You are here

Article Text

Article menu
Environment
Original article
Household solid fuel use and pulmonary function in an urban population in Shanghai, China
  1. Mi-Sun Lee1,
  2. Jing-qing Hang2,
  3. Feng-ying Zhang2,
  4. Bu-yong Zheng2,
  5. Li Su1,
  6. Yang Zhao1,
  7. He-lian Dai2,
  8. Hong-xi Zhang2,
  9. David C Christiani1,3
  1. 1Environmental and Occupational Medicine and Epidemiology Program, Department of Environmental Health, Harvard School of Public Health, Boston, Massachusetts, USA
  2. 2Shanghai Putuo District People's Hospital, Shanghai, China
  3. 3The Massachusetts General Hospital/Harvard Medical School, Boston, Massachusetts, USA
  1. Correspondence to Dr David C Christiani, Environmental and Occupational Medicine and Epidemiology Program, Department of Environmental Health, Harvard School of Public Health, 665 Huntington Ave, Building I Room 1401, Boston, MA 02115, USA; dchris{at}hsph.harvard.edu

Abstract

Objectives We examined the association between household solid fuel exposure and lung function in a densely populated district in urban Shanghai, China.

Methods Spirometry was performed in 12 506 subjects, aged 18 and over, residing in the Putuo District in Shanghai, China, in a cross-sectional survey. Exposure to solid fuel use at home was assessed by an administered questionnaire, estimating duration and total amount of solid fuel use at home during the lifetime.

Results After adjusting for confounders, the subjects with exposure to household solid fuel had a 1.3% (95% CI 0.57 to 2.02) decrease in forced expiratory volume in 1 s (FEV1) percent predicted and 3.5% (95% CI 2.74 to 4.18) decrease in forced vital capacity (FVC) percent predicted, respectively. Trends towards decreased pulmonary function measures were seen for longer duration and greater amount of household fuel use at home, in the highest compared with lowest tertile (p values for trend <0.001). We observed decrease in FEV1 and FVC percent predicted across increase in tertile of body mass index in association with in-home solid fuel exposure.

Conclusions This study suggests that in-home solid fuel exposure is associated with reduced lung function in an urban population.

https://doi.org/10.1136/oemed-2011-100569

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    What this paper adds

    • Few epidemiologic studies have examined the relation between household solid fuel use and lung function in an urban population, and none its potential effect modification by body mass index (BMI).

    • We found that exposure to household solid fuel is associated with reduced lung function, as indicated by forced expiratory volume in 1 s and forced vital capacity, in an urban population, and BMI modifies this association.

    • Our results underscore the importance of the control of exposure to solid fuel use at home and respiratory health surveillance, in particular, among obese subjects who may be more susceptible to in-home air pollution.

    Introduction

    The source of domestic fuels is changing rapidly, shifting from traditional to modern fuels, such as natural gas. However, more than 50% of the world's population still uses solid fuels (mainly biomass and coal) for home for cooking and heating, contributing greatly to the burden of disease.1–3

    Epidemiologic evidence suggests that solid fuel exposure is associated with respiratory diseases, such as acute respiratory infections (ARIs), chronic obstructive pulmonary disease and lung cancer.4–7 One of the possible underlying mechanisms includes direct oxidative stress, induced by combustion by-products, and indirectly via enhanced release of reactive oxygen species (ROS), and impaired ventilatory function.8 In addition to this imbalance of oxidative status, obesity, a central feature of the metabolic syndrome (MetS), can increase ROS production, since MetS is associated with elevated oxidative stress.9 The prevalence of Mets has increased in China significantly over recent decades with increasing body mass index (BMI).10 Although an association between indoor solid fuel exposure and impaired lung function is presumed, the interaction between household solid fuel exposure and BMI has not been examined.

    We therefore performed a large-scale population-based study to assess the relation between exposure to household solid fuel, estimated by total amount and duration of exposure throughout the subjects’ lives, and lung function. We also hypothesised that the effect of exposure to household solid fuel on lung function is greater among individuals with higher BMI. Modern China is facing an obesity epidemic, as are many other rapidly developing countries. At the same time, China remains the world's largest consumer and producer of coal (figure 1).11 ,12

    Figure 1
    Figure 1

    China's coal production and consumption, and the prevalence (%) of overweight and obese (body mass index≥25 kg/m2) in Chinese adults. Based on data from International Energy Annual (http://www.eia.gov), and the China Health and Nutrition Survey 1989–1997 (Du et al, 2002) and the 2002 China National Nutrition and Health Survey (Chinese Ministry of Public Health 2004, Wang et al, 2007). *Based on data from this study. This figure is only reproduced in colour in the online version.

