Risk of severe cardiovascular events and incidence rate of cardiopulmonary events among patients with COPD in China: a retrospective cohort study

Abstract

Objectives

To examine the risk of severe cardiovascular (CV) events in patients with chronic obstructive pulmonary disease (COPD) across different time periods following COPD exacerbations and the incidence rate of cardiopulmonary events in a real-world setting in China.

Design

Retrospective cohort study.

Setting

Regional electronic health records database from Yinzhou District of Ningbo City, China.

Participants

A total of 14 713 patients aged ≥40 years with a first COPD diagnosis between 1 January 2014 and 1 March 2022.

Primary and secondary outcome measures

The risk of severe CV events (ie, hospitalisation and a primary or secondary discharge code for acute coronary syndrome, heart failure decompensation, cerebral ischaemia, arrhythmia and CV-related death) during different exposed time periods following a COPD exacerbation, the incidence rate of overall cardiopulmonary events (ie, severe exacerbation of COPD, all-cause mortality, inpatient CV events, inpatient ischaemic stroke and inpatient tachyarrhythmia/atrial fibrillation) and the incidence rate stratified by COPD exacerbation history.

Results

We included a total of 14 713 patients. During a median (IQR) follow-up of 2.8 (4.0) years, 20.1% experienced severe CV events. Compared with the unexposed period, the risk of severe CV events was the highest in the first 10 days following a COPD exacerbation (adjusted HR 10.00, 95% CI 8.16 to 12.25). The risk of severe CV events decreased over time but remained significantly elevated up to 90 days post exacerbation. We found that 32.7% of COPD patients experienced cardiopulmonary events, with a crude incidence rate of 9.38 (95% CI 9.09 to 9.69) per 100 person-years.

Conclusion

This study is the largest retrospective cohort study investigating CV and cardiopulmonary events among patients with COPD in China. Our findings highlight an elevated risk of CV events closer to the time of COPD exacerbations and show that nearly one-third of COPD patients experience cardiopulmonary events.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Large sample size from a regional electronic health records database enabled robust risk estimates of severe cardiovascular (CV) events following chronic obstructive pulmonary disease (COPD) exacerbations.

• Retrospective cohort design allowed assessment of temporal associations between COPD exacerbations and CV events.

• Generalising these findings to broader populations should account for variation in COPD diagnosis and management, geographic and demographic characteristics and differences between hospital and non-hospital care settings.

Introduction

Chronic obstructive pulmonary disease (COPD), a common respiratory condition characterised by chronic airflow limitation, is a leading contributor to the global disease burden.¹,³ In China, which accounts for one-fourth of the global COPD population, the incidence and prevalence surged by 61.2% (2.39–3.97 million) and 67.8% (27.75–45.17 million), respectively, between 1990 and 2019.⁴ As the population continues to age rapidly in China, the burden of COPD is projected to continue escalating.

COPD exacerbations are strongly associated with heightened risks of cardiovascular (CV) events.⁵,¹⁴ Large trials, including IMPACT, SUMMIT and UPLIFT, consistently demonstrate that CV risks peak within 30 days post exacerbation and remain elevated for up to 1 year, irrespective of baseline CV history or exacerbation severity,^{8 9 15} while the UPLIFT trial reports a 6-month post-exacerbation surge in CV events (eg, cardiac failure, ischaemic heart disease, stroke), even in patients without prior cardiac disease.¹⁵ This relationship may stem from shared pathophysiological mechanisms between COPD and CV diseases, such as systemic inflammation and hypoxia.¹¹

Despite global evidence, the COPD-CV risk association remains understudied in China, hindering tailored management for COPD patients with CV comorbidities. Additionally, the impact of COPD exacerbation severity on CV risk is poorly characterised. This study addresses these gaps by contributing Chinese real-world data to the global EXAcerbations of COPD and their OutcomeS on CardioVascular disease (EXACOS-CV) Programme.¹⁶ Given the expanding understanding of the intricate biological pathways and interlinks between COPD and CV diseases,¹¹ it is essential to investigate COPD-associated cardiopulmonary risk, which is defined as the risk of serious respiratory and/or CV events in patients with COPD as proposed by Singh et al.¹⁷ To the best of our knowledge, cardiopulmonary events, which have never been studied in patients with COPD in China, were introduced in our study for the first time.

To address these knowledge gaps, we conducted this study to examine the risk of severe CV events in patients with COPD across different time periods following COPD exacerbations and to assess the incidence rate of cardiopulmonary events in a real-world setting in China.

