Prevalence of non-smoking chronic obstructive pulmonary disease and its risk factors in China: a systematic review and meta-analysis

Abstract

Background

Chronic obstructive pulmonary disease (COPD) is a prevalent chronic disorder in China, impacting a significant proportion of individuals aged > 40 years. In China, the prevalence of and risk factors for COPD among non-smokers remain largely unexplored. In this study, we aimed to determine the prevalence of COPD in non-smokers within the Chinese population and identify potential risk factors associated with COPD in non-smokers.

Methods

Web of Science, PubMed, Embase, Chinese WanFang, Chinese China National Knowledge Infrastructure, and Weipu databases from inception to August 5, 2024, were searched. Studies reporting the percentage of never-smokers among those diagnosed with COPD and investigations exploring the risk factors associated with COPD in never-smokers in China were examined. Summary proportions and odds ratios (OR), along with their corresponding 95% confidence intervals (95% CI), were measured.

Results

In total, 112 investigations with 491,812 participants were included. The percentage of never-smokers in people with COPD was 41.1% (95% CI: 37.5–44.6%). The prevalence of never-smokers among males diagnosed with COPD was 22.3% (95% CI: 18.8–25.7%), which differed from that among women (81.3%, 95% CI: 75.3–87.2%). The results showed an association between the utilization of biomass fuel and the occurrence of COPD in never-smokers (OR: 1.25, 95% CI: 1.06–1.44). Among never-smokers, the data showed a close association between being underweight (OR: 1.89, 95% CI: 1.78–2.00), tuberculosis history (OR: 1.71, 95% CI: 1.53–1.88) and COPD. Never-smokers living in rural areas or those with low educational status were more susceptible to COPD.

Conclusion

This review confirmed the highly different proportions of never-smokers among male and female patients with COPD.

Trial registration

PROSPERO: CRD42023420786.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-024-20170-z.

Introduction

In recent decades, there has been a worldwide increase in the morbidity and mortality rates associated with chronic obstructive pulmonary disease (COPD). According to the findings of the China Pulmonary Health (CPH) investigation, COPD is a highly prevalent chronic ailment, impacting an estimated 13.7% of people aged > 40 years within the population [1]. Patients with COPD often complain of dyspnea, cough and sputum production, which could be exacerbated in the chronic course of the disease, leading to lung function decline and even death [1]. Notably, COPD presents a significant economic and social challenge in China.

COPD is distinguished by the presence of persistent airflow obstruction, which arises from a multitude of contributing factors, among which cigarette smoking is the most significant [2]. However, COPD can occur in a considerable number of non-smokers, who account for 15–50% of COPD patients globally [3]. Besides the well-documented causative factors (such as cigarette smoking), a variety of other factors are also involved in the susceptibility to or the pathogenesis of COPD [3]. Based on the findings of the international, population-based Burden of Obstructive Lung Disease investigation, 28% of individuals diagnosed with COPD was never-smokers, across 14 countries included in the analysis [4]. However, this proportion rose to 51% in China, as reported in the CPH study [1]. The notably higher prevalence in China, in comparison to that of other countries, warrants further investigation.

To date, the risk factors associated with COPD in non-smokers include genetic susceptibility, exposure to biomass fuel, a history of recurrent lower respiratory tract infections during childhood, intrauterine growth retardation, a history of pulmonary tuberculosis, inadequate nutrition, and low socioeconomic status [3], and these life-course exposures may impact lung function trajectory and COPD development in later life [5]. Recent studies have significantly advanced our understanding of the etiology of COPD and have been instrumental in guiding the screening of high-risk populations. Considering the increasing research in this area, it is imperative to conduct a systematic review and meta-analysis to comprehensively synthesize the existing evidence and delineate the incidence and risk factors of COPD in China.

In this meta-analysis, we systematically reviewed all relevant publications, estimated the overall and sex-specific proportions of non-smokers among patients with COPD, and conducted an assessment of the risk factors associated with COPD within the non-smoking population of China.

Methods

Search strategy

The present investigation employed a systematic review and meta-analysis methodology in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [6]. We searched electronic databases encompassing PubMed, Embase, Chinese China National Knowledge Infrastructure, Web of Science, Chinese WanFang, and Chinese Weipu databases from inception to August 5, 2024, with the following keywords: “Chronic Obstructive Pulmonary Diseases”/“COPD,” “China,” and “never-smokers”/“non-smoking”/“risk factors”/“risk factor.” The selected terms were integrated employing Boolean logical operators (OR, AND, and NOT). Details of the search terms are provided in Additional file 1. In addition, we evaluated review articles and references of the selected articles for additional eligible studies. Neither geographic nor language restrictions were applied in our search. The PROSPERO study registration number is CRD42023420786.

Study selection

The two reviewers, YZ and HLC, identified the studies meeting the specified eligibility criteria: (1) original articles encompassing study designs with no limitations, (2) reporting the prevalence of never-smokers diagnosed with COPD in China, (3) detailing COPD risk factors in never-smokers in China, and (4) unrestricted by language or geographical location. The exclusion criteria for the studies were as follows: (1) investigations with irrelevant topics; (2) studies focusing solely on findings from animal models; (3) conference abstracts, study protocols, or review articles; (4) lack of data on smoking status in COPD patients, or unutilizable data for evaluation; and (5) ecological studies. An review of the title, abstract, and the full text of the study was conducted by two reviewers independently. Any disagreements regarding study selection were resolved through consensus.

