Joint association of cigarette smoking and PM_2.5 with COPD among urban and rural adults in regional China

Chronic obstructive pulmonary disease (COPD) is a global public health problem, and the present prevalence of COPD was 11.7–15.8% worldwide [1‐3]. For China, the most populous country in the world, COPD has also caused heavy disease burden. The recently estimated prevalence of COPD, defined according to spirometry, was 8.6% among overall adults in China, and this figure was even as high as 13.7% in those aged 40 + years [4]. Moreover, COPD might account for 697.63 million years lived with disability and nearly one million deaths every year in China [4, 5]. Thus, it is a public health priority to reduce the disease burden caused by COPD through population-based intervention programs in China.

Identifying specific modifiable risk factors is critically important for tailored COPD prevention at population level. Cigarette smoking and air pollution are two of such risk factors of COPD [6]. Cigarette smoking has been examined to be significantly associated with COPD either measured as smoking versus non-smoking or assessed based on the number of cigarettes smoked [4, 7, 8]. And it was further documented that cigarette smoking could account for 80–90% of COPD cases [8, 9]. For another modifiable risk factor of COPD, outdoor air pollution was commonly indicated with concentration of particulate matter with a diameter less than 2.5 μm (PM_2.5) [2, 4]. Moreover, a cutoff of 75 μg/m³ was usually employed to classify PM_2.5 concentration as “low” versus “high” level in population-based studies regarding association of air pollution (indicated with PM_2.5) with COPD, showing that PM_2.5 concentration was in significantly positive relation to COPD [4, 10].

The individual link between cigarette smoking, PM_2.5 and COPD has been investigated, but the joint association of cigarette smoking and PM_2.5 with COPD was not explored yet. Identifying the potential joint association would be of help for developing risk factor-specific intervention strategies against COPD. To bridge this gap, we conducted a population-based study to examine the combined association of cigarette smoking and PM_2.5 with COPD among adults in regional China, with a hypothesis that cigarette smoking and PM_2.5 might exert additive effect on COPD.

Initially, 3600 eligible participants were recruited, and 3407 (94.6%) completed both the questionnaire survey and spirometry. No significant differences were examined between respondents and non-respondents in terms of age, gender, education, occupation or body weight status. Table 1 displays selected personal characteristics of participants by smoking status in this study. For overall participants, their mean (standard deviation) age was 57.2 (9.9) years; 41.5% were elders (aged 60+ years); 45.8% were men; and 49.1% resided in urban areas. There was no difference in each of these three main socio-demographic characteristics (age, gender and residential location) between our sample population and the standard population of Jiangsu Province in 2015 [18]. Moreover, 33.5% lived in a area with PM_2.5 ≥ 75 μg/m³; 26.1% were biomass fuel users; 31.9% subjects had positive parental history of chronic respiratory diseases; and 17.9% were obese adults.

Table 2 shows the prevalence of COPD among participants by smoking and PM_2.5 exposure status. The COPD prevalence was 11.9% (95% CI 10.9%, 13.0%) among overall participants, while 6.4% (95% CI 5.3%, 7.4%), 20.6% (95% CI 18.4%, 22.8%), 10.9% (95% CI 9.6%, 12.1%), and 14.1% (95% CI 12.1%, 16.1%) for non-smokers, smokers, those living areas with PM_2.5 < 75 μg/m³ and PM_2.5 ≥ 75 μg/m³, respectively. The prevalence of COPD tended to be higher among men, household biomass fuel users, and those had positive parental history of chronic respiratory diseases. Moreover, COPD prevalence became higher as participants aged, but became lower with their educational attainment or BMI increased. There was no difference in COPD prevalence between urban and rural subjects, or blue and white collars.