    Methods

    Study design and data collection

    The Shanghai Putuo Cohort Study, a collaboration between the Harvard School of Public Health, and Shanghai Putuo District People's Hospital, began recruiting subjects from August 2007 to July 2009 from Shanghai Putuo District in China. Detailed information regarding our study population has been reported previously.13 Briefly, study subjects were recruited from the Shanghai Putuo District in China using random selection from census track data. Of the total 19 620 subjects, 15 183 provided written informed consent to participate in this study with the baseline clinical examination and questionnaire. Of these 15 183 subjects, 14 068 (71.7% of 19 630), aged 18 years and older, were included in this study. Of the 14 068 participants, 12 947 subjects had no missing data on spirometry. Finally, from the 12 947 subjects, individuals with missing covariates, including education, height, smoking, pack-years of smoking or second-hand smoke (n=441) were excluded. Thus, a total of 12 506 (96.6% of 12 947) subjects were included in the analysis of solid fuel use and pulmonary function measures. Of the 12 506 subjects, participants with missing data on duration (n=59) and total amount of solid fuel use (n=2509) were additionally excluded in the analysis. This left 12 447 subjects (96.1% of 12 947) for the final analysis of duration of solid fuel use, and 9997 individuals (77.2% of 12 947) for the final analysis of total amount and lifetime average amount of solid fuel use. Of the 12 506 participants, 9750 subjects have ever used solid fuel for cooking and/or heating, and 99% of 2756 non-users use natural gas as a cooking fuel. Participants reported being in jobs classified as white collar (41.4%), blue collar (7.0%), homemaker (1.9%), retiree (36.8%), others (12.5%) and missing (0.4%); 77.8% of participants are presently living in an apartment, or another type of house (22.1%) including 1–3-storied houses and missing (0.1%).

    All participants were interviewed in person by trained personnel, using structured questionnaires. The questionnaires include sociodemographic factors (age, gender, education, marital status and household income), smoking history (smoking status, pack-years of smoking and second-hand smoke), occupational history and household fuel exposures. Individual household exposures to solid fuels (coal and biomass) were assessed in the questionnaire: ever use (yes or no), duration, total amount and lifetime average amount. We did not assess current use or last use in the questionnaire, but were only able to assess ever exposed in the questionnaire. Currently, the Putuo District home fuel supply is natural gas, and no more biomass burning is allowed in the neighbourhood. The assessment of duration of exposure (number of years for cooking/heating using solid fuels, year) was based on ‘How many years have you used coal as heating?’ and ‘How many years have you used briquet, coal or firewood for cooking?’ The assessment of annual amount (kg/year) was asked as ‘How many kilograms per year of this fuel have you used for cooking and heating?’ Based on these, we calculated the total amount (as calculated by multiplying duration of exposure by the annual amount of fuel used (kg)) and the lifetime average amount of exposure (as calculated by multiplying duration by annual amount by dividing age, kg/year).13 These factors have been shown to influence the risk of lung cancer,14 ,15 but limited data are available with regard to their influence on pulmonary function in a population-based study. Anthropometric measurements, such as height, weight and waist circumference were taken during the physical examination. BMI was calculated by dividing body weight in kilograms by the square of height in metres. Waist circumference was measured in centimetres at the mid-distance between the top of the iliac crest and the bottom of rib cage. Trained personnel performed all measurements. The study protocol was approved by the institutional review boards of the Harvard School of Public Health and the Shanghai Putuo District People's Hospital.

    Spirometry

    Forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1) were measured by trained and supervised technicians using calibrated hand-held spirometers (Micro Plus, Micro Medical Ltd, Rochester, UK), which are portable and have shown to give good reproducibility,16 according to American Thoracic Society guidelines.17 All testing was done in a sitting position. Intrasubject variability was assessed by coefficient of variation, as calculated by SD/mean×100, of FVC<20%, a criteria used for reliability in a large multicenter trial.18 The largest FVC and FEV1 were recorded from at least three curves. FEV1 and FVC were expressed in litres, and also as the percentage of predicted values based on the prediction equations from the Chinese adult population.19 We computed correlation between the percentage of predicted values from the published prediction equation for a Chinese adult population,19 and the percentage of predicted values from a prediction equation from our sample. From this, we obtained coefficients of 0.97 for FEV1 percent predicted and 0.93 for FVC percent predicted.