Data extraction and quality evaluation

Two authors, YZ and HLC, conducted separate reviews of full-text articles that were potentially relevant. There were no attempts to contact the study authors in case of missing data, and the aforementioned articles were later omitted. The extraction of data was conducted using standardized data collection forms. Discrepancies were resolved through the process of consulting with the corresponding author, YCS. Following a comprehensive and thorough deliberation on divergent viewpoints, the authors reached a unanimous consensus regarding the ultimate categorization and incorporation of all research studies. The data were extracted as follows:

(1) Study information: first author, publication year, study design, province or city of China, and patient age.

(2) Participants: Definition of COPD, number of COPD patients, number of never-smokers with COPD, sex, and overall and sex-specific percentage of never-smokers with COPD.

(3) Risk factors: biomass use, education level, history of tuberculosis, body mass index (BMI), sex, residence, and passive smoking.

(4) Effect sizes: summary proportion, odds ratio (OR), and relative risk (RR), each accompanied by their respective 95% confidence intervals (95%CI).

Employing the Newcastle-Ottawa Scale [7] and the cross-sectional investigation quality methodology checklist [8], two researchers evaluated the quality of the case-control, cohort, and cross-sectional investigations. The framework comprises three distinct domains, namely population selection, comparability, and exposure or result evaluation, encompassing a cumulative sum of nine points distributed across eight items. The scoring system classified scores from 7 to 9 as “high,” those from 4 to 6 as “medium,” and those from 1 to 3 as “poor.” Inconsistencies in quality assessments were resolved through communication [9].

Statistical analysis

To ascertain the proportion of never-smokers within the COPD population, we conducted a single-proportion meta-analysis using the metan command in STATA 14.0 (StataCorp, College Station, TX, USA) as outlined by DerSimonian and Laird [10]. A forest plot was used to estimate the overall incidence and effect of each study with their respective 95% CIs using a random-effects model. The same statistical methods were used to evaluate the sex-specific percentage of never-smokers with COPD. We employed the inverse variance index (I²) as a means of evaluating the level of heterogeneity present within the articles that were included (low, < 25%; moderate, 25–50%; and high, > 50%) [11]. Sensitivity analyses were conducted by separately eliminating one investigation at a time from each analysis to evaluate the effect on the overall percentage and robustness of the results. To identify any potential publication biases in the combined findings, we conducted funnel plots and Egger’s tests, employing a significance threshold of P < 0.05 [12].

Given the inconsistent findings regarding the COPD risk factors in never-smokers, as reported by various studies, and the presence of overlapping risk factors across multiple studies, we carried out a meta-analysis to systematically measure the risk factors associated with COPD in never-smokers. Risk factors were estimated using ORs with 95% CIs, and I² > 50% was defined as high heterogeneity [12]. Considering the variability in clinical settings across the studies, it was assumed that there would be heterogeneity. Therefore, a random-effects model was used for all subsequent analyses. The utilization of this methodology resulted in more cautious conclusions, as it considered variations both within and between studies when determining the error terms employed in the analysis. The combined OR was considered statistically significant if the 95% CI did not encompass 1.00. We did not assess sensitivity analyses and publication bias for articles that evaluated the likelihood factors for COPD in never-smokers, as there were fewer than ten studies in this section.

Results

Search results and study characteristics

The systematic search and selection process is illustrated in Fig. 1, showcasing the various steps involved. In total, 14,029 articles were analyzed. After removing duplicate entries and screening titles and abstracts, 3,517 articles were retained. Following a comprehensive review of the available literature, 112 investigations satisfied the predetermined criteria for inclusion and were subsequently incorporated into the present study [1, 13–123].

Fig. 1.

Flow diagram of study selection

Table 1 displays a comprehensive overview of the key attributes of the investigations that have been incorporated into this analysis. The published studies encompassed a time frame spanning from 2003 to 2024, comprising a total of 3 retrospective investigations, 14 cohort investigations, 17 case-control investigations, and 78 cross-sectional investigations. All the studies (n = 112) were conducted in China. Regarding study quality, 75 were categorized as high quality, and the remaining were categorized as medium quality (Additional file 2). The current systematic review included a sample size ranging from 75 to 317,000, with a total of 491,812 participants. All studies examined the associations between non-smoking status and COPD or COPD risk factors in never-smokers.

Table 1.