Table 3 presents individual and joint associations of cigarette smoking and PM_2.5 with COPD among participants. Among overall participants, after adjustment for age, gender, residence, education, occupation, biomass use, parental history of respiratory diseases, body weight status, PM_2.5/cigarette smoking and potential clustering effects at study area level, smokers were at 2.46 times odds to experience COPD than non-smokers (OR = 2.46, 95% CI 1.76, 3.43), while participants living in areas with PM_2.5 ≥ 75 μg/m³ were more likely to develop COPD relative to their counterparts living in areas with PM_2.5 < 75 μg/m³ (OR = 1.29, 95% CI 1.02, 1.64). Furthermore, with control for potential confounding factors, compared to non-smokers who exposed to < 75 μg/m³ PM_2.5, non-smokers who exposed to ≥ 75 μg/m³ PM_2.5 (OR = 1.10, 95% CI 0.74, 1.64), smokers who exposed to < 75 μg/m³ PM_2.5 (OR = 2.22, 95% CI 1.51, 3.27) and ≥ 75 μg/m³ PM_2.5 (OR = 3.14, 95% CI 2.15, 4.59) were, respectively, more likely to develop COPD. Such an association was non-significant between non-smoker who exposed to < 75 μg/m³ PM_2.5 and ≥ 75 μg/m³ PM_2.5. However, for those smokers who exposed to ≥ 75 μg/m³ PM_2.5 were at 1.36 (95% CI 1.01, 1.83) times likelihood to experience COPD relative to their counterparts who exposed to < 75 μg/m³ PM_2.5.

In this population-based study, we mainly aimed to explore the joint association of cigarette smoking and PM_2.5 (predicting outdoor air pollution) with COPD among representative urban and rural adults in regional China. It was observed that either of cigarette smoking or PM_2.5 concentration was positively associated with COPD, and, furthermore, they additively exerted positive effect on COPD, after control of potential confounding factors. These findings suggested that cigarette smoking and PM_2.5-based outdoor air pollution might be jointly used as indicators to identify people at high risk in population-based precision intervention campaigns against COPD in China.

The prevalence of spirometry-defined COPD was 11.9% (95% CI 10.9%, 13.0%) among overall participants in our study, 18.3% (95% CI 16.3%, 20.2%) among men and 6.6% (95% CI 5.5%, 7.7%) among women, which were similar to the national estimates among participants with the same age-group in eastern region of China [4, 5]. It has also been estimated that 36.2–40.2% of smokers self-reported being diagnosed with COPD and the population fraction of COPD attributed to cigarette smoking was 22.2% in China in 2015 [4, 5, 19, 20]. Consistent with findings documented in previous studies [4, 5, 19, 20], 20.6% of smokers were diagnosed as COPD patients, and about 67.6% of COPD patients were smokers in our study. Our study shows that COPD prevalence was 14.1% (95% CI 12.1%, 16.1%) among those who resided in areas with PM_2.5 ≥ 75 μg/m³ in this study, which were higher than the figure (9.7%) estimated in a national survey conducted in the same year [4].

Cigarette smoking was the most well-studied and the most important modifiable risk factor of COPD. The independent and positive relationship was solidly established between cigarette smoking and COPD regardless of that cigarette smoking was assessed with categorical (smokers/ex-smokers or non-smokers) or continuous measure (number of cigarettes smoked) [4, 5, 7, 12]. Moreover, even if quitting smoking in time, a participant would continue to experience decline in lung function for years due to the lag-effect of inflammation caused by smoking [20]. In our study, as expected, cigarette smoking was examined to be significantly associated with COPD. After adjustment for potential confounding factors and community-level clustering effects, smokers had a 2.46 times likelihood to experience COPD relative to non-smokers. Such a likelihood for smokers to develop COPD in our study was greater than that for their counterparts in national survey conducted in the same year where the likelihood was less than 2.0 [4, 5].

Another important modifiable risk factor of COPD was outdoor and indoor air pollution [4, 21‐23]. With the rapid economic growth over past decades in China, residents might obtain more and more earning and income, and consequently clean energy and kitchen ventilators might have become easily affordable and been widely used for cooking/heating in households, leading to indoor air quality becoming better and better [24, 25]. Moreover, recent studies reported that domestic fuels used, kitchen ventilation and heating in winter was not in significant relation to COPD in China [12, 26]. Therefore, we had indoor air pollution adjusted for in the analysis and then paid particular attention to the relationship between outdoor air pollution (indicated with PM_2.5 concentration) and COPD in this study. Participants living in air-polluted areas with PM_2.5 ≥ 75 μg/m³ were at 1.29-folds odds for experiencing COPD compared to those living in areas with PM_2.5 < 75 μg/m³ [4], suggesting that air pollution was also in positive relation to COPD.

