, respectively). Pehen and wood use were not statistically significantly associated with all-cause or subtype-specific mortality.
Sex specific analyses yielded a potential signal for greater risk of kerosene/diesel burning on all-cause and total CVD mortality in women versus men (10-year HR 1.09, 95% CI 1.03-1.15, versus 1.04 (0.99-1.09); HR 1.14, 95% CI 1.06-1.22, versus 1.08 (1.01-1.16), respectively (Table 2). Significant associations for ischemic heart disease mortality kerosene/diesel burning were similar among men and women. The 10-year HR for stroke mortality due to kerosene/diesel burning, although not statistically significant, was higher for women than men. The benefits of gasoline burning, relative to other fuels, on all-cause and total CVD mortality were only found among women. Mediation and effect modification analyses for age (Supplemental Table 3), obesity (Supplemental Table 4), diabetes mellitus (Supplemental Table 5), hypertension (Supplemental Table 6), removal of comorbidity variables (Supplemental Table 7), physical activity level (Supplemental Table 8), smoking (Supplemental Table 9), opium consumption (Supplemental Table 10) and alcohol consumption (Supplemental Table 11) were also performed. There was no clear evidence for either mediation or effect modification, but the number of mortality events was small for many of these analyses and therefore precluded a definitive conclusion. Further analysis, including adjusting for cigarette pack-years in lieu of mere history of tobacco use revealed no significant difference from the base model (Supplemental Table 12) and no evidence for mediation or effect modification among men (Supplemental Table 13).
Incremental 10-year exposure to kerosene/diesel burning was associated with increasing all-cause, total CVD, ischemic heart disease and stroke related mortality (p for trend ≤ 0.022 for all, Figure 1). Traditional cardiovascular risk factors—including hypertension, tobacco, and diabetes—were also associated with increased all-cause and cardiovascular mortality (data not shown).
Figure 1. Cox proportional hazards models for increasing levels of kerosene/diesel exposure.
Hazard ratios are on a logarithmic scale for all-cause, total CVD, ischemic heart disease, and cerebrovascular accident related mortality by increasing lifetime exposure to kerosene/diesel burning (p for trend ≤ 0.022 for all). Reference groups include those with 0 years of exposure. Cox proportional hazards models were adjusted for age at enrollment, sex, rural living, socioeconomic status, education level, Turkmen ancestry, physical activity, body mass index, tobacco use, alcohol use, opiate use, self-reported diabetes mellitus, hypertension, and ischemic heart disease, and total lifetime use of other fuels.
Discussion
In summary, we found an increased risk for all-cause and CVD mortality (largely driven by ischemic heart disease) with kerosene/diesel burning, while there was a decreased risk for all-cause and CVD mortality with natural gas burning, relative to other household fuels used for cooking and heating. Similar trends were observed for death due to stroke. There is a potential signal for greater risk of all-cause and CVD mortality due to kerosene/diesel burning among women versus men, possibly driven by excess stroke related mortality, and could reflect increased daily household fuel burning exposure times; however this possible sex effect modification was not robust and definitive conclusions cannot be made as the model was underpowered to detect such differences. Per 25-year period of kerosene/diesel burning, the relative increases in all-cause (15%), total CVD (25%), and ischemic heart disease mortality (30%) are quite large.