Characteristics of the included studies

Study		Study design	Age	Area	Definition of COPD	The number of COPD patients (n)	The number of never-smokers with COPD (n)	Prevalence of never-smokers among people with COPD	Quality
Liu et al. (2024) [13]		Cross-Sectional	Aged 65 years or older	A university hospital in southwestern China	Post-bronchodilator FEV1/FVC < 0.7	319	130	40.8%	High
Tan et al. (2024) [14]		Cross-Sectional	Aged 20 years or older	First Affiliated Hospital of Guangzhou Medical University	Post-bronchodilator FEV1/FVC < 0.7	1,282	411	32.1%	High
Fan et al. (2024) [15]		Cohort	Aged 40 years or older	10 medical centers in China	Post-bronchodilator FEV1/FVC < 0.7	235	115	48.9%	High
Zhang et al. (2024) [16]		Cohort	Aged 35–70 years	12 communities in different districts in Eastern China	Post-bronchodilator FEV1/FVC < 0.7	1,102	832	75.5%	High
Zhao et al. (2024) [17]		Cross-Sectional	Aged 60 years or older	5 public hospitals in Ningxia	Post-bronchodilator FEV1/FVC < 0.7	627	366	58.3%	Medium
Zhao et al. (2024) [18]		Cohort	Aged 60 years or older	Affiliated Hospital of Guangdong Medical University	Post-bronchodilator FEV1/FVC < 0.7	579	124	21.4%	High
Cui et al. (2023) [19]		Cohort	Aged 40 years or older	12 tertiary hospitals in Hunan and Guangxi provinces	Post-bronchodilator FEV1/FVC < 0.7	845	120	14.2%	High
Liu et al. (2023) [20]		Cross-Sectional	Not defined	Three tertiary hospitals from Anhui Province	Post-bronchodilator FEV1/FVC < 0.7	761	334	43.9%	Medium
Liu et al. (2023) [21]		Cross-Sectional	Aged 40 years or older	Beijing Chao-Yang Hospital	Post-bronchodilator FEV1/FVC < 0.7	226	32	14.2%	High
Zhu et al. (2023) [22]		Retrospective study	Aged 40–90 years	Southern and central districts of Beijing	Post-bronchodilator FEV1/FVC < 0.7	1,348	552	41.0%	High
Song et al. (2023) [23]		Cross-Sectional	Not defined	6 hospitals in Hunan province	Post-bronchodilator FEV1/FVC < 0.7	910	233	25.6%	High
Lin et al. (2023) [24]		Cohort	Aged 18 years or older	The Second Xiangya Hospital of Central South University	Post-bronchodilator FEV1/FVC < 0.7	461	72	15.6%	Medium
Lin et al. (2023) [25]		Cross-Sectional	Aged 40 years or older	Longnan City, Jiuquan City, Qingyang City, and Gannan City in Gansu	Post-bronchodilator FEV1/FVC < 0.7	508	266	52.4%	Medium
Zou et al. (2023) [26]		Retrospective study	Not defined	The First Affiliated Hospital of Hebei North University	Post-bronchodilator FEV1/FVC < 0.7	174	44	25.3%	Medium
Lv et al. (2023) [27]		Cohort	Not defined	Hefei, Bozhou, and Fuyang City	Post-bronchodilator FEV1/FVC < 0.7	789	352	44.6%	High
Zhan et al. (2023) [28]		Cohort	Aged 35–80 years	12 hospitals nationwide	Post-bronchodilator FEV1/FVC < 0.7	404	109	27.0%	High
Tang et al. (2022) [29]		Retrospective study	Aged 40–80 years	Zhejiang hospital in Zhejiang Province	Post-bronchodilator FEV1/FVC < 0.7	375	202	53.9%	High
Wu et al. (2022) [30]		Case-Control	Not defined	Hainan Provincial People's Hospital	Post-bronchodilator FEV1/FVC < 0.7	498	280	56.2%	Medium
Li et al. (2022) [31]		Cohort	Not defined	Tangshan City of Hebei Province	Post-bronchodilator FEV1/FVC < 0.7	474	295	62.2%	High
Wen et al. (2022) [32]		Cross-Sectional	Aged 40–80 years	Guangzhou, Shao Guan, and He Yuan communities of Guangdong Province	Post-bronchodilator FEV1/FVC < 0.7	895	102	11.4%	High
Zhang et al. (2022) [33]		Cohort	Not defined	People’s Hospital of Xinjiang Uygur Autonomous Region	Post-bronchodilator FEV1/FVC < 0.7	174	149	85.6%	Medium
	Tang et al. (2022) [34]	Cross-Sectional	Aged 50–92 years	Nanjing City in Jiangsu Provence	Post-bronchodilator FEV1/FVC < 0.7	170	35	20.6%	High
	Yang et al. (2022) [35]	Cohort	Aged 40 years or older	50 hospitals across six geographical regions	Post-bronchodilator FEV1/FVC < 0.7	4,978	1,280	25.7%	High
	Pan et al. (2022) [36]	Case-Control	Aged 18 years or older	Xiangya Second Hospital of Central South University and the 1st People′s Hospital of Huaihua	Post-bronchodilator FEV1/FVC < 0.7	959	112	11.7%	High
	Dai et al. (2022) [37]	Cross-Sectional	Aged 60 years or older	Several hospitals in Yunnan Province	Post-bronchodilator FEV1/FVC < 0.7	624	478	76.7%	Medium