    Statistical analysis

    All statistical analyses were performed using SAS V.9.2 (SAS Institute Inc, Carry, North Carolina, USA). The primary outcome measures were FEV1 percent predicted, FVC percent predicted, and FEV1/FVC ratio. Household exposure to solid fuel for cooking and heating was expressed as the use of solid fuels during a lifetime (ever users vs non-users), duration of exposure in years, total amount of solid fuel use in kilograms, and lifetime average amount of solid fuel use in kilograms per year. This is done by assessing the exposure on a continuous scale, and categorising each into tertiles based on each exposure distribution. For FEV1/FVC ratio, potential confounders, such as age (years), gender, height (cm), education (less than high school, high school, above college or higher), second-hand smoke, smoking status (never, former, current), and pack-years of smoking were considered. For percent predicted values of FEV1 and FVC, to avoid overadjustment, age and height were not included because the percent predicted lung function parameters were already adjusted for age and height. Demographic and clinical differences between solid fuel users and non-user groups were tested using the χ2 test and t test. Linear regressions were applied to estimate the adjusted regression coefficients (β), which represents a difference in the mean in solid fuel ever users compared with non-users, and the corresponding 95% CIs for the association between lung function measures and household solid fuel exposure.

    To assess the linear trend in associations, trend tests were performed by categorising the continuous exposure variables into tertiles, and treating ordinal scores as continuous in regression models. To assess effect modification by BMI, multiplicative interaction terms, along with the main effects, were included in regression models.

    Results

    The distribution of study population demographic characteristics stratified by household solid fuel use is shown in table 1. The study population consisted of 5847 men (47%) and 6659 women (53%), with mean age of 48 years. Overall 3479 (27.8%) were current or former smokers. The combined prevalence of overweight and obesity defined as BMI≥25 was 26%. Of the 12 506 subjects, 9750 (78%) participants were using solid fuels for cooking and heating in their home. Raw FEV1 and FVC values were 2.7 and 3.0 l, respectively. Statistical differences between solid fuel users and non-users were found in age, education, BMI, second-hand smoke and smoking status. Lung function parameters (FEV1 and FVC) were significantly lower in users compared with those in non-users (mean of 96.6 vs 99.9 for FEV1 percent predicted and 88.2 vs 94.6 for FVC percent predicted, respectively, p<0.001).

    View this table:
    Table 1

    General characteristics of the study population, stratified by solid fuel use, n (%) or mean±SD

    The adjusted associations of household solid fuel exposure with FEV1 percent predicted, FVC percent predicted and FEV1/FVC ratio are shown in table 2. After adjusting for confounders, the subjects with exposure to household solid fuel had a 1.29% (95% CI 0.57 to 2.02) decrease in FEV1 percent predicted and 3.46% (95% CI 2.74 to 4.18) decrease in FVC percent predicted, respectively, whereas no association was seen with FEV1/FVC ratio. Compared with individuals in the lowest tertile of the duration, total amount and lifetime average amount of solid fuel exposure, the highest tertile exposure group was associated with significant decreases in percent predicted FEV1 and FVC, respectively (p value for trend <0.001).

    View this table:
    Table 2

    Adjusted estimates (95% CIs) for percent predicted FEV1 and FVC, and FEV1/FVC ratio associated with household solid fuel exposures