Tobacco use and ambient air pollution are known risk factors for cardiovascular mortality. Increases in exposure to fine particulate matter ≤ 2.5μm (PM2.5) in aerodynamic diameter from ambient air pollution and equivalent estimates of inhaled PM2.5 due to cigarette smoking are associated with increasing relative risk of cardiovascular and cardiopulmonary mortality in adulthood.16 A consistent cardiovascular mortality benefit is observed among residents in relatively clean cities with less exposure to PM2.5 versus passive smokers, as well as light smokers versus heavy smokers.16 This is the first study to our knowledge that prospectively shows a significant, multivariable-adjusted increased risk for all-cause, total CVD and ischemic heart disease mortality from increasing lifetime exposures to household air pollution from kerosene/diesel burning. It bridges a gap in medical literature, as previous studies reporting such a link have largely relied on modeling.23,24
Kerosene and diesel are chemical mixtures produced from the distillation of crude oil. Various mechanisms may contribute to the association between kerosene/diesel burning and cardiovascular mortality. One animal model replicating HAP from kerosene burning implicated increased atherosclerosis based on aortic plaque induction found on autopsy among guinea pigs exposed to high levels of kerosene exhaust versus saline aerosols.25 Additionally, Mills and colleagues reported acutely attenuated forearm blood flow among healthy individuals exposed to diesel exhaust despite concurrent administration of vasodilators including acetylcholine, bradykinin, sodium nitroprusside and verapamil, consistent with decreased vasomotor function leading to ischemia.26 Impaired endogenous fibrinolysis, due to decreased tissue plasminogen activator release, was also shown in this study and a follow-up study in men with a history of ischemic heart disease.26,27
The mechanisms relating HAP and CVD events may result from inhaled toxicants stimulating systemic inflammation, oxidative stress, and the autonomic nervous system, leading to increased leukocyte epigenetic changes, coagulation, thrombosis, and vasoconstriction.28,29 At a minimum, increased PM2.5 exposures from cooking are associated with CVD risk factors including increased systolic and diastolic blood pressure, carotid intima-medial thickness, carotid atherosclerotic plaque and reduced heart rate variability.7,12,30 Clean cook-stove interventions and air purifiers have led to measured reduction in exposures to fine particulate matter with corresponding improvements in circulating inflammatory markers, systolic and diastolic blood pressure and the magnitude of ST segment depression.8,9,31,32 Due to our study design, we are unable to determine the exact mechanism by which diesel/kerosene was associated with cardiovascular mortality. It is possible that the repeated exposures could have led to recurrent acute ischemia resulting in chronic, adverse changes to the vascular endothelium. It is also possible that the fuel exposure could have a triggering effect on events. Nonetheless, such mechanistic determination merits further investigation.
Overall, the study population had longer exposures to natural gas and kerosene/diesel burning as opposed to wood and pehen burning. Compared to natural gas, kerosene/diesel and solid fuel burning are associated with higher indoor PM2.5 and polyaromatic hydrocarbon concentrations,33-35 which could contribute to downstream adverse health consequences including the observed increases in all-cause and CVD mortality seen in the present study. Insufficient exposure times to solid fuel burning in the present study could explain the lack of an observed effect on mortality.
Further areas of investigation include assessing how time since last exposure to various fuel sources impacts mortality. Additionally, an evaluation of ventilation methods attached to various stoves could be illuminating. These data can also be coupled to investigations using gravimetric PM2.5 measurement to better estimate the daily and seasonal volume of combustion-related household air pollution from various fuel sources as well as a competing risk analysis given increasing comorbid illness as the population ages.
The main strengths of this study include the 99% successful follow-up rate, the time for follow-up to track mortality, and the systematic method used to track attributable causes of death. In addition, we were able to gather information on lifetime exposure to various fuel sources for cooking and heating. Although the data on fuel exposure were collected by recall, the detailed questionnaire allowed the investigators to account for overlap periods of fuel exposure from cooking and heating. Furthermore, mortality trends were robust when exposures were considered as continuous or categorical variables.
A major limitation of our study is the lack of quantifiable exposures to particulate matter or hours spent per day exposed to fuel burning. Hence, years of exposure to various types of fuel burning were used as a surrogate for the likely amount of household particulate matter to which individuals in our study population were exposed over time. Information on seasonal variation in quantifiable fuel exposure and adjustments for various household and stove ventilation methods is currently lacking. As a result, there is potential misclassification of exposure to indoor air pollution given the use of recall to estimate duration of exposures. Nonetheless, in the absence of quantifiable PM2.5 data, information on household fuel burning abstracted from survey data has previously been used to establish associations between household air pollution and childhood mortality,36-38 therefore supporting our methodology. Also noteworthy is that people of lower socioeconomic status are likely to use inexpensive higher polluting fuels and be subject to other external, modifiable factors that contribute to a higher rate of death. Hence there is potential residual confounding by socioeconomic status, despite adjustment for several predefined socioeconomic indicators,20 and by various forms of household and stove ventilation methods. Other potential limitations include lack of information on family history of heart disease and a referent group for fuel exposures. While we were able to achieve a 99% follow-up rate of all study participants, and medical personal examined verbal autopsy questionnaires in addition to ancillary pathology information, there is a small possibility of misclassification of causes of death with ICD-10 coding.