	Wu et al. (2022) [38]	Cross-Sectional	Aged 60 years or older	Dancheng Branch of Xiangshan First People’s Hospital Medical and Health Group	Post-bronchodilator FEV1/FVC < 0.7	219	53	24.2%	Medium
	Chen et al. (2022) [39]	Cross-Sectional	Not defined	Three tertiary hospitals in Chongqing city	Post-bronchodilator FEV1/FVC < 0.7	335	100	29.9%	Medium
	Wang et al. (2021) [40]	Cross-Sectional	Aged 35 years or older	Northwest China	Post-bronchodilator FEV1/FVC < 0.7	394	320	81.2%	High
	Wang et al. (2021) [41]	Cross-Sectional	Aged 20 years or older	Shanxi Province	Post-bronchodilator FEV1/FVC < 0.7	363	160	44.0%	Medium
	Liu et al. (2021) [42]	Cross-Sectional	Aged 40 to 80 years	Chinese Epidemiological Survey of COPD (CESCOPD) study	Post-bronchodilator FEV1/FVC < 0.7	1,582	608	38.4%	High
	Li et al. (2021) [43]	Cross-Sectional	Aged 40 years or older	The Uyghur population in the Kashi region	Post-bronchodilator FEV1/FVC < 0.7	504	388	77.0%	Medium
	Li et al. (2021) [44]	Case-Control	Not defined	Hainan General Hospital	Post-bronchodilator FEV1/FVC < 0.7 and FEV1 < 80% predicted	315	166	52.7%	Medium
	Abudureheman et al. (2021) [45]	Case-Control	Not defined	Kashi in XinJiang	Post-bronchodilator FEV1/FVC < 0.7 and FEV1 < 80% predicted	541	417	77.1%	High
	Zhao et al. (2021) [46]	Cross-Sectional	Aged 40 years or older	10 communities of Shijingshan District in Beijing	Post-bronchodilator FEV1/FVC < 0.7	416	192	46.2%	High
	Dong et al. (2021) [47]	Cross-Sectional	Not defined	Dali Region in Yunnan Provence	Post-bronchodilator FEV1/FVC < 0.7	113	39	34.5%	Medium
	Chen et al. (2021) [48]	Cross-Sectional	Not defined	Guiyang public health treatment center	Post-bronchodilator FEV1/FVC < 0.7	315	184	58.4%	Medium
	Li et al. (2021) [49]	Cross-Sectional	Aged 18 years or older	Outpatient or inpatient department of 3 hospitals in Fujian province	Post-bronchodilator FEV1/FVC < 0.7	248	51	20.6%	Medium
	Zhou et al. (2020) [50]	Cross-Sectional	Aged 40 years or older	Outpatient department of the Second Xiangya Hospital	Post-bronchodilator FEV1/FVC < 0.7	1,241	263	21.2%	High
	Shi et al. (2020) [51]	Cross-Sectional	Not defined	Hainan Affiliated Hospital of Hainan Medical University	Post-bronchodilator FEV1/FVC < 0.7	313	164	52.4%	High
	Dong et al. (2020) [52]	Cross-Sectional	Not defined	Eight hospitals in Jilin Province	Post-bronchodilator FEV1/FVC < 0.7	306	87	28.4%	High
	Xiong et al. (2020) [53]	Case-Control	Aged 40 years or older	Southern part of China	Post-bronchodilator FEV1/FVC < 0.7	513	126	24.6%	High
	Ding et al. (2020) [54]	Case-Control	Not defined	Hainan Han population	Post-bronchodilator FEV1/FVC < 0.7	313	164	52.4%	High
	Zhang et al. (2020) [55]	Case-Control	Not defined	Sichuan Academy of Medical Sciences & Sichuan Provincial People's Hospital	Post-bronchodilator FEV1/FVC < 0.7	480	197	41.0%	Medium
	Zhu et al. (2020) (part of a national key research and development program about COPD) [56]	Cohort	Aged 40 years or older	Four Chinese tertiary hospitals	Post-bronchodilator FEV1/FVC < 0.7	1,644	325	19.8%	High
	Yan et al. (2020) [57]	Cross-Sectional	Aged 40 years or older	Suzhou City in Jiangsu Province	Post-bronchodilator FEV1/FVC < 0.7	586	397	67.8%	High
	Jia et al. (2020) [58]	Cohort	Aged 18 years or older	Outpatient department of West China Hospital	Post-bronchodilator FEV1/FVC < 0.7	75	15	20.0%	High
	Cheng et al. (2020) [59]	Cross-Sectional	Aged 20 years or older	4 districts in Shanghai	Post-bronchodilator FEV1/FVC < 0.7	591	339	57.4%	High
	Duan et al. (2020) [60]	Cross-Sectional	Aged 40 years or older	12 Grade-A hospitals in Hunan Province and Guangxi Zhuang Autonomous Prefecture	Post-bronchodilator FEV1/FVC < 0.7	5,183	1,495	28.8%	High
	Wang et al. (2020) [61]	Case-Control	Aged 35 years or older	Permanent residents in Hulunbuir city	Post-bronchodilator FEV1/FVC < 0.7	2,320	184	7.9%	Medium
	Ma et al. (2020) [62]	Cross-Sectional	Not defined	A tertiary general hospital in Qinghai	Post-bronchodilator FEV1/FVC < 0.7	3,264	2,085	63.9%	Medium
	Sheng et al. (2019) [63]	Cross-Sectional	Aged 60 years or older	Island area of Ningbo	Post-bronchodilator FEV1/FVC < 0.7	254	66	26.0%	High
	Zha et al. (2019) [64]	Cross-Sectional	Aged 40 years or older	Anhui Province	Post-bronchodilator FEV1/FVC < 0.7	272	92	33.8%	High