Methods

Study design and data source

This observational retrospective cohort study, part of the multinational EXACOS-CV Programme,¹⁶ used data from the Yinzhou regional electronic health records database—a population-based system covering 98% of permanent residents in Yinzhou, a developed district of Ningbo City, China. Since 2009, the database has integrated inpatient/outpatient records, public health surveillance, disease management and mortality data via a unique and encoded patient ID.

Study population

COPD patients aged ≥40 years with ≥2 primary diagnosis items/codes (ICD-10 (International Classification of Diseases, Tenth Revision) codes with J44, J44.0, J44.1, J44.8 or J44.9) for COPD in the outpatient setting or ≥1 primary or secondary diagnosis items/codes for COPD in the inpatient setting between 1 January 2014 and 1 March 2022 were eligible. The index date was defined as the first COPD diagnosis during this period. Baseline data spanned 24 months preceding the index date. Follow-up continued until 1 March 2023, loss to follow-up or death (whichever occurred first). Patients were followed up starting from 1 day after the index date until the earliest of (a) 1 March 2023 (administrative right censoring); (b) loss to follow-up or (c) all-cause death (death data were available until 30 June 2022). Patients with <24 months of pre-index data or alpha-1 antitrypsin deficiency were excluded (figure 1).

Figure 1. Study design scheme. COPD, chronic obstructive pulmonary disease; CV, cardiovascular.

Exposures and outcomes

Exacerbations during follow-up were categorised as moderate (outpatient visit with general practitioner, pulmonologist or internist, with new corticosteroid and/or antibiotic prescription ≥3 days and <30 days) or severe (hospitalisation with ICD-10 codes J44/J44.0/J44.1 as the main discharge diagnosis or J44.0 as a secondary discharge diagnosis). Two exacerbations within 14 days were merged, with severity assigned to the highest level and the date of the first exacerbation as day 1. We should acknowledge that this could slightly shift the exposure window for the second exacerbation, potentially underestimating risk immediately after the second event. Accordingly, analyses of moderate exacerbations reflect episodes that did not escalate to severe exacerbations within the 14 days, whereas exacerbations that escalated were classified as severe. Exposure periods post exacerbation (1–10 days, 11–30 days, 31–90 days, 91–180 days, 181–365 days and >365 days) and unexposed time (pre-first exacerbation) were analysed. See figure 2 for a detailed illustration of unexposed and exposed subtime periods.

Figure 2. Graphical description for unexposed and exposed subtime periods categorised by the incidence of exacerbation during the follow-up period. Note: the period after the first severe CV event was not taken into account as unexposed period in analysis. AECOPD, acute exacerbation of COPD; COPD, chronic obstructive pulmonary disease; CV, cardiovascular.

Outcomes of interest were severe CV events and cardiopulmonary events that occurred during the follow-up period. Severe CV events of any type were defined as hospitalisation and a primary or secondary discharge code for acute coronary syndrome, heart failure decompensation, cerebral ischaemia, arrhythmia and CV-related death (online supplemental table 1). Cardiopulmonary events of any type were defined as severe exacerbation of COPD, all-cause mortality, inpatient CV events (including acute heart failure, acute myocardial infarction, angina pectoris, cardiac arrest and other ischaemic heart disease/acute coronary syndrome), inpatient ischaemic stroke and inpatient tachyarrhythmia/atrial fibrillation. The two outcomes, severe CV events and cardiopulmonary events, were developed for different analytic purposes and therefore overlap but are not identical. These outcomes aligned with prior studies.^{8 9 11 15}

Covariates

Baseline and time-dependent covariates listed in online supplemental table 2 were prespecified as potential confounders based on discussion with clinical experts and the Yinzhou database scientists.

Statistical analysis

The full COPD cohort was described in terms of demographics and clinical characteristics during the baseline period. Continuous variables were summarised using mean, SD, median, IQR and minimum and maximum values. Binary/categorical variables were summarised using frequency and percentage. Count variables were described as continuous variables, using the mean value, SD, median, IQR, minimum and maximum values and using n (%) of each count category where appropriate. For each variable, n (%) of missing data was reported where appropriate.

For severe CV events, patients were followed up from index date, up until (a) 1 March 2023 (administrative right censoring); (b) the end of a specific time frame (ie, 1 year, 3 years, 5 years post index date); (c) loss to follow-up; (d) death or (e) any severe CV events, whichever occurred first. The crude incidence rate of the first severe CV event of any type and of each individual type per 100 person-years in the full COPD cohort was reported overall and separately by each time period of exposure and non-exposure with the corresponding 95% CI. Time-dependent Cox proportional hazard models were used to estimate the association between each exposed time period, as binary indicators, following a moderate or severe exacerbation and the outcome of time to the first severe CV event, compared with the unexposed time period. Analyses were carried out for time to first severe CV event of any type and for separate categories of severe CV events. Separate models were also fitted for moderate exacerbations and for severe exacerbations. All models were fitted with and without adjustment for confounders. Confounders were selected from the baseline covariates shown in online supplemental table 2, specifically, time-varying covariates were updated at the time of exacerbation and severe CV events. HRs of the first severe CV event between each exposed time period and unexposed time period prior to an exacerbation were estimated with their corresponding 95% CIs. Data were processed using SQL and R (V.4.2.2+), with no imputation for missing data.