The global burden of household air pollution is substantial. The WHO estimates that 3 billion people, mostly in LMICs, cook and heat their homes using solid fuels with open fires and inefficient stoves. Despite efforts to transition to cleaner burning fuels such as natural gas or to electricity, as many as 500 million households worldwide still use kerosene for lighting purposes whereas urbanization in LMICs is accompanied by an increase in kerosene use as opposed to coal and biomass burning for cooking purposes.39 Given the effects on cardiovascular morbidity and mortality, decreasing household air pollution is an undeniably vital target for non-communicable disease management.
These results support the stance that communities that transition away from high-pollution fuels to low-pollution fuels can experience a relative decrease in all-cause and CVD mortality. More studies should be conducted to replicate these findings. Furthermore, in combination with other results demonstrating improvements in CVD risk profile with interventions to reduce exposures to household air pollution, our results support the ongoing efforts to develop and implement alternative solutions to reduce exposure to household air pollution. Currently the Global Alliance for Clean Cookstoves launched by the United Nations Foundation aims to help 100 million impoverished homes adopt clean and efficient stoves and fuels by 2020.3 Successful implementation will require the involvement of local governments, industry, and non-governmental organizations to propel this effort forward and reduce the worldwide burden of cardiovascular morbidity and mortality attributable to household air pollution.
Supplementary Material
Clinical Perspective.
Household air pollution ranks amongst the highest risk factors for global disease burden. In this study of 50045 adults in the Golestan Cohort Study, we were able to demonstrate that increasing lifetime duration of kerosene and diesel fuel burning exposure is significantly associated with increased all-cause and total cardiovascular disease mortality. Subtype-specific analyses revealed significant associations with ischemic heart disease mortality with an adjusted 10-year hazard ratio of 1.14 (95% CI 1.06-1.21). This is the first study to our knowledge that prospectively finds a significant adjusted increased risk for all-cause and cardiovascular disease mortality due to household air pollution. It bridges a gap in medical literature between the impact of ambient air pollution and direct tobacco inhalation on cardiovascular death. To address the high burden of cardiovascular disease the world over it is important for practitioners to identify and act upon modifiable risk factors for cardiovascular disease mortality. More studies are needed to replicate these findings and investigate transitioning communities to low-pollution fuels to improve all-cause and cardiovascular disease mortality.
Acknowledgments
We thank Dr. Ann Mwangi of the Department of Behavioral Sciences of the Moi University School of Medicine, Eldoret, Kenya for statistical support.
Funding Sources: The Golestan Cohort Study was funded by Tehran University of Medical Sciences (grant No: 81/15, Tehran Iran), Cancer Research UK (grant No: C20/A5860, London, UK), the Intramural Research Program of the United States National Cancer Institute (Bethesda, MD, USA), and through various collaborative research agreements with the International Agency for Research on Cancer (Lyon, France). Dr. R. Vedanthan receives grant funding from the Fogarty International Center of the NIH (K01 TW 009218 – 05, Bethesda, MD, USA). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
Abbreviations
- HAP
household air pollution
- CVD
cardiovascular disease
- LMIC
low- and middle-income countries
- HR
hazard ratio
- BMI
body mass index
- WHO
world health organization
- PM2.5
particulate matter ≤ 2.5μm PM2.5
Footnotes
Disclosures: None.