	Jia et al. (2018) [65]	Cross-Sectional	Not defined	Eleven tertiary hospitals	Post-bronchodilator FEV1/FVC < 0.7	1,698	497	29.3%	High
	Zhang et al. (2018) [66]	Case-Control	Aged 40 years or older	Ningxia Hui Autonomous Region	Post-bronchodilator FEV1/FVC < 0.7	491	219	44.6%	High
	Wang et al. (2018) [1]	Cross-Sectional	Aged 20 years or older	Ten provinces, autonomous regions, and municipalities	Post-bronchodilator FEV1/FVC < 0.7	4,908	2,479	50.5%	High
	Wu et al. (2018) [67]	Cross-Sectional	Aged 40 years or older	Shandong Province and Shanghai City	Post-bronchodilator FEV1/FVC < 0.7	744	316	42.5%	High
	Fang et al. (2018) [68]	Cross-Sectional	Aged 40 years or older	Seven major regions of China	Post-bronchodilator FEV1/FVC < 0.7	9,134	3,326	36.4%	High
	Hu et al. (2018) [69]	Cross-Sectional	Aged 40 years or older	Respiratory Medicine Clinic of Zhujiang Hospital	Post-bronchodilator FEV1/FVC < 0.7	631	197	31.2%	High
	Ma et al. (2018) [70]	Cross-Sectional	Aged 40 years or older	Linxia prefecture	Post-bronchodilator FEV1/FVC < 0.7	403	203	50.4%	Medium
	Shi et al. (2018) [71]	Cross-Sectional	Not defined	Hebei Province	Post-bronchodilator FEV1/FVC < 0.7	116	45	38.8%	Medium
	Zhang et al. (2017) [72]	Case-Control	Aged 40–80 years	Guangdong and Hubei Province	Post-bronchodilator FEV1/FVC < 0.7	989	167	16.9%	High
	Deng et al. (2017) [73]	Case-Control	Not defined	The Armed Police Corps Hospital	Post-bronchodilator FEV1/FVC < 0.7	120	55	45.8%	High
	Xiao et al. (2017) [74]	Cross-Sectional	Not defined	Beijing, Shanghai, Chengdu and Guangzhou City	Post-bronchodilator FEV1/FVC < 0.7	678	218	32.2%	High
	Yan et al. (2017) [75]	Cross-Sectional	Aged 35–70 years	12 provinces of China	Post-bronchodilator FEV1/FVC < 0.7	3,690	2,614	70.8%	High
	Li et al. (2017) [76]	Cross-Sectional	Aged 43–94 years	10 Beijing suburb hospitals	Post-bronchodilator FEV1/FVC < 0.7	447	141	31.5%	Medium
	Duan et al. (2017) [77]	Cross-Sectional	Aged 40 years or older	Wuwei City in Gansu Province	Post-bronchodilator FEV1/FVC < 0.7	190	76	40.0%	Medium
	Gong et al. (2017) [78]	Cross-Sectional	Aged 30–94 years	3 third-grade class-A hospitals in Shenyang City	Post-bronchodilator FEV1/FVC < 0.7	2,400	858	35.8%	High
	Lv et al. (2017) [79]	Cross-Sectional	Aged 40 years or older	Maoming City in Gangdong Province	Post-bronchodilator FEV1/FVC < 0.7	203	61	30.0%	High
	Zhang et al. (2017) [80]	Cross-Sectional	Aged 40 years or older	Changchun urban area in Jilin Province	Post-bronchodilator FEV1/FVC < 0.7	176	83	47.2%	High
	Zhu et al. (2017) [81]	Cross-Sectional	Aged 60–89 years	Zhejiang Province	Post-bronchodilator FEV1/FVC < 0.7	248	70	28.2%	High
	Qi et al. (2017) [82]	Cross-Sectional	Aged 40 years or older	Xinjiang Province	Post-bronchodilator FEV1/FVC < 0.7	500	110	22.0%	Medium
	Xiao et al. (2017) [83]	Cross-Sectional	Not defined	Beijing, Shanghai, Chengdu and Guangzhou city	Post-bronchodilator FEV1/FVC < 0.7	678	218	32.2%	High
	Cai et al. (2017) [84]	Cross-Sectional	Not defined	Chaozhou-Shantou Region of Guangdong Province	Post-bronchodilator FEV1/FVC < 0.7	876	260	29.7%	Medium
	Cui et al. (2017) [85]	Cross-Sectional	Not defined	Luoyang City in Henan Provence	Lung function test record	1,356	436	32.2%	Medium
	Ding et al. (2016) [86]	Case-Control	Aged 40 years or older	Hlai (the Li) ethnicity, Hainan Province	Post-bronchodilator FEV1/FVC < 0.7	212	154	72.6%	Medium
	Lu et al. (2016) [87]	Case-Control	Aged 40–80 years	Guangzhou City	FEV1/FVC < the lower limit of normal (LLN)	855	213	24.9%	High
	Liao et al. (2015) [88]	Cross-Sectional	Aged 40–70 years	Chengdu City in Sichuan Provence	Post-bronchodilator FEV1/FVC < 0.7	151	100	66.2%	High
	Han et al. (2015) [89]	Cross-Sectional	Aged 40 years or older	Heilongjiang Provence	Post-bronchodilator FEV1/FVC < 0.7	297	157	52.9%	High
	Yang et al. (2015) [90]	Case-Control	Not defined	Guangzhou City in Guangdong Provence and Suzhou city in Jiangsu Provence	Post-bronchodilator FEV1/FVC < 0.7	1,791	888	49.6%	Medium
	Feng et al. (2015) [91]	Cross-Sectional	Not defined	Wuhan cohort study	Post-bronchodilator FEV1/FVC < 0.7	228	169	74.1%	High
	Li et al. (2015) [92]	Cross-Sectional	Aged 40 years or older	Ningbo City in Zhejiang Provence	Post-bronchodilator FEV1/FVC < 0.7	396	172	43.4%	Medium
