Preplanned Studies: The Association Between Preserved Ratio Impaired Spirometry and Mortality — 10 CKB Study Areas, China, 2004–2022

Preserved ratio impaired spirometry (PRISm) is a non-obstructive spirometry phenotype described as a transient state associated with the progression of chronic obstructive pulmonary disease (COPD) (1). The prevalence of COPD is increasing in China (2). Previous studies excluded PRISm from respiratory studies (3), while recent studies on PRISm have predominantly been conducted in Western or high-income countries (1), leaving the significance of this lung function pattern in China uncertain. To address these gaps, this study analyzed the association between PRISm and all-cause and cause-specific mortality based on the China Kadoorie Biobank (CKB), a large-scale prospective cohort. Elevated risks of mortality from all-cause, cardiovascular, neoplastic, respiratory, and infectious and parasitic diseases were observed in PRISm, with rates comparable to those observed in obstructive spirometry. PRISm should be given more attention to avoid its progression to COPD.

The CKB, which has been reported previously (4), recruited 512,724 participants aged 30 to 79 from 10 areas across China at baseline from 2004 to 2008. All individuals underwent interviewer-administered questionnaires, physical measurements, and spirometry tests by trained technicians according to standard operating procedures. In the present study, participants with previously self-reported physician-diagnosed ischemic heart disease, stroke, cancer, or asthma at baseline, lost to follow-up shortly after baseline, with an FEV₁/FVC ratio >1.0, or missing data for covariates were omitted from the study, leaving 484,301 participants.

For baseline spirometry, the higher of two measurements for both FEV1 and FVC was used to calculate the FEV₁/FVC ratio. Predicted FEV₁ was calculated using the Global Lung Function Initiative 2012 equations for the Southeast Asian and Northeast Asian populations (5). Normal lung function was defined as an FEV₁/FVC ratio ≥0.7 and FEV₁ ≥80% predicted, PRISm as an FEV₁/FVC ratio ≥0.7 and FEV₁ <80% predicted, and obstructive spirometry as an FEV₁/FVC ratio <0.7.

Mortality data were obtained from official residential records and the Disease Surveillance Points system, supplemented by annual door-to-door investigations among those not linked to the database. Based on the International Classification of Disease, 10th Revision (ICD-10), the outcomes comprised death from all causes and cause-specific mortality, including circulatory diseases (ICD-10: I00–I99), neoplasms (C00–D48), respiratory diseases (J00–J99), infectious and parasitic diseases (A00–B99), ischemic heart disease (I20–I25), intracerebral hemorrhage (I61), ischemic stroke (I63), lung cancer (C34), COPD (J41–J44), pneumonia (J12–J18), and respiratory tuberculosis (A15–A16). Participants were censored upon death, loss to follow-up, or December 31, 2022, whichever occurred first.

Baseline variable means and prevalences were calculated for normal, PRISm, and obstructive spirometry groups using linear regression for continuous variables and logistic regression for categorical variables, adjusted for age, sex, and 10 study areas when appropriate. Mortality rates per 100,000 person-years for each group were standardized to the age structure of the 7^_th National Population Census data (2020). For all-cause mortality, stratified Cox proportional hazards regression was used to estimate hazard ratios (HRs) with 95% confidence intervals (CIs). A proportional subdistribution hazards regression model for cause-specific mortality was fitted to account for competing risks from other causes. All analyses were stratified by age (in 5-year groups), sex, and 10 study areas, and adjusted for education, occupation, household income, marital status, alcohol consumption, smoking status, passive smoking status, physical activity levels, primary cooking and heating fuel use, consumption frequency of fresh fruits, fresh vegetables, and meat, general obesity, and abdominal obesity. The proportional hazards assumption was verified using Schoenfeld residuals. Associations between mortality and PRISm were analyzed after stratification by the presence of self-reported cough or sputum, ever-smoking, physician-diagnosed bronchitis or emphysema, and tuberculosis at baseline. Subgroup analyses were conducted across sex, age (≥60 years or not), and region (urban/rural).