For cardiopulmonary events, the crude incidence rate of the first cardiopulmonary event of any type and of each individual type per 100 person-years in the full COPD cohort was reported overall and separately by COPD exacerbation history (no exacerbation, 1 moderate exacerbation, ≥2 moderate exacerbations, ≥1 severe exacerbation) at the 24-month pre-index baseline period. The crude incidence rate of a cardiopulmonary event was reported in different time periods during follow-up: 1 year, 3 years and 5 years and overall.

Time intervals between exacerbations and subsequent cardiopulmonary events were analysed, accounting for recurrent events (eg, intervals between sequential severe exacerbations). Time intervals between exacerbations and subsequent cardiopulmonary events were analysed, with four handling rules for recurrent events:

1. If a patient had multiple exacerbations (not severe exacerbation simultaneously) before multiple cardiopulmonary events, only the time interval after the latest exacerbation and before the first cardiopulmonary event was considered.

2. If a patient had multiple exacerbations and cardiopulmonary sequences, all time intervals between exacerbation and the following cardiopulmonary event were included in the analysis.

3. If a patient had two severe exacerbations during the follow-up period, the time interval between the first severe exacerbation and the second severe exacerbation was included since severe exacerbation is one of the definitions of cardiopulmonary events.

4. If a patient had three or more severe exacerbations in a row during the follow-up period, the time intervals between the prior severe exacerbation and next severe exacerbation were included in the analysis.

Patient and public involvement statement

The study was a retrospective cohort study using a secondary database. Patients were not involved in the design, conduct, reporting or dissemination plans of this research.

Results

From 1 January 2014 to 1 March 2022, 14 713 patients with COPD were included in the final analysis (table 1). The mean (SD) age was 72.0 (11.5) years, and 9862 (67.0%) were male. Comorbid asthma was present in 6142 (41.7%) patients, and 12 409 (84.3%) were on cardiac medications (baseline characteristics in online supplemental table 3). During a median (IQR) follow-up of 2.8 (4.0) years, 7017 (47.7%) patients experienced ≥1 exacerbation of any severity; 7696 (52.3%) had no exacerbations.

Table 1. Patient attrition flowchart.

Step	Description	Number
0	All patients in the database between 1 January 2014 and 1 March 2022 (screening period)	8 214 735*
1	Patients with at least one COPD diagnosis during screening period	44 697
2	Patients with at least two diagnoses for COPD in the outpatient setting or at least one diagnosis for COPD in the inpatient setting within the screening period	17 087
3	Age ≥40 years on the first identified diagnosis (index date)	16 820
4	Data continuously available for at least 24 months before index date	14 713
5	Patients with a record of alpha-1 antitrypsin deficiency during the baseline period will be excluded	14 713

^*8 214 735 is the number of patients that had at least one clinical record in the database during the screening period, including 1 807 835 registered residents of Yinzhou, who had continual contact with the hospital system.

COPD, chronic obstructive pulmonary disease.

Incidence of severe CV events

During follow-up, 2394/11 896 (20.1%) patients experienced severe CV events, with a crude incidence rate of 5.48 (95% CI 5.26 to 5.70) per 100 person-years (table 2). Heart failure decompensation (11.7%) was the most common event, followed by arrhythmias (8.9%), cerebral ischaemia (5.6%), CV-related death (4.6%) and acute coronary syndrome (0.3%). Crude incidence rates per 100 person-years were the highest for heart failure decompensation (3.09; 95% CI 2.94 to 3.25) and the lowest for acute coronary syndrome (0.08; 95% CI 0.06 to 0.11).

Table 2. Incidence rate of severe CV events during follow-up.