First Author Surname/Brief Title: Mitter/ Household fuel use and CVD mortality
Subject Codes: Cardiovascular Disease, Cerebrovascular, Epidemiology, Mortality/Survival
References
- 1.Rehfuess E, Mehta S, Prüss-Ustün A. Assessing household solid fuel use: multiple implications for the Millennium Development Goals. Environ Health Perspect. 2006;114:373–378. doi: 10.1289/ehp.8603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Kuhn R, Rothman DS, Turner S, Solórzano J, Hughes B. Beyond Attributable Burden: Estimating the Avoidable Burden of Disease Associated with Household Air Pollution. PLoS ONE. 2016;11:e0149669. doi: 10.1371/journal.pone.0149669. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Martin WJ, Glass RI, Balbus JM, Collins FS. Public health. A major environmental cause of death. Science. 2011;334:180–181. doi: 10.1126/science.1213088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Alam DS, Chowdhury MAH, Siddiquee AT, Ahmed S, Hossain MD, Pervin S, Streatfield K, Cravioto A, Niessen LW. Adult Cardiopulmonary Mortality and Indoor Air Pollution. Glob Heart. 2012;7:215–221. doi: 10.1016/j.gheart.2012.06.008. [DOI] [PubMed] [Google Scholar]
- 5.Lee M-S, Hang J-Q, Zhang F-Y, Dai H-L, Su L, Christiani DC. In-home solid fuel use and cardiovascular disease: a cross-sectional analysis of the Shanghai Putuo study. Environ Health. 2012;11:18. doi: 10.1186/1476-069X-11-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Dutta A, Mukherjee B, Das D, Banerjee A, Ray MR. Hypertension with elevated levels of oxidized low-density lipoprotein and anticardiolipin antibody in the circulation of premenopausal Indian women chronically exposed to biomass smoke during cooking. Indoor Air. 2011;21:165–176. doi: 10.1111/j.1600-0668.2010.00694.x. [DOI] [PubMed] [Google Scholar]
- 7.Baumgartner J, Schauer JJ, Ezzati M, Lu L, Cheng C, Patz JA, Bautista LE. Indoor air pollution and blood pressure in adult women living in rural China. Environ Health Perspect. 2011;119:1390–1395. doi: 10.1289/ehp.1003371. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.McCracken JP, Smith KR, Díaz A, Mittleman MA, Schwartz J. Chimney stove intervention to reduce long-term wood smoke exposure lowers blood pressure among Guatemalan women. Environ Health Perspect. 2007;115:996–1001. doi: 10.1289/ehp.9888. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.McCracken J, Smith KR, Stone P, Díaz A, Arana B, Schwartz J. Intervention to lower household wood smoke exposure in Guatemala reduces ST-segment depression on electrocardiograms. Environ Health Perspect. 2011;119:1562–1568. doi: 10.1289/ehp.1002834. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kargin R, Kargin F, Mutlu H, Emiroglu Y, Pala S, Akcakoyun M, Aung SM, Baran R, Ozdemir N. Long-term exposure to biomass fuel and its relation to systolic and diastolic biventricular performance in addition to obstructive and restrictive lung diseases. Echocardiography. 2011;28:52–61. doi: 10.1111/j.1540-8175.2010.01278.x. [DOI] [PubMed] [Google Scholar]
- 11.Bloomfield GS, Lagat DK, Akwanalo OC, Carter EJ, Lugogo N, Vedanthan R, Velazquez EJ, Kimaiyo S, Sherman CB. Waiting to inhale: An exploratory review of conditions that may predispose to pulmonary hypertension and right heart failure in persons exposed to household air pollution in low- and middle-income countries. Glob Heart. 2012;7:249–259. doi: 10.1016/j.gheart.2012.06.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Painschab MS, Davila-Roman VG, Gilman RH, Vasquez-Villar AD, Pollard SL, Wise RA, Miranda JJ, Checkley W, CRONICAS Cohort Study Group Chronic exposure to biomass fuel is associated with increased carotid artery intima-media thickness and a higher prevalence of atherosclerotic plaque. Heart. 2013;99:984–991. doi: 10.1136/heartjnl-2012-303440. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Fatmi Z, Coggon D, Kazi A, Naeem I, Kadir MM, Sathiakumar N. Solid fuel use is a major risk factor for acute coronary syndromes among rural women: a matched case control study. Public Health. 2014;128:77–82. doi: 10.1016/j.puhe.2013.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Miller KA, Siscovick DS, Sheppard L, Shepherd K, Sullivan JH, Anderson GL, Kaufman JD. Long-term exposure to air pollution and incidence of cardiovascular events in women. New Engl J Med. 2007;356:447–458. doi: 10.1056/NEJMoa054409. [DOI] [PubMed] [Google Scholar]