    In a sensitivity analysis, we modelled the missing data for smoking as a separate category to check whether missing data in smoking caused bias in our observed findings. The sensitivity analysis showed similar results as our main model: solid fuel ever users had a 1.29% (95% CI 0.58 to 2.00) decrease in FEV1 percent predicted, and 3.43% (95% CI 2.72 to 4.14) decrease in FVC percent predicted. Therefore, it is unlikely that our observed findings are due to possible bias by lack of data in smoking. In another sensitivity analyses that restricted data to only among the never-smoking population, we found an association between in-home solid fuel exposure and lung function: in-home solid users had a 0.73% (95% CI −0.07 to 1.54, p=0.07) decrease in FEV1 percent predicted, and 3.1% (95% CI 2.29 to 3.92) decrease in FVC percent predicted. Compared with those in the lowest tertile of duration and total amount of exposure, participants in the highest tertile had greater decrease in FEV1 percent predicted and FVC percent predicted among never smokers after adjusting for covariates (p values for trend <0.001) (Appendix). For further sensitivity analysis, we additionally adjusted for BMI and found similar results: in-home solid fuel ever users had a 1.17% (95% CI 0.44 to 1.90) decrease in FEV1 percent predicted, and a 3.22% (95% CI 2.50 to 3.94) decrease in FVC percent predicted, respectively. When we performed separate sensitivity analysis with additionally adjusting for waist circumference, the number of subjects was reduced to 12 502 due to missing waist circumference (n=4), and the results showed similar results as our main model: in-home solid fuel ever users had a 1.20% (95% CI 0.47 to 1.92) decrease in FEV1 percent predicted, and 3.18% decrease (95% CI 2.46 to 3.90) in FVC percent predicted.

    We observed significant effect modification by BMI in the association between solid fuel exposure and lung function parameters, after adjusting for confounders (figure 2). Household fuel exposure was negatively associated with FEV1 percent predicted and FVC percent predicted among subjects with the highest tertile of BMI (p value for interaction is 0.004 for FEV1 percent predicted, and is 0.003 for FVC percent predicted). We also assessed effect modification by BMI, as treated in four groups according to international classification by WHO (<18.50 (underweight), 18.50–24.90 (normal), 25.00–29.90 (overweight), and 30≤ (obese)). We observed borderline significant effect modification by BMI in the association between solid fuel exposure and adjusted lung function parameters. Household fuel exposure was negatively associated with FEV1 percent predicted and FVC percent predicted, respectively, among obese subjects with BMI≥30 (p value for interaction is 0.068 for FEV1 percent predicted, and is 0.080 for FVC percent predicted).

    Figure 2
    Figure 2

    Adjusted estimates (95% CI) for FEV1 percent predicted and forced vital capacity percent predicted associated with household solid fuel exposure by body mass index (tertile, kg/m2). Each model was adjusted for education, second-hand smoke, smoking status, and pack-years of smoking. The regression coefficients (β) are indicated by the symbol and their 95% CIs by the solid lines. *p<0.001. This figure is only reproduced in colour in the online version.

    Discussion

    In this large-scale population-based study, household solid fuel exposure was associated with reduced FEV1 and FVC. We also observed larger decrements in lung function in association with a greater amount and longer duration of exposure. Furthermore, the associations of solid fuel with lung function measures were stronger among subjects with higher BMI values than among non-obese subjects. These data suggest that obese people may experience a larger detrimental effect from household exposure to solid fuel than do non-obese people.

    The mechanism for potential toxic effects of household solid fuels on lung function is still not completely understood, but one of the main mechanisms includes oxidative stress pathways. The burning of solid fuels emits multiple toxicants in gaseous and particle phases. Of specific concerns are household solid fuel-derived particulates, including polycyclic aromatic hydrocarbons (PAHs) and metals, which are capable of promoting ROS and lung inflammation. It has been suggested that oxidative stress, as measured using a lipid peroxidation indicator, is associated with impaired lung function in the general population.8 ,20 We could find no prior studies that examined the modifying effects of BMI as a key trait of MetS in relation to in-home solid fuel exposure and pulmonary responses in the general population. Although the underlying mechanism is not clear, obesity may increase one's susceptibility to the adverse effects of particulate matter (PM), a major by-product from combustion of solid fuels, since lipophilic substances, such as PAHs, accumulate in adipose tissue. Recent animal study in rats showed that exposure to PM2.5 enhanced insulin resistance in rats fed a high-fat diet.21 The magnitude of decline in lung function in response to short-term PM exposure was different; two to five times higher among obese children than those of normal weight.22 In healthy adults, BMI was positively associated with exhaled nitric oxide, a marker of pulmonary inflammation.23 More recent in vivo study shows that some toxic PAHs, such as benzo(a)pyrene, potentiate high-fat diet effects on inflammation,24 a key feature of metabolic disorders, including obesity and diabetes.25 Proinflammatory cytokines, such as tumour necrosis factor-α (TNF-α) and other inflammatory mediators are overexpressed in the adipose tissue of experimental obese mouse models and in humans, suggesting a clear link in the development of chronic inflammation and MetS.26 Alternatively, it may also be possible that high ventilation in obese people may enhance the association between solid fuel exposure and lung function. In healthy children, BMI was associated with an increase in the estimated total lung dose of deposited fine particles, suggesting that obese individuals may be associated with increased risk from inhaled ambient particles.27