	Total N*	N (%)	Crude incidence rate (95% CI)†, per 100 person-years
Any severe CV events	11 896	2394 (20.1%)	5.48 (5.26 to 5.70)
Acute coronary syndrome	14 647	48 (0.3%)	0.08 (0.06 to 0.11)
Heart failure decompensation	13 016	1521 (11.7%)	3.09 (2.94 to 3.25)
Cerebral ischaemia	13 854	769 (5.6%)	1.45 (1.35 to 1.56)
Arrhythmias	13 403	1199 (8.9%)	2.36 (2.23 to 2.50)
CV-related death‡	14 710	682 (4.6%)	1.19 (1.11 to 1.29)

^*Total number of patients (N) for different CV events varied as patients who (1) had the specific severe CV event any time prior to or on the index date or (2) died on the index date or (3) were lost to follow-up on the index date were excluded from the incidence rate calculation.

^†The crude incidence rate was defined as the number of CV events within a specific time frame per 100 person-years during follow-up.

^‡Mortality data were only available until 30 June 2022.

CV, cardiovascular.

Association between COPD exacerbations and severe CV events

The risk of experiencing severe CV events was analysed concerning periods following COPD exacerbations. The crude incidence rates of the first severe CV event of any type—with moderate or severe exacerbation during follow-up as exposure—are detailed in online supplemental table 4. Generally, the incidence rates were higher during exposed periods compared with unexposed periods. The results with moderate exacerbation as the exposure and severe exacerbation as the exposure are shown in online supplemental table 5 and online supplemental table 6, respectively, with severe exacerbations showing higher crude incidence rates of severe CV events compared with moderate exacerbations.

The association between severe CV events and time periods after COPD exacerbation of any severity, after moderate COPD exacerbation and after severe COPD exacerbation is shown in table 3. Compared with unexposed periods, the adjusted HR for any severe CV event was the highest within the first 10 days post exacerbation (10.00; 95% CI 8.16 to 12.25) and remained elevated up to 90 days. Beyond 90 days post exacerbation of any severity, there was still a trend of increased risk of severe CV events, although effects were not significant.

Table 3. Covariates adjusted HRs for the time to the severe CV events together or individually, comparing different time periods following COPD exacerbation to the unexposed period.