- 15.Pope CA, Burnett RT, Krewski D, Jerrett M, Shi Y, Calle EE, Thun MJ. Cardiovascular mortality and exposure to airborne fine particulate matter and cigarette smoke: shape of the exposure-response relationship. Circulation. 2009;120:941–948. doi: 10.1161/CIRCULATIONAHA.109.857888. [DOI] [PubMed] [Google Scholar]
- 16.Smith KR, Peel JL. Mind the gap. Environ Health Perspect. 2010;118:1643–1645. doi: 10.1289/ehp.1002517. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Berry JD, Dyer A, Cai X, Garside DB, Ning H, Thomas A, Greenland P, Van Horn L, Tracy RP, Lloyd-Jones DM. Lifetime risks of cardiovascular disease. New Engl J Med. 2012;366:321–329. doi: 10.1056/NEJMoa1012848. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Bonjour S, Adair-Rohani H, Wolf J, Bruce NG, Mehta S, Prüss-Ustün A, Lahiff M, Rehfuess EA, Mishra V, Smith KR. Solid fuel use for household cooking: country and regional estimates for 1980-2010. Environ Health Perspect. 2013;121:784–790. doi: 10.1289/ehp.1205987. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Pourshams A, Khademi H, Malekshah AF, Islami F, Nouraei M, Sadjadi AR, Jafari E, Rakhshani N, Salahi R, Semnani S, Kamangar F, Abnet CC, Ponder B, Day N, Dawsey SM, Boffetta P, Malekzadeh R. Cohort Profile: The Golestan Cohort Study--a prospective study of oesophageal cancer in northern Iran. Int J Epidemiol. 2010;39:52–59. doi: 10.1093/ije/dyp161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Islami F, Kamangar F, Nasrollahzadeh D, Aghcheli K, Sotoudeh M, Abedi-Ardekani B, Merat S, Nasseri-Moghaddam S, Semnani S, Sepehr A, Wakefield J, Møller H, Abnet CC, Dawsey SM, Boffetta P, Malekzadeh R. Socio-economic status and oesophageal cancer: results from a population-based case-control study in a high-risk area. Int J Epidemiol. 2009;38:978–988. doi: 10.1093/ije/dyp195. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Khademi H, Malekzadeh R, Pourshams A, Jafari E, Salahi R, Semnani S, Abaie B, Islami F, Nasseri-Moghaddam S, Etemadi A, Byrnes G, Abnet CC, Dawsey SM, Day NE, Pharoah PD, Boffetta P, Brennan P, Kamangar F. Opium use and mortality in Golestan Cohort Study: prospective cohort study of 50,000 adults in Iran. BMJ. 2012;344:e2502–e2502. doi: 10.1136/bmj.e2502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Khademi H, Etemadi A, Kamangar F, Nouraie M, Shakeri R, Abaie B, Pourshams A, Bagheri M, Hooshyar A, Islami F, Abnet CC, Pharoah P, Brennan P, Boffetta P, Dawsey SM, Malekzadeh R. Verbal autopsy: reliability and validity estimates for causes of death in the Golestan Cohort Study in Iran. PLoS ONE. 2010;5:e11183. doi: 10.1371/journal.pone.0011183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Galea KS, Hurley JF, Cowie H, Shafrir AL, Sánchez Jiménez A, Semple S, Ayres JG, Coggins M. Using PM2.5 concentrations to estimate the health burden from solid fuel combustion, with application to Irish and Scottish homes. Environ Health. 2013;12:50. doi: 10.1186/1476-069X-12-50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Mestl HES, Aunan K, Seip HM. Health benefits from reducing indoor air pollution from household solid fuel use in China--three abatement scenarios. Environ Int. 2007;33:831–840. doi: 10.1016/j.envint.2007.03.012. [DOI] [PubMed] [Google Scholar]
- 25.Noa M, Illnait J. Induction of aortic plaques in guinea pigs by exposure to kerosene. Arch Environ Health. 1987;42:31–36. doi: 10.1080/00039896.1987.9935792. [DOI] [PubMed] [Google Scholar]
- 26.Mills NL, Törnqvist H, Robinson SD, Gonzalez M, Darnley K, Macnee W, Boon NA, Donaldson K, Blomberg A, Sandström T, Newby DE. Diesel exhaust inhalation causes vascular dysfunction and impaired endogenous fibrinolysis. Circulation. 2005;112:3930–3936. doi: 10.1161/CIRCULATIONAHA.105.588962. [DOI] [PubMed] [Google Scholar]