    Recent epidemiologic data suggests that household air pollution from biomass fuel has been associated with respiratory diseases, such as acute respiratory infections28 in Asian countries where solid fuels for cooking and heating are used extensively. In a study by Regalado et al (2006) conducted in the rural village of Solis, Mexico, they measured indoor concentrations of PM (PM10, particles with a diameter of 10 µm or less) in the kitchen when the cooking biomass stove was burning. They found significant decrease in FEV1 (81 ml) and in FVC (122 ml) for women with a peak PM10 concentrations higher than 2.6 mg/m3 compared with those with less than 2.6 mg/m3. However, little is known about other exposure indicators other than the type of fuel used at home, such as amount and duration of fuel use throughout lifetime, since these are important factors for characterising indoor exposures when comparing developing and developed countries.5 In addition, whether obesity, a leading public health crisis, may have mechanistic implications for the pulmonary effects of household air pollution remains unanswered. To address these gaps in existing research, our community-based large population study found evidence that the use of solid fuel in the home had a pronounced effect on lung function in the overweight subpopulation, implying a multiplier effect that makes obese individuals more susceptible to household air pollution effects.

    In our study, household solid fuel exposure indicators do not have significant impacts on the FEV1/FVC ratio. Because restrictive lung disease causes decrease in both FEV1 and FVC, the FEV1/FVC ratio can remain approximately normal (http://meded.ucsd.edu/isp/1998/asthma/html/spirexp.html). A recent longitudinal study among children aged 6–13 years living in Chinese cities reported that use of coal as a household fuel was associated with reduced growth rates of lung function, 33% and 39% of FEV1 and FVC, respectively,29 but they did not see any significant association with FEV1/FVC ratio as consistent with our study, suggesting that indoor air pollution from household solid fuel use may cause mainly a restrictive pattern. Further study is needed to confirm the possibility that in-home solid fuel exposures cause pulmonary restriction.

    Potential residual confounding by smoking may raise a concern. However, our sensitivity analyses performed in never smokers suggest that residual confounding by cigarette smoking may not explain the association between in-home exposure to solid fuel and lung function.

    Despite the significance of this study, the findings need to be interpreted with caution. First, due to the cross-sectional design, we cannot rule out a possibility of reverse causation. However, we found consistent results with solid fuel exposures (duration and total amount), so this should be unlikely. Future research with a longitudinal approach would be important in confirming the association observed in this study. Second, household fuel exposure data were obtained by questionnaire, without personal monitoring measurements, so we do not know the precise levels of household fuel exposures. Although we estimated the total amount and duration of solid fuel used throughout life in the questionnaire-based approach, more integrated approaches using qualitative and quantitative exposure assessments are needed for better assessing of an individual's in-home fuel exposure and, thereby, identify household solid fuel-derived specific contaminants that play an adverse role in pulmonary function. Third, for assessing duration, total amount and lifetime average amount of in-home solid fuel exposure, there is potential for measurement error due to recall based on a questionnaire approach, which may be particularly inaccurate in the older age group. Among solid fuel users, there was a positive correlation between age and duration of exposure (r=0.724, p<0.0001). When we analyse the association stratified by tertile of age (18–39, 40–55, and >55 years), significant decreases in percent-predicted values of lung function measures in association with duration of solid fuel exposure were found among participants of ages ≤40, ≤55 and >55 years, but no association was found among subjects with lowest tertile of age. Generally, non-differential exposure misclassification can bias results toward the null. However, we observed greater effects of duration of exposure on lung function among participants aged >55 years and older. Therefore, it is unlikely that misclassification of exposure induced bias considerably. Possible selection bias due to lack of data on lifetime average amount of solid fuel use may raise concern. To check this, we analysed the association between outcome and missing in lifetime average amount of solid fuel use among ever users. A decreased FEV1 percent predicted was found among the subjects with missing data in lifetime average amount of solid fuel use compared with the subjects without missing data, whereas an increase in FVC percent predicted was found among the subjects with missing data in lifetime average amount of solid fuel use compared with the subjects without missing data. We thus may not completely rule out possible selection bias due to lack of data on lifetime average amount of solid fuel use. An additional limitation is that our regression models may not fully adjust for potential confounders, such as occupational exposure, although we were able to control for some important confounders, such as age, education, smoking and pack-years of smoking.5 Possible effects of unmeasured confounders may need to be considered in future studies.