Severe CV event	Time periods following an exacerbation (vs unexposed period*)	Moderate or severe exacerbation during follow-up as exposure		Moderate exacerbation during follow-up as exposure		Severe exacerbation during follow-up as exposure
		aHR† (95% CI)	P value	aHR† (95% CI)	P value	aHR† (95% CI)	P value
Any severe CV event	Unexposed (reference)
	1–10 days	10.00 (8.16 to 12.25)	<0.01**	7.94 (6.10 to 10.35)	<0.01**	17.17 (12.64 to 23.31)	<0.01**
	11–30 days	2.23 (1.67 to 2.98)	<0.01**	2.05 (1.43 to 2.93)	<0.01**	3.20 (1.97 to 5.19)	<0.01**
	31–90 days	1.57 (1.25 to 1.98)	<0.01**	1.52 (1.15 to 2.01)	<0.01**	2.09 (1.39 to 3.14)	<0.01**
	91–180 days	1.24 (0.98 to 1.58)	0.08	1.29 (0.97 to 1.70)	0.08	1.36 (0.86 to 2.16)	0.19
	181–365 days	1.18 (0.96 to 1.46)	0.12	1.27 (1.00 to 1.61)	0.05	1.13 (0.75 to 1.72)	0.55
	>365 days	1.06 (0.90 to 1.26)	0.46	1.10 (0.91 to 1.32)	0.34	1.21 (0.90 to 1.62)	0.20
Acute coronary syndrome	Unexposed (reference)
	1–10 days‡	9.61 (1.11 to 83.34)	0.04*	10.96 (1.18 to 102.07)	0.04*	–	–
	11–30 days‡	–	–	–	–	–	–
	31–90 days	9.69 (2.68 to 34.96)	<0.01**	6.72 (1.39 to 32.45)	0.02*	24.68 (2.90 to 209.76)	<0.01**
	91–180 days	4.46 (1.05 to 18.93)	0.04*	3.42 (0.61 to 19.28)	0.16	9.77 (0.77 to 124.26)	0.08
	181–365 days	3.14 (0.74 to 13.43)	0.12	2.39 (0.42 to 13.57)	0.33	7.21 (0.56 to 92.44)	0.13
	>365 days	2.09 (0.60 to 7.30)	0.25	1.60 (0.37 to 6.86)	0.53	4.80 (0.55 to 41.59)	0.15
Heart failure decompensation	Unexposed (reference)
	1–10 days	21.53 (16.96 to 27.33)	<0.01**	15.82 (11.54 to 21.70)	<0.01**	38.20 (27.31 to 53.43)	<0.01**
	11–30 days	3.32 (2.27 to 4.84)	<0.01**	3.17 (2.00 to 5.03)	<0.01**	4.16 (2.21 to 7.81)	<0.01**
	31–90 days	1.84 (1.33 to 2.54)	<0.01**	1.84 (1.25 to 2.70)	<0.01**	2.09 (1.19 to 3.68)	0.01*
	91–180 days	1.58 (1.15 to 2.17)	<0.01**	1.74 (1.20 to 2.51)	<0.01**	1.35 (0.72 to 2.52)	0.35
	181–365 days	1.28 (0.95 to 1.71)	0.10	1.22 (0.86 to 1.74)	0.26	1.57 (0.97 to 2.55)	0.07
	>365 days	0.88 (0.70 to 1.10)	0.27	0.88 (0.68 to 1.15)	0.34	1.02 (0.71 to 1.46)	0.94
Cerebral ischaemia	Unexposed (reference)
	1–10 days	5.18 (3.21 to 8.36)	<0.01**	3.18 (1.58 to 6.41)	<0.01**	9.33 (4.88 to 17.83)	<0.01**
	11–30 days	1.91 (1.10 to 3.30)	0.02*	1.81 (0.92 to 3.54)	0.08	2.03 (0.80 to 5.14)	0.14
	31–90 days	2.14 (1.48 to 3.10)	<0.01**	1.26 (0.74 to 2.13)	0.40	4.22 (2.55 to 6.98)	<0.01**
	91 - 180 days	1.43 (0.96 to 2.15)	0.08	1.28 (0.78 to 2.10)	0.33	1.73 (0.89 to 3.35)	0.10
	181–365 days	1.58 (1.13 to 2.21)	0.01*	1.51 (1.01 to 2.25)	0.04*	1.67 (0.95 to 2.95)	0.08
	>365 days	1.37 (1.04 to 1.81)	0.03*	1.17 (0.84 to 1.63)	0.34	1.90 (1.25 to 2.90)	<0.01**
Arrhythmias	Unexposed (reference)
	1–10 days	13.25 (9.83 to 17.87)	<0.01**	11.40 (7.85 to 16.53)	<0.01**	17.00 (10.65 to 27.12)	<0.01**
	11–30 days	1.78 (1.06 to 3.00)	0.03*	1.67 (0.89 to 3.13)	0.11	1.94 (0.78 to 4.85)	0.15
	31–90 days	1.80 (1.27 to 2.56)	<0.01**	1.65 (1.08 to 2.51)	0.02*	2.05 (1.13 to 3.72)	0.02*
	91–180 days	1.52 (1.07 to 2.16)	0.02*	1.48 (0.98 to 2.23)	0.06	1.52 (0.80 to 2.88)	0.20
	181–365 days	1.36 (1.00 to 1.85)	0.05	1.42 (1.00 to 2.02)	0.05	1.07 (0.58 to 1.98)	0.82
	>365 days	1.14 (0.90 to 1.44)	0.28	1.11 (0.85 to 1.46)	0.44	1.10 (0.76 to 1.61)	0.61
CV-related death§	Unexposed (reference)
	1–10 days	11.39 (6.76 to 19.20)	<0.01**	15.29 (7.89 to 29.64)	<0.01**	8.47 (3.71 to 19.34)	<0.01**
	11–30 days	6.93 (4.19 to 11.46)	<0.01**	5.42 (2.50 to 11.75)	<0.01**	9.04 (4.69 to 17.43)	<0.01**
	31–90 days	4.71 (3.08 to 7.19)	<0.01**	4.00 (2.17 to 7.36)	<0.01**	6.01 (3.41 to 10.59)	<0.01**
	91–180 days	3.59 (2.29 to 5.62)	<0.01**	4.10 (2.29 to 7.34)	<0.01**	3.40 (1.74 to 6.63)	<0.01**
	181–365 days	3.67 (2.48 to 5.45)	<0.01**	5.16 (3.19 to 8.36)	<0.01**	2.35 (1.21 to 4.57)	0.01*
	>365 days	3.70 (2.69 to 5.08)	<0.01**	4.14 (2.77 to 6.21)	<0.01**	3.69 (2.37 to 5.73)	<0.01**

*p<0.05; **p<0.01

^*Unexposed period is time prior to first exacerbation.

^†Adjusted for baseline covariates including year of cohort entry, age, gender, education level, prevalent COPD indication during baseline, comorbidities any time prior to index date including index date (obesity, diabetes mellitus type 1 or diabetes mellitus type 2, any disorders of lipoprotein metabolism and other lipidaemias, ischaemic heart diseases, hypertensive diseases, heart failure, venous thromboembolism, arrhythmias, current asthma, chronic kidney disease, mental illness); time-varying covariates updated at time of exacerbation, including the number of moderate or severe exacerbations, long-acting and short-acting COPD drugs, theophylline, cardiac drugs within last 12 months at time of exacerbation; competing risk, including other severe CV outcomes, was updated on occurrence when examining the individual CV event.

