In this study, subjects aged 20 years or older were selected through a multistage stratified cluster sampling method from Shanxi Province, China, between June 2012, and May 2015. Firstly, a large city, a midsize city, an economically developed county, and an underdeveloped county (based on average GDP) in Shanxi Province were selected randomly. Secondly, two urban districts from each city, two towns from the developed county, and two towns from the underdeveloped county were randomly chosen. Thirdly, we selected at random individuals aged 20 years or older from these communities. Only one person could be selected from one family without replacement. Lastly, we stratified these participants by age and gender according to the Chinses census data from 2010.10

Only permanent residents who had lived in Shanxi Province for more than one year could be included in this study. In addition, individuals were excluded if they were physically incapable of taking a spirometry test, such as patients with myocardial infarction in the past three months or heart rate of more than 120 beats per minute.

This study was approved by the Ethics Review Committee of Beijing Capital Medical University and other participating agencies, and all participants signed the informed consent.

We obtained survey data from local health centers, and then designed a standardized questionnaire for the participants,8 which included demographic data, medical history, family history of respiratory disease, and other risk factors. People who had smoked more than 100 cigarettes in their lifetime were defined as current smokers, and people who resided with smokers were defined as passive smokers.

Biomass utilization meant that the individuals cooked by wood or animal feces in the past six months. History of childhood pneumonia or bronchitis was defined as admission to hospital at least once before the age of 14, which was further divided into frequent (cumulative >3 months per year), occasional (1–3 months per year), and rare (<1 month per year).

History of chronic bronchitis referred to cough and sputum over at least three months within two years. The exposure to ambient particles with diameter less than 2.5 μm (PM2.5) was obtained by the aerosol optical depth model, which was recovered from the regional satellite.11 The PM 2.5 concentration model (2010) of each selected region was used as the air contamination exposure indicator for the subjects.

The pulmonary function tests were performed by certified technicians using the MasterScreen Pneumo PC spirometer (CareFusion, Yorba Linda, CA, USA), which was calibrated daily using a 3 L syringe. Subjects were asked to perform eight exhalations until 1 s of forced vital capacity (FVC) and forced expiratory volume (FEV1) could be restored within 150 mL.12 The bronchodilator (salbutamol 400 μg) was inhaled with same standard at 500 mL intervals, and then spirometry was performed after 20 min. All participants performed all spirometry measurements, including in a sitting posture, with nose clips, and disposable mouthpiece.

Data quality control was performed according to the American Thoracic Society and the European Respiratory Society standards.

Based on the Global Initiative for Chronic Obstructive Lung Disease (GOLD) standard (2017), the ratio of FEV1: FVC of less than 70% after bronchodilator use was defined as COPD.13 The ratio of FEV1: FVC and the obstruction level were predicted based on the general population references in the United States (GOLD 1, ≥80%, GOLD 2, ≥50% to <80%, GOLD 3, ≥30% ∼ <50%).14 The lower limit of normal (LLN) reference values in China were used to analyze the sensitivity of COPD,15 and COPD diagnoses were made by clinicians according to the self-reported spirometry results.

Based on the survey data, five age groups (20–39 years, 40–49 years, 50–59 years, 60–69 years, and ≥70 years) were designed for both men and women. The sample size was calculated using PASS software (NCSS, Kaysville, UT, USA). The weights and age-standardized prevalence were calculated based on 2010 China census population data. The ANOVA was used to assess the diversities between continuous variables and categorical variables with the χ2 test. A multivariable logistic regression analysis was performed to investigate risk factors for all participants and only nonsmokers, respectively.

All data analysis was conducted with SUDAAN (version 11.0; Research Triangle Institute, Research Triangle Park, NC, USA) and SAS (version 9.4; SAS Institute, Cary, NC, USA).

A total of 6400 individuals were invited to participate in our survey, of which 240 (3.75%) individuals declined to participate. After excluding unreliable results, 5636 cases (2468 men and 3168 women) were included for further analysis (Fig. 1).

Figure 1.

Demographic characteristics of participants.

(0.17MB).