    In conclusion, our analysis provides evidence that household fuel use may induce reductions in lung function in adults, particularly among obese subjects who may be more susceptible to in-home air pollution.

    Acknowledgments

    The authors gratefully acknowledge the contribution of all participants and the research team in Shanghai, and the research assistance of Ms Marcia Chertok.

    References

    1. ↵
      1. Zhang J,
      2. Mauzerall DL,
      3. Zhu T,
      4. et al
      . Environmental health in China: progress towards clean air and safe water. Lancet 2010;375:1110–19.
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      1. Lin HH,
      2. Murray M,
      3. Cohen T,
      4. et al
      . Effects of smoking and solid-fuel use on COPD, lung cancer, and tuberculosis in China: a time-based, multiple risk factor, modelling study. Lancet 2008;372:1473–83.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      IARC. Household use of solid fuels and high-temperature frying/IARC Monographs on the Evaluation of Carcinogenic Risks to Humans. IARC 2010;95:45.
      OpenUrl
    4. ↵
      1. Tang L,
      2. Lim WY,
      3. Eng P,
      4. et al
      . Lung cancer in Chinese women: evidence for an interaction between tobacco smoking and exposure to inhalants in the indoor environment. Environ Health Perspect 2010;118:1257–60.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Hosgood HD,
      2. Boffetta P,
      3. Greenland S,
      4. et al
      . In-home coal and wood use and lung cancer risk: a pooled-analysis of the International Lung Cancer Consortium. Environ Health Perspect 2010;118:1743–7.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Hu G,
      2. Zhou Y,
      3. Tian J,
      4. et al
      . Risk of COPD from exposure to biomass smoke: a metaanalysis. Chest 2010;138:20–31.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Regalado J,
      2. Perez-Padilla R,
      3. Sansores R,
      4. et al
      . The effect of biomass burning on respiratory symptoms and lung function in rural Mexican women. Am J Respir Crit Care Med 2006;174:901–5.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Ochs-Balcom HM,
      2. Grant BJ,
      3. Muti P,
      4. et al
      . Oxidative stress and pulmonary function in the general population. Am J Epidemiol 2005;162:1137–45.
      OpenUrlAbstract/FREE Full Text
    9. ↵
      1. Hansel B,
      2. Giral P,
      3. Nobecourt E,
      4. et al
      . Metabolic syndrome is associated with elevated oxidative stress and dysfunctional dense high-density lipoprotein particles displaying impaired antioxidative activity. J Clin Endocrinol Metab 2004;89:4963–71.
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Gu D,
      2. Reynolds K,
      3. Wu X,
      4. et al
      . Prevalence of the metabolic syndrome and overweight among adults in China. Lancet 2005;365:1398–405.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Du S,
      2. Lu B,
      3. Zhai F,
      4. et al
      . A new stage of the nutrition transition in China. Public Health Nutr 2002;5:169–74.
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Wang Y,
      2. Mi J,
      3. Shan XY,
      4. et al
      . Is China facing an obesity epidemic and the consequences? The trends in obesity and chronic disease in China. Int J Obes (Lond). 2007;31:177–88.
      OpenUrlCrossRefPubMed
    13. ↵
      1. Lee MS,
      2. Hang JQ,
      3. Zhang FY,
      4. et al
      . In-home solid fuel use and cardiovascular disease: a cross-sectional analysis of the Shanghai Putuo study. Environ Health 2012;11:18.
      OpenUrlCrossRefPubMed
    14. ↵
      1. Hosgood HD 3rd.,
      2. Chapman R,
      3. Shen M,
      4. et al
      . Portable stove use is associated with lower lung cancer mortality risk in lifetime smoky coal users. Br J Cancer 2008;99:1934–9.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Lan Q,
      2. Chapman RS,
      3. Schreinemachers DM,
      4. et al
      . Household stove improvement and risk of lung cancer in Xuanwei, China. J Natl Cancer Inst 2002;94: 826–35.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Dirksen A,