	Liu et al. (2015) [93]	Cross-Sectional	Aged 40–75 years	Jilin Provence	Post-bronchodilator FEV1/FVC < 0.7	77	41	53.2%	Medium
	Xiao et al. (2015) [94]	Cross-Sectional	Not defined	Linxia Hui Autonomous Prefecture	Post-bronchodilator FEV1/FVC < 0.7	269	60	22.3%	Medium
	Yu et al. (2014) [95]	Cross-Sectional	Aged 40 years or older	Jiading Districts of Shanghai	Post-bronchodilator FEV1/FVC < 0.7	165	63	38.2%	High
	Qiu et al. (2013) [96]	Cross-Sectional	Aged 40 years or older	Ningxia Hui Autonomous Region	Post-bronchodilator FEV1/FVC < 0.7	360	180	50.0%	High
	Jiang et al. (2013) [97]	Cross-Sectional	Not defined	Guangzhou City in Guangdong Provence	Post-bronchodilator FEV1/FVC < 0.7	911	230	25.2%	High
	Miao et al. (2013) [98]	Prospective cohort	Not defined	Zhengzhou City in Henan Provence	Post-bronchodilator FEV1/FVC < 0.7	1,166	507	43.5%	High
	Zou et al. (2013) [99]	Cross-Sectional	Aged 40 years or older	Haiyang City in Shandong Province	Post-bronchodilator FEV1/FVC < 0.7	159	67	42.1%	High
	Zhang et al. (2013) [100]	Cross-Sectional	Aged 40 years or older	Kunming City in Yunnan Province	Post-bronchodilator FEV1/FVC < 0.7	92	26	28.3%	Medium
	Lou et al. (2012) [101]	Cross-Sectional	Aged 40–75 years	Rural area of Xuzhou City	Post-bronchodilator FEV1/FVC < 0.7	5,650	3,136	55.5%	High
	Zhou et al. (2012) [102]	Case-Control	Aged 40 years or older	Eight hospitals in Beijing City	Post-bronchodilator FEV1/FVC < 0.7	193	81	42.0%	High
	Wang et al. (2012) [103]	Cross-Sectional	Aged 40–91 years	Chengdu City in Sichuan Province	Post-bronchodilator FEV1/FVC < 0.7	1,335	307	23.0%	Medium
	Hu et al. (2012) [104]	Cross-Sectional	Aged 15 years or older	Cixi City in Zhejiang Province	Post-bronchodilator FEV1/FVC < 0.7	1,650	472	28.6%	Medium
	Wang et al. (2011) [105]	Case-Control	Not defined	Qilu Hospital and the Second Affiliated Hospital of Shandong University of Traditional Chinese Medicine	Post-bronchodilator FEV1/FVC < 0.7	409	190	46.5%	High
	Ling et al. (2011) [106]	Cross-Sectional	Aged 15–92 years	Xinjiang rural areas	Post-bronchodilator FEV1/FVC < 0.7	138	101	73.2%	High
	Du et al. (2011) [107]	Cross-Sectional	Aged 40 years or older	Jilin City in Jilin Province	Post-bronchodilator FEV1/FVC < 0.7	403	80	19.9%	Medium
	Luo et al. (2011) [108]	Cross-Sectional	Not defined	Beijing City	Post-bronchodilator FEV1/FVC < 0.7	134	38	28.4%	Medium
	Weng et al. (2011) [109]	Cross-Sectional	Aged 40 years or older	Chongqing City	Post-bronchodilator FEV1/FVC < 0.7	160	42	26.3%	High
	Lam et al. (2010) [110]	Cross-Sectional	Aged 50 years or older	The Guangzhou Biobank Cohort Study	Post-bronchodilator FEV1/FVC < 0.7	522	364	69.7%	High
	Zhang et al. (2009) [111]	Cross-Sectional	Aged 40 years or older	Rural areas of Liaoning province	Post-bronchodilator FEV1/FVC < 0.7	172	61	35.5%	High
	Zhou et al. (2009) [112]	Cross-Sectional	Aged 40 years or older	Rural area in 7 provinces or cities of China	Post-bronchodilator FEV1/FVC < 0.7	830	321	38.7%	High
	Wang et al. (2009) [113]	Case-Control	Aged 30 years or older	Eight hospitals in Beijing City	Post-bronchodilator FEV1/FVC < 0.7	306	105	34.3%	High
	Ko et al. (2008) [114]	Cohort	Aged 60 years or older	Hong Kong City	Post-bronchodilator FEV1/FVC < 0.7	261	172	65.9%	High
	Zhong et al. (2007) [115]	Cross-Sectional	Aged 40 years or older	Seven provinces/cities	Post-bronchodilator FEV1/FVC < 0.7	1,668	644	38.6%	High
	Liu et al. (2007) [116]	Cross-Sectional	Aged 40 years or older	Guangdong Province	Post-bronchodilator FEV1/FVC < 0.7	310	114	36.8%	High
	Xu et al. (2007) [117]	Cross-Sectional	Aged 35 years or older	Nanjing City in Jiangsu Provence	Post-bronchodilator FEV1/FVC < 0.7	1,743	1,097	62.9%	High
	Wang et al. (2005) [118]	Cross-Sectional	Aged 40 years or older	Northern part of Guangdong province	Post-bronchodilator FEV1/FVC < 0.7	176	70	39.8%	High
	Liu et al. (2005) [119]	Cross-Sectional	Aged 40 years or older	Guangdong Province	Post-bronchodilator FEV1/FVC < 0.7	310	128	41.3%	High
	Li et al. (2003) [120]	Cross-Sectional	Not defined	First Affiliated Hospital of Zhongshan University	Post-bronchodilator FEV1/FVC < 0.7	713	378	53.0%	Medium