Cigarette smoking and outdoor air pollution could separately exert positive influence on COPD, and more interestingly and importantly an additive influence by these two risk factors on COPD was observed in this study. The odds for experiencing COPD was 3.14 for smokers who exposed to ≥ 75 μg/m³ PM_2.5, 2.22 for smokers living areas with PM_2.5 < 75 μg/m³ and 1.10 for non-smokers exposed to ≥ 75 μg/m³ PM_2.5, respectively, relative to non-smokers who exposed to < 75 μg/m³ PM_2.5, suggesting a positively gradient association of cigarette smoking and PM_2.5 with COPD in the study population. Interestingly and meaningfully, the likelihood for experiencing COPD between those exposed to high and low level of PM_2.5 was significant for smokers but not for non-smokers. It implied that PM_2.5 might exert influence on the risk of experiencing COPD for cigarette smokers only. This might be due to the relatively fewer COPD patients identified among non-smokers or some unknown underlying mechanisms. Well designed studies with sufficient sample size are in need to explore the PM2.5-COPD association among non-smokers in future.

With respect to the potential mechanisms behind this scenario, there are at least two main explanations. The first is that either cigarette smoking or PM_2.5 could stimulate some key oxidative and pro-inflammatory molecules in the process of COPD [22, 27, 28]. The second is that both of these two factors could also increase participants’ susceptibility to bacterial and/or viral infections [22, 29, 30]. Therefore, for cigarette smokers exposed to high level of PM_2.5, they would be more vulnerable to COPD due to such double impact exerted by smoking and PM_2.5.

This study has several strengths. First, all the COPD patients were identified using spirometry, an objective lung function assessment. Second, outdoor air pollution was also objectively assessed with PM_2.5 concentration. Third, participants were from urban and rural areas with representativeness of overall population. Fourth, classical confounding factors and clustering effects at study area level were controlled for in the analysis. Fifth, individual and joint association of cigarette smoking and PM_2.5 with COPD was separately investigated, showing these two modifiable factors might exert additive influence on COPD. Sixth, although collected in 2015, data used in this study were the most recently available population-based information regarding COPD in Jiangsu Province as well as China. Finally, the interesting findings from this study has important public health implications that both cigarette smoking and outdoor PM_2.5 should be jointly put into consideration for developing population-based precision prevention campaigns to fight COPD.

Regardless of the strengths, this study also has some limitations. First, because of the nature of cross-sectional study, the association examined in this study did not imply any causal direction. Second, due to lack of data, cigarette smoking, one of the two main explanatory variables, could not be used based on the number of cigarettes smoked, and thus the dose–response relationship between cigarettes smoked and CODP was not able to be assessed. Third, as only yearly mean values of PM_2.5 were available, we could examine the association between PM_2.5 and COPD using PM_2.5 as continuous or peak measures in our study. Fourth, the cutoff of PM_2.5 concentrations used to classify participants was 75 μg/m³ in this study. Although this cutoff was officially recommended by Ministry of Ecology and Environment of China, it was much higher than that recommended by the WHO or the U.S. Environmental Protection Agency. Therefore, when the association between PM_2.5 and COPD in this study was interpreted, the specific PM_2.5 cutoff should be put into consideration. Fifth, participants’ personal protective approaches against outdoor air pollution were not considered due to data limitation. Thus, the findings in this study should be interpreted prudently. In future, well-designed large-scale prospective population-based observational or even experimental studies are warranted to further investigate the joint association of cigarette smoking and PM_2.5 with COPD in China.

Cigarette smoking and PM_2.5 were individually in positive relation to COPD, and moreover they might exert additive influence on COPD among urban and rural adult smokers in regional China. This study has important public health significance that population-based precision COPD prevention campaigns should be tailored for specific participants with consideration of multiple risk factors.