^‡There were very few patients having acute coronary syndrome during certain time periods following an exacerbation, and the HR from the models is close to 0. Therefore, the results were not presented.

^§Mortality data were only available until 30 June 2022.

aHR, adjusted HR; COPD, chronic obstructive pulmonary disease; CV, cardiovascular.

Severe exacerbations conferred greater risk than moderate exacerbations (adjusted HR at 10 days: 17.17 (95% CI 12.64 to 23.31) vs 7.94 (95% CI 6.10 to 10.35)), and the overall risk for both groups remained significantly greater than unexposed periods for up to 90 days post exacerbation.

Regarding individual severe CV event (heart failure, cerebral ischaemia, arrhythmias, CV-related death), the risk remained significantly elevated for varying durations post exacerbation, with acute coronary syndrome, heart failure decompensation and arrhythmia risks elevated up to 180 days, cerebral ischaemia risk up to 365 days and CV-related death risk beyond 365 days. Compared with moderate exacerbations, the risk of non-fatal severe CV events was higher following severe exacerbations, especially during the first 90 days post exacerbation.

Incidence of cardiopulmonary events

During the follow-up period, 3785 patients (32.7%) experienced cardiopulmonary events (crude incidence rate: 9.38 (95% CI 9.09 to 9.69) per 100 person-years). All-cause death (20.1%) was the most frequent, followed by severe COPD exacerbations (16.7%), CV events (4.7%), ischaemic stroke (2.7%) and tachyarrhythmia/atrial fibrillation (1.4%) (table 4). Crude incidence rates decreased with longer follow-up duration (1 year >3 years >5 years).

Table 4. Incidence rate of cardiopulmonary events during follow-up in all patients.

	Total*	N (%)	Crude incidence rate (95% CI)†, per 100 person-years
Overall
Cardiopulmonary event‡	11 580	3785 (32.7%)	9.38 (9.09 to 9.69)
Severe exacerbation of COPD	12 316	2054 (16.7%)	4.72 (4.52 to 4.93)
All-cause death §	14 710	2951 (20.1%)	5.17 (4.98 to 5.36)
Cardiovascular events ¶	14 195	669 (4.7%)	1.23 (1.14 to 1.33)
Ischaemic stroke	14 404	391 (2.7%)	0.71 (0.64 to 0.78)
Tachyarrhythmia/atrial fibrillation	14 595	204 (1.4%)	0.36 (0.31 to 0.42)
Within 1 year during follow-up
Cardiopulmonary event‡	11 580	1411 (12.2%)	13.53 (12.83 to 14.26)
Severe exacerbation of COPD	12 316	830 (6.7%)	7.47 (6.97 to 7.99)
All-cause death §	14 710	907 (6.2%)	6.62 (6.20 to 7.07)
Cardiovascular events ¶	14 195	180 (1.3%)	1.37 (1.17 to 1.58)
Ischaemic stroke	14 404	101 (0.7%)	0.75 (0.61 to 0.92)
Tachyarrhythmia/atrial fibrillation	14 595	41 (0.3%)	0.30 (0.22 to 0.41)
Within 3 years during follow-up
Cardiopulmonary event‡	11 580	2693 (23.3%)	10.36 (9.97 to 10.76)
Severe exacerbation of COPD	12 316	1499 (12.2%)	5.38 (5.11 to 5.66)
All-cause death §	14 710	1893 (12.9%)	5.39 (5.15 to 5.64)
Cardiovascular events ¶	14 195	404 (2.9%)	1.20 (1.09 to 1.32)
Ischaemic stroke	14 404	236 (1.6%)	0.69 (0.60 to 0.78)
Tachyarrhythmia/atrial fibrillation	14 595	122 (0.8%)	0.35 (0.29 to 0.42)
Within 5 years during follow-up
Cardiopulmonary event‡	11 580	3323 (28.7%)	9.60 (9.28 to 9.94)
Severe exacerbation of COPD	12 316	1816 (14.8%)	4.88 (4.66 to 5.11)
All-cause death §	14 710	2504 (17.0%)	5.23 (5.03 to 5.44)
Cardiovascular events ¶	14 195	534 (3.8%)	1.17 (1.07 to 1.27)
Ischaemic stroke	14 404	321 (2.2%)	0.69 (0.62 to 0.77)
Tachyarrhythmia/atrial fibrillation	14 595	173 (1.2%)	0.37 (0.31 to 0.42)

^*For each outcome, this total number of patients for different cardiopulmonary events varied as patients who (1) had the specific cardiopulmonary event any time prior to or on the index date or (2) died on the index date or (3) were lost to follow-up on the index date were excluded from the incidence rate calculation.