Abbreviations: COPD, chronic obstructive pulmonary disease; FEV₁, forced expiratory volume in one second; FVC, forced vital capacity

Percentage of never-smokers in the COPD population

One hundred and nine studies (72 high-quality studies) reported the percentage of never-smokers among individuals with COPD. The outcomes showed that the percentage of never-smokers in the COPD population was 41.1% (95%CI: 37.5–44.6%, I² = 99.4%) (Fig. 2).

Fig. 2.

The percentage of never-smokers in COPD patients

Meta-analysis of the 21 studies [1, 41, 46, 47, 59, 60, 63–65, 68, 78, 79, 99–101, 103, 112, 117–120] reported the prevalence of never-smokers among COPD men. Our study indicated that the percentage of male never-smokers with COPD was 22.3% (95%CI: 18.8–25.7%, I² = 97.6%) (Fig. 3 and Additional file 3).

Fig. 3.

The percentage of never-smokers in male COPD patients

Twenty studies [1, 41, 46, 47, 59, 60, 63–65, 68, 78, 99–101, 103, 112, 117–120] reported the percentage of female non-smokers with COPD. The investigation revealed that the proportion of COPD female never-smokers was 81.3% (95%CI: 75.3–87.2%, I² = 99.2%) (Fig. 4 and Additional file 3).

Fig. 4.

The percentage of never-smokers in female COPD patients

COPD risk factors in never-smokers

Five studies have reported the risk factors of COPD in never-smokers (Additional file 4). High-quality evidence based on the results from two studies of biomass exposure [1, 121] showed an association between biomass fuel and COPD in never-smokers (OR: 1.25, 95% CI: 1.06–1.44) (Additional file 5). Correspondingly, evidence from four investigations [1, 121–123] showed a positive association between being underweight (BMI < 18.5 kg/m²) and COPD in never-smokers (OR: 1.89, 95% CI 1.78–2.00) (Additional file 6). A significant association between the history of tuberculosis [1, 110, 122, 123] and COPD was observed in never-smokers (OR: 1.71, 95% CI 1.53–1.88) (Additional file 7). No significant associations were identified between passive smoking [1, 121–123] and COPD in never-smokers (OR: 0.90, 95% CI 0.87–0.94) (Additional file 8). Moreover, never-smokers who resided in rural regions [1, 122, 123] were more susceptible to COPD (OR: 1.20; 95% CI 1.16–1.25) than those who did not (Additional file 9). Four additional investigations [1, 121–123] showed that never-smokers with low educational status (primary school and lower) were more prone to developing COPD (OR: 1.17, 95% CI 1.12–1.22) than those with a high educational status (Additional file 10).

Sensitivity analysis and publication bias

Sensitivity analysis suggested that there was no substantial alteration in the percentage of never-smokers among patients with COPD after systematically eliminating one study at a time; we have verified the robustness and consistency of our findings (Additional file 11). As for the prevalence of never-smokers among male or female COPD patients, the results of sensitivity analysis remain stable after leave-one-out analysis (Additional file 12 and Additional file 13).

As most of the included studies were small-scale, the asymmetrical funnel plot indicated a potential publication bias. However, after performing Egger’s test, no significant publication bias was observed in relation to the percentage of never-smokers among COPD patients (P = 0.069) (Additional file 14).

Discussion

The high incidence, associated disability, and mortality rates of COPD in China present a significant public health concern [1]. Cigarette smoking is identified as the primary risk factor associated with COPD [124]. Nevertheless, other environmental factors and host susceptibility are also involved. This systematic review and meta-analysis encompassed a total of 112 investigations involving a substantial participant pool of 491,812 individuals. The investigation outcomes reveal that the incidence of non-smokers among COPD patients in China was as high as 41.1% (95%CI: 37.5–44.6%). In the investigations encompassed in the meta-analysis, the sex ratio of COPD patients was not completely consistent, which could influence the prevalence of non-smokers within the population of individuals with COPD. To further explore the differences in sex-specific proportions, we summarized 21 articles that contained relevant sex-based data separately. We found that the number of non-smokers among COPD patients greatly varied between men and women. Specifically, the proportion of non-smokers among male patients was 22.3% (95%CI: 18.8–25.7%), while that in female patients was as high as 81.3% (95%CI: 75.3–87.2%).

The implications of the study’s findings are significant in terms of comprehending the risk factors linked to COPD among non-smokers. The outcomes of this meta-analysis indicate that the utilization of biomass fuel is associated with the onset of COPD among non-smokers. In contrast to developed nations, indoor air pollution, encompassing coal and biomass combustion, has been considered as a significant factor in COPD development in China and other developing countries [116, 125, 126], particularly among non-smokers. Biomass encompasses wood, grass, charcoal, crop stems, and animal dung, which are utilized as fuel sources for cooking and heating [127]. The primary constituents of biomass smoke that pose a significant risk to human health consist of nitrogen oxides, oxysulfides, carbonic oxides, polycyclic organic compounds, and hydrocarbon compounds. These substances are generated as a result of the incomplete combustion of biomass [3]. According to data derived from China, in rural areas, it was found that approximately 60% of households utilized biomass, while approximately 31% of households relied on coal for cooking or heating [128]. Hence, the combustion of biomass fuel emits smoke that constitutes a significant risk factor contributing to COPD development within the Chinese population. The significance of these findings lies in the fact that a considerable proportion of non-smokers in China, particularly those residing in rural areas, are exposed to biomass smoke on a regular basis through cooking or heating activities. Hence, it is imperative to consider the contribution of biomass smoke in the progression of COPD.