- 27.Mills NL, Törnqvist H, Gonzalez MC, Vink E, Robinson SD, Söderberg S, Boon NA, Donaldson K, Sandström T, Blomberg A, Newby DE. Ischemic and thrombotic effects of dilute diesel-exhaust inhalation in men with coronary heart disease. New Engl J Med. 2007;357:1075–1082. doi: 10.1056/NEJMoa066314. [DOI] [PubMed] [Google Scholar]
- 28.Brook RD, Rajagopalan S, Pope CA, Brook JR, Bhatnagar A, Diez Roux AV, Holguin F, Hong Y, Luepker RV, Mittleman MA, Peters A, Siscovick D, Smith SC, Whitsel L, Kaufman JD, American Heart Association Council on Epidemiology and Prevention, Council on the Kidney in Cardiovascular Disease, and Council on Nutrition, Physical Activity and Metabolism Particulate matter air pollution and cardiovascular disease: An update to the scientific statement from the American Heart Association. Circulation. 2010;121:2331–2378. doi: 10.1161/CIR.0b013e3181dbece1. [DOI] [PubMed] [Google Scholar]
- 29.Chin MT. Basic mechanisms for adverse cardiovascular events associated with air pollution. Heart. 2015;101:253–256. doi: 10.1136/heartjnl-2014-306379. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Unosson J, Blomberg A, Sandström T, Muala A, Boman C, Nyström R, Westerholm R, Mills NL, Newby DE, Langrish JP, Bosson JA. Exposure to wood smoke increases arterial stiffness and decreases heart rate variability in humans. Part Fibre Toxicol. 2013;10:20. doi: 10.1186/1743-8977-10-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Clark ML, Bachand AM, Heiderscheidt JM, Yoder SA, Luna B, Volckens J, Koehler KA, Conway S, Reynolds SJ, Peel JL. Impact of a cleaner-burning cookstove intervention on blood pressure in Nicaraguan women. Indoor Air. 2013;23:105–114. doi: 10.1111/ina.12003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Chen R, Zhao A, Chen H, Zhao Z, Cai J, Wang C, Yang C, Li H, Xu X, Ha S, Li T, Kan H. Cardiopulmonary benefits of reducing indoor particles of outdoor origin: a randomized, double-blind crossover trial of air purifiers. J Am Coll Cardiol. 2015;65:2279–2287. doi: 10.1016/j.jacc.2015.03.553. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Watson JG, Chow JC, Chen L-WA, Lowenthal DH, Fujita EM, Kuhns HD, Sodeman DA, Campbell DE, Moosmüller H, Zhu D, Motallebi N. Particulate emission factors for mobile fossil fuel and biomass combustion sources. Sci Total Environ. 2011;409:2384–2396. doi: 10.1016/j.scitotenv.2011.02.041. [DOI] [PubMed] [Google Scholar]
- 34.Wu F, Wang W, Man YB, Chan CY, Liu W, Tao S, Wong MH. Levels of PM2.5/PM10 and associated metal(loid)s in rural households of Henan Province, China. Sci Total Environ. 2015;512-513:194–200. doi: 10.1016/j.scitotenv.2015.01.041. [DOI] [PubMed] [Google Scholar]
- 35.Li T, Cao S, Fan D, Zhang Y, Wang B, Zhao X, Leaderer BP, Shen G, Zhang Y, Duan X. Household concentrations and personal exposure of PM2.5 among urban residents using different cooking fuels. Sci Total Environ. 2016;548-549:6–12. doi: 10.1016/j.scitotenv.2016.01.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Bruce NG, Dherani MK, Das JK, Balakrishnan K, Adair-Rohani H, Bhutta ZA, Pope D. 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. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Ezeh OK, Agho KE, Dibley MJ, Hall JJ, Page AN. The effect of solid fuel use on childhood mortality in Nigeria: evidence from the 2013 cross-sectional household survey. Environ Health. 2014;13:113. doi: 10.1186/1476-069X-13-113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Stevens GA, Dias RH, Ezzati M. The effects of 3 environmental risks on mortality disparities across Mexican communities. Proc Natl Acad Sci USA. 2008;105:16860–16865. doi: 10.1073/pnas.0808927105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Lam NL, Smith KR, Gauthier A, Bates MN. Kerosene: a review of household uses and their hazards in low- and middle-income countries. J Toxicol Environ Health B Crit Rev. 2012;15:396–432. doi: 10.1080/10937404.2012.710134. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.