      2. Madsen F,
      3. Pedersen OF,
      4. et al
      . Long-term performance of a hand held spirometer. Thorax 1996;51:973–6.
      OpenUrlAbstract/FREE Full Text
    17. ↵
      ATS. Standardization of Spirometry, 1994 Update. American Thoracic Society. Am J Respir Crit Care Med 1995;152:1107–36.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Ramsey BW,
      2. Dorkin HL,
      3. Eisenberg JD,
      4. et al
      . Efficacy of aerosolized tobramycin in patients with cystic fibrosis. N Engl J Med 1993;328:1740–6.
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Ip MS,
      2. Ko FW,
      3. Lau AC,
      4. et al
      . Updated spirometric reference values for adult Chinese in Hong Kong and implications on clinical utilization. Chest 2006;129:384–92.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Schunemann HJ,
      2. Muti P,
      3. Freudenheim JL,
      4. et al
      . Oxidative stress and lung function. Am J Epidemiol 1997;146:939–48.
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Yan YH,
      2. Chou CC,
      3. Lee CT,
      4. et al
      . Enhanced insulin resistance in diet-induced obese rats exposed to fine particles by instillation. Inhal Toxicol 2011;23: 507–19.
      OpenUrlCrossRefPubMed
    22. ↵
      1. Luttmann-Gibson H,
      2. Dockery DW
      . Short-term effects of air pollution on lung function: are obese children at higher risk? (Abstract). Am J Respir Crit Care Med 2004;169:A19.
      OpenUrlCrossRef
    23. ↵
      1. De Winter-de Groot KM,
      2. Van der Ent CK,
      3. Prins I,
      4. et al
      . Exhaled nitric oxide: the missing link between asthma and obesity? J Allergy Clin Immunol 2005;115:419–20.
      OpenUrlCrossRefPubMedWeb of Science
    24. ↵
      1. Khalil A,
      2. Villard PH,
      3. Dao MA,
      4. et al
      . Polycyclic aromatic hydrocarbons potentiate high-fat diet effects on intestinal inflammation. Toxicol Lett 2010;196:161–7.
      OpenUrlCrossRefPubMed
    25. ↵
      1. Hotamisligil GS
      . Inflammation and metabolic disorders. Nature 2006;444:860–7.
      OpenUrlCrossRefPubMedWeb of Science
    26. ↵
      1. Wellen KE,
      2. Hotamisligil GS
      . Inflammation, stress, and diabetes. J Clin Invest 2005;115:1111–19.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. Bennett WD,
      2. Zeman KL
      . Effect of body size on breathing pattern and fine-particle deposition in children. J Appl Physiol 2004;97:821–6.
      OpenUrlAbstract/FREE Full Text
    28. ↵
      1. Ezzati M,
      2. Kammen D
      . Indoor air pollution from biomass combustion and acute respiratory infections in Kenya: an exposure-response study. Lancet 2001;358:619–24.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Roy A,
      2. Chapman RS,
      3. Hu W,
      4. et al
      . Indoor air pollution and lung function growth among children in four Chinese cities. Indoor Air 2012;22:3–11.
      OpenUrlCrossRefPubMedWeb of Science

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

      Files in this Data Supplement:

    Footnotes

    • MSL, JQH contributed equally.

    • Contributors MSL and JQH participated in study design, statistical analysis and interpretation of the results, and wrote the manuscript. DCC originated study concept and design, advised analyses and interpretation of data, and revised the entire manuscript critically. FYZ, BYZ, LS, YZ, HLD and HXZ collected data, analysis and interpretation of data. All authors read and approved the final version of manuscript.

    • Funding This study was supported by grants from the National Institute of Environmental Health Science (Grant ES000002); and the National Institute for Occupational Safety and Health (Grant R010421); and Shanghai Putuo District People's Hospital.

    • Competing interests None.

    • Patient consent Obtained.

    • Ethics approval Institutional Review Boards of the Harvard School of Public Health and the Putuo District People's Hospital.

    • Provenance and peer review Not commissioned; externally peer reviewed.