^†per 100 person-years.

^‡All cardiopulmonary events were identified in the inpatient setting except for mortality.

^§The all-cause death referred to all deaths during the follow-up period (starting from 1 day after index date). Mortality data were only available until 30 June 2022.

^¶Cardiovascular events (inpatient setting): acute heart failure, acute myocardial infarction, angina pectoris, cardiac arrest and other ischaemic heart diseases/acute coronary syndrome.

COPD, chronic obstructive pulmonary disease.

Higher baseline exacerbation frequency correlated with increased cardiopulmonary event risk: 28.8% (no exacerbations), 38.9% (one moderate exacerbation) and 54.6% (≥2 moderate exacerbations). Severe baseline exacerbations were associated with the highest rates of all-cause death and non-fatal events (online supplemental table 7).

The median (IQR) interval between COPD exacerbation (any severity) and cardiopulmonary events was 4.17 (11.30) months, with shorter intervals after moderate exacerbations (4.30 (13.58) months) compared with severe exacerbations (5.88 (13.07) months) (table 5).

Table 5. Time interval between exacerbation of COPD and cardiopulmonary events during follow-up in all patients.

	Number of cases	Number of patients	Time interval in months, median (IQR)
Any exacerbation
Any time after exacerbation	5321	2897	4.17 (11.30)
Within 3 months after exacerbation	2314	1450	0.89 (1.48)
Within 6 months after exacerbation	3059	1759	1.38 (2.50)
Within 1 year after exacerbation	3941	2163	2.17 (4.76)
Within 2 years after exacerbation	4630	2485	3.02 (7.46)
Within 3 years after exacerbation	4962	2665	3.61 (9.13)
Severe exacerbation
Any time after exacerbation	3713	1770	5.88 (13.07)
Within 3 months after exacerbation	1289	782	1.28 (1.28)
Within 6 months after exacerbation	1871	1002	1.97 (2.51)
Within 1 year after exacerbation	2563	1281	2.99 (5.09)
Within 2 years after exacerbation	3137	1493	4.24 (8.38)
Within 3 years after exacerbation	3404	1605	5.01 (10.16)
Moderate exacerbation
Any time after exacerbation	2976	2247	4.30 (13.58)
Within 3 months after exacerbation	1319	1032	0.46 (1.31)
Within 6 months after exacerbation	1666	1252	0.92 (2.52)
Within 1 year after exacerbation	2122	1581	1.71 (4.99)
Within 2 years after exacerbation	2523	1876	2.73 (8.25)
Within 3 years after exacerbation	2732	2042	3.48 (10.25)

COPD, chronic obstructive pulmonary disease.

Discussion

This retrospective longitudinal cohort study investigated the risk of severe CV events in COPD patients during distinct post-exacerbation periods and explored cardiopulmonary event patterns. Key findings include: 20.1% of patients experienced severe CV events over a mean follow-up of 3.2 years (crude incidence: 5.48/100 person-years). The risk of severe CV events peaked within the first 10 days post exacerbation, remained elevated up to 90 days and gradually declined but stayed significant for 1 year compared with pre-exacerbation periods. These results align with prior studies but provide novel temporal granularity.

The IMPACT trial reported heightened CV risk during and within 30 days post exacerbation,⁸ while the EXACOS-CV studies found prolonged risk profiles: the US cohort noted elevated CV risk for 1 year (peaking at 30 days),¹⁸ the Netherlands study reported 180-day risks after moderate exacerbations and 1-year risks after severe ones¹⁴ and the Canadian cohort observed risks persisting beyond 1 year.¹⁹ Another population-based study in England investigating temporal association between non-fatal CV events and COPD exacerbation also revealed that an immediate period of CV relative rate after any exacerbation (1–14 days), followed by progressively declining yet maintaining elevated rates after 1 year.²⁰ Our study complements these findings but refines the temporal risk stratification using shorter intervals (1–10 days, 11–30 days, 31–90 days, etc), offering sharper insights into early post-exacerbation vulnerability.