According to the survey results, the prevalence of COPD was estimated to be 6.4% (95% CI 5.8–7.1) of the population (age ≥ 20) in Shanxi Province (Table 1 and Supplementary Table 1). The COPD prevalence was also calculated based on each characteristic at different GOLD stages. In male participants, the prevalence of COPD was 6.4%, 2.8%, and 0.4% at GOLD stages I, II, and III, while it was 2.6%, 1.1%, and 0.2% in female, respectively. The prevalence of COPD increased with age, and the prevalence of the 20−39 age group was 1.5% (95% CI 1.0–2.0), which increased to 9.3% (95% CI 8.4–10.3) in the group aged 40 years or older (p < 0.0001 for the age difference) (Tables 2 and 3).

More than 90% of the participants were urban residents, and the average age was 43.8 ± 14.0 for men and 46.9 ± 12.6 for women. Furthermore, compared to the less educated population (high school or lower), the prevalence of COPD was significantly lower in the higher education population (college or higher).

Smoking was also an important factor for COPD prevalence. The age-specific and age-standardized prevalence of spirometry-defined COPD was furthered calculated for the non-smoking group and the smoking group (Supplementary Table 2). The non-smoking and smoking populations showed differences in all age groups. A total of 253 subjects were current smokers (229 men and 24 women), and the prevalence of COPD calculated by age showed that the current smoking group (15.8%, 95% CI 11.3–20.3) and the former smoking group (10.6%, 95% CI 9.0–12.1) were significantly higher than the non-smoking group (4.2%, 95% CI 3.5–4.8, p < 0.0001). There was an obvious dose-response relationship between smoking years and the COPD prevalence (p < 0.0001). However, living with smokers was not related to an increase in COPD prevalence (p = 0.870).

Chronic cough in childhood (p < 0.0001) and parents with respiratory illness history (p < 0.0001) were related to an increase in COPD prevalence, and biomass utilization was also significantly associated with a rise in COPD prevalence in men (p < 0.0001).

Body mass index (BMI) was 25.16 ± 3.78 kg/m2 in men and 24.61 ± 3.78 kg/m2 in women, respectively. The BMI did not show any significant effect on COPD, and there was also no difference in COPD prevalence between rural (6.1%, 95% CI 3.9–8.4) and city (6.5%, 95% CI 5.8–7.1, p = 0.785) regions (Table 2) (Fig. 2A).

Figure 2.

COPD in population over 20 years old in Shanxi Province. A) Prevalence and absolute number of COPD in population over 20 years old in Shanxi Province. Bars represent percentage or average, and error bars represent 95% CI. The age-specific prevalence of COPD in men and women in urban regions (a) and rural regions (b). The number of COPD cases in men and women by age in urban regions (c) and in rural regions (d). B) Proportion of participants performed lung function test before or COPD cases in population over 20 years old in Shanxi Province. Bars represent proportions, and error bars represent 95% CI. Proportion with lung function test history in COPD cases, by gender and urbanization (a). Proportion with lung function test history in all participants, by gender and urbanization (b). Proportion with lung function test history in COPD cases, by age and urbanization (c). Proportion with lung function test history in all the participants, by age and urbanization (d).

(0.26MB).

Only 1‰ (5/5636) patients reported typical symptoms of COPD in their daily life, such as continuous cough and recurrent asthma. Nor were exposure to PM2.5 and tuberculosis history associated with COPD prevalence in the non-smoking and smoking groups.

The multiple-adjusted analysis showed that sex (p < 0.0001), age (p < 0.0001), and education level (p < 0.0001) were significantly related to the prevalence risk of COPD (Supplementary Table 3). The number of smoking years (p = 0.004) was positively correlated with the risk of COPD.

It should be noted that chronic cough in childhood (age ≤ 14 years, p < 0.0001) and parents with respiratory disease history (p = 0.006) were also significantly correlated with the prevalence of COPD in Shanxi. However, the other factors including rural residents, living with smokers, pneumonia or bronchitis history in childhood, BMI, biomass utilization, exposure to PM2.5, and parental tuberculosis history were not significantly correlated with COPD.

Among COPD patients identified by spirometry in our study, only five urban residents were previously aware that they had COPD disease (one man and four women). In addition, the percentage with lung examination history was higher in the city (Fig. 2B). Young COPD patients were more likely to undergo pulmonary testing.