Competing-risk analysis was performed using SAS (version 9.4, SAS Institute Inc, Cary, NC, USA), and all other statistical analyses were performed using R (version 4.3.1, R Foundation for Statistical Computing, Vienna, Austria). All tests were two-tailed, and P<0.05 were considered statistically significant.

The study included 484,301 participants with a mean age of 51.5 [standard deviation (SD)=10.5] years at baseline, 59.1% women, and 32.3% ever-smokers. At baseline, 117,210 (24.2%) had PRISm, with a mean FEV₁ % predicted of 83.9%. Compared with the normal group, individuals with PRISm or an obstructive spirometry pattern were more likely to be older, male, current or former smokers, reside in rural areas, have a lower socioeconomic level, and report poorer self-rated health status (Table 1).

During a median follow-up of 16.0 years, 73,288 deaths were documented, and the corresponding all-cause age-standardized mortality rates for normal, PRISm, and obstructive spirometry were 1,070.5, 1,574.4, and 1,943.3 per 100,000 person-years, respectively. Compared with the normal group, PRISm had a higher adjusted all-cause (HR=1.37, 95% CI: 1.35, 1.40), circulatory disease [subdistribution hazard ratio (SHR)=1.36, 95% CI: 1.33, 1.40], neoplasm (SHR=1.07, 95% CI: 1.04, 1.11), and infectious and parasitic disease (SHR=1.47, 95% CI: 1.24, 1.73) mortality, comparable to those observed in obstructive spirometry (Table 2). For respiratory disease mortality, the SHR in PRISm was 2.45 (95% CI: 2.30, 2.60), lower than those with obstructive spirometry (SHR=5.05, 95% CI: 4.73, 5.39). For disease-specific deaths, PRISm was associated with a higher risk of death from ischemic heart disease (SHR=1.37, 95% CI: 1.31, 1.42), intracerebral hemorrhage (SHR=1.31, 95% CI: 1.25, 1.38), ischemic stroke (SHR=1.30, 95% CI: 1.20, 1.40), lung cancer (SHR=1.26, 95% CI: 1.19, 1.34), COPD (SHR=3.06, 95% CI: 2.84, 3.31), pneumonia (SHR=1.47, 95% CI: 1.28, 1.68), and respiratory tuberculosis (SHR=2.63, 95% CI: 1.82, 3.82).

When stratified by respiratory symptoms of PRISm individuals, those with frequent coughing or sputum (HR=1.62, 95% CI: 1.57, 1.68) and self-reported prior bronchitis or emphysema (HR=1.81, 95% CI: 1.72, 1.91) had a much higher risk of all-cause mortality. However, the risk of all-cause mortality in the PRISm group did not appear to be substantially affected by smoking or passive smoking status (Figure 1). Analyses of sex, age, and regional subgroups of PRISm mortality risk did not show large differences (Supplementary Table S1).

Figure 1. 

Associations of PRISm with symptoms, smoking, bronchitis or emphysema, tuberculosis at baseline with all-cause mortality in China Kadoorie Biobank, 2004–2022.

Note: The mortality rate is per 100,000 person-years and is age-standardized based on the 7^th National Population Census data (2020). HRs were stratified by age (in 5-year intervals), sex, and study areas, and adjusted for education, occupation, household income, marital status, alcohol consumption, smoking status, passive smoking status, physical activity levels, cooking and heating fuel usage, consumption frequency of fresh fruits, fresh vegetables, meat, and general and abdominal obesity. Symptoms were defined as frequent coughing or sputum. PRISm was defined as FEV₁/FVC ratio ≥0.7 and FEV₁ <80%.

Abbreviation: PRISm=preserved ratio impaired spirometry; HR=hazard ratio; CI=confidence interval.

*PRISm individuals were grouped according to whether they never/occasionally smoked or smoked regularly, and whether they were passively exposed to smoke.