China is confronted with a significant burden of pulmonary tuberculosis. The findings indicated that the occurrence of COPD in individuals without a smoking history who have tuberculosis was greater than that in the overall population, supporting the notion that pulmonary tuberculosis is an independent COPD risk factor in non-smokers. However, the mechanisms underlying airflow obstruction after tuberculosis remain largely unknown. The presence of bronchostenosis, characterized by inflammation and lesions, has been demonstrated to be a consequence of tuberculosis affecting the endobronchial region, in addition to contributing to the development of secondary tuberculous lymphadenopathy [129]. Moreover, tuberculosis can cause the dysregulation of matrix metalloproteinases and widespread parenchymal lung destruction [130]. Thus, it is possible that a combination of airway and lung parenchyma impairment leads to the development of airflow obstruction and abnormalities in ventilation, resulting in reduced lung function. In addition, cumulative effects of life-course exposure may also impact on the association [5].

We found that COPD risk increased in non-smoking patients with lower educational attainment and in those residing in rural regions, which may be attributable to occupational exposure to dust, smoke, biomass smoke, unfavorable living conditions, malnutrition, and further risk factors associated with the lifestyle of these patients. The meta-analysis revealed a correlation between low socioeconomic status, as indicated by educational attainment, and the occurrence of COPD in non-smokers. However, this association could potentially be influenced by various established risk factors, including occupational exposure, respiratory disease throughout childhood, malnutrition, and limited access to healthcare [121].

Tobacco smoke exposure is common in China. Moreover, secondhand smoke has been linked to an increased level of bronchial responsiveness and a progressive decrease in pulmonary function [125, 131]. However, our analysis concluded that passive smoking was not a significant COPD risk factor in non-smokers in China. Wang et al. [1] assessed the prevalence of cohabitating smokers among individuals diagnosed with COPD. Furthermore, passive smoking, as defined by Long et al. [122], refers to the act of inhaling smoke by never-smokers who may or may not reside with smokers. Conversely, passive smoking, as defined by Smith et al. [123] is characterized by the quantification of exposure through the assessment of frequency and duration of hours per week. The Guangzhou Biobank Cohort investigation [132] revealed that the duration of passive smoking experienced at home or in the workplace was significantly associated with the development of COPD. This association was found to be more influential than the one involving the mere presence of smokers within the same household or indoor setting. Hence, whether passive smoking is a risk factor for non-smoker COPD, the duration of exposure to passive smoking, and the prevalence of smokers in the surrounding environment are crucial factors to consider.

Moreover, we identified a strong association between low BMI (BMI < 18.5 kg/m²) and COPD risk. A diet with a greater consumption of vegetables, fruits, and fish could decrease COPD risk [133, 134]. Previous research has revealed that a low BMI is a significant and independent indicator of the risk of mortality [135]. In addition, it has been observed that COPD is associated with a gradual decline in skeletal muscle mass [4]. The outcomes of this investigation indicate that COPD is related to dietary patterns and weight.

In this meta-analysis, the proportion of non-smokers was found to be greater among female COPD patients than among their male counterparts. This may be explained by the fact that females are more likely to be exposed to biomass fuel for cooking in China [121]. Current evidence suggests a plausible sex-specific effect on COPD development related to exposure to different risk factors [136, 137].

In this meta-analysis, we summarized 112 articles on the smoking status of COPD patients. In total, 78 study designs were cross-sectional, with 491,812 participants. To the best of our knowledge, this is the first report investigating the disparities between sexes regarding the prevalence of non-smokers diagnosed with COPD within the Chinese population. Furthermore, this meta-analysis is the first to evaluate the COPD risk factors in non-smokers in China.

Limitations

Our investigation has several limitations. First, the records of smoking history and risk factor exposures relied solely on questionnaires; therefore, recall and responder biases may be unavoidable. Second, heterogeneity in the combined proportions was primarily caused by different study designs, procedures, and measurements. The heterogeneity presented in the summary estimates was primarily owing to potential confounders; hence, it is advisable to exercise caution when interpreting the outcomes of the analysis.

Conclusion

In conclusion, this meta-analysis elucidated the prevalence of and potential risk factors for COPD in non-smokers. The results suggested that lower BMI, less education, living in a rural residence, biomass use, and history of pulmonary tuberculosis were associated with a greater COPD risk among never-smokers. Compared with that in smokers, COPD in non-smokers may be caused by different mechanisms, and alternative preventive or therapeutic modalities are needed.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

BMI

Body mass index

CI

Confidence interval

COPD

Chronic obstructive pulmonary disease

CPH

China pulmonary health

I²

Inverse variance index

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

RR

Relative risk

Author contributions

YCS designed the study. YZ, XYG, HLC, JGQ, LL and YCS analyzed and interpreted the data. YZ, XYG and YCS conducted the literature screening. YZ and YCS participated in the statistical analysis. YCS and YZ wrote the manuscript, and YCS revised the manuscript. All the authors read and accepted this paper.

Funding

The research described in this investigation received financial support from the National Natural Science Foundation of China (grant number 81970041), the Capital’s Funds for Health Improvement and Research (grant number 2022-2G-40910), and the Key Clinical Projects of Peking University Third Hospital (grant number BYSYZD2022014).

Data availability

The entirety of the data utilized or investigated in the course of this investigation has been incorporated within this published article, along with its accompanying Supplementary Material.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.