Mechanistically, COPD exacerbations may drive CV risk through systemic inflammation (elevated cytokines, endothelial dysfunction, plaque instability),²¹,²³ respiratory stress (hypoxaemia, hypercapnia, haemodynamic strain)²⁴,²⁶ and potential medication effects.^{27 28} These mechanisms affect CV event types differently. Heart failure decompensation is the most frequent probably because exacerbations imposed significant fluid and pressure load on already compromised ventricles.^{19 29} Arrhythmias are also common, driven by hypoxaemia, and acid-base imbalance destabilises cardiac conduction.^{19 30} In contrast, cerebral ischaemia and CV-related death occur less often,³¹ reflecting secondary effects of hypoxaemia and hypercoagulability, while acute coronary syndrome remains rare as plaque rupture is less likely directly triggered by short-term respiratory stress.^{32 33} These findings underscored the need for vigilant CV monitoring post exacerbation.

The concept of cardiopulmonary events has evolved and gained increasing attention in COPD research. It is widely acknowledged that COPD is a disease affecting both the CV and respiratory systems of the body. To the best of our knowledge, the concept of cardiopulmonary event in the real-world setting is introduced and studied in China for the first time, which offered a valuable perspective that complemented and extended the findings from randomised clinical trial environments to real-world scenarios. In a post-hoc analysis of the IMPACT trial,³⁴ the time to on-treatment cardiopulmonary events was investigated. Our focus diverged from the IMPACT trial by examining the incidence rate of cardiopulmonary events during the follow-up period. It was found that about one in three patients with COPD experienced cardiopulmonary events in this study with a median follow-up duration of approximately 3 years. The occurrence of cardiopulmonary events, especially severe exacerbation and death, increased with the frequency and severity of COPD exacerbation history. This indicates that patients with COPD who have more frequent and severe prior exacerbations are more likely to experience subsequent severe COPD exacerbations, which aligns with previous studies.^{35 36} It was also found that the median time interval between severe exacerbation and cardiopulmonary events was longer than that between moderate exacerbation(s) and cardiopulmonary events, which was counterintuitive. It might be due to patients who experience a severe exacerbation(s) being likely to receive more intensive treatments while in a hospital setting, which could delay or alter the onset of a subsequent cardiopulmonary event(s).³⁷

Many patients with COPD are comorbid with CV-related diseases, such as hypertension, coronary artery disease and heart failure. The exacerbation-related physiological stressors and inflammatory responses can worsen these pre-existing CV conditions, leading to an increased risk of decompensation and acute severe CV events, as stated previously. In clinical practice, close monitoring immediately after exacerbations is essential, particularly for patients with known CV comorbidities. Parameters such as oxygen saturation, blood gases, ECG changes and cardiac biomarkers (eg, N‑terminal pro–B‑type Natriuretic Peptide, troponin) may help identify patients at higher risk of heart failure or arrhythmias. Addressing these factors can help healthcare professionals in enhancing positive outcomes and alleviating the impact of CV complications among patients with COPD. Therefore, healthcare providers should implement strategies to manage systemic inflammation, optimise oxygenation, carefully monitor CV risk factors and medication and assess medication effects to reduce the risk of CV and cardiopulmonary events. While our study did not establish new predictive markers, future research should focus on developing risk stratification models and validating biomarkers to guide preventive strategies. Additionally, integrated care models and syndemics approaches that address both respiratory and CV aspects of COPD management are essential for comprehensive and effective patient care.³⁸,⁴⁰

Limitations

Our study has several limitations. First, measurement bias may exist due to potential misclassification of exposures and outcomes. For instance, symptoms of COPD exacerbation (eg, dyspnoea) could overlap with those of heart failure decompensation, creating diagnostic uncertainty. To mitigate this, the first 1–10 days post exacerbation were analysed cautiously to avoid conflating exacerbations with acute CV events. Furthermore, outpatient visits with new respiratory prescriptions might include patients without true exacerbations (eg, those with pneumonia or colds), potentially diluting the moderate exacerbation subgroup. Second, the unexposed group may include healthier individuals who never experienced an exacerbation, which could bias (likely overestimate) the HR estimates, even though multiple covariates were adjusted in the models. Future studies should consider advanced approaches such as propensity score methods or marginal structural models to further address this issue. Third, the incidence of acute coronary syndrome (ACS) in our COPD cohort (0.3%) was lower than rates reported in international EXACOS-CV studies.^{14 41} This discrepancy may stem from our case identification methodology: we used keywords like “acute coronary syndrome” and “unstable angina” but excluded specific terms such as “ST-elevation myocardial infarction” or “non-ST-elevation myocardial infarction.” This omission likely led to under capturing ACS cases. Fourth, data were based on cumulative counts of patients identified during screening and follow-up, which included both permanent residents and temporary ‘visiting’ patients. Transient patients might have had incomplete long-term exposure or outcome data, potentially affecting generalisability. In addition, the mortality data from the Yinzhou database were available up to 30 June 2022. In late 2022, China experienced a prolonged COVID-19 outbreak.⁴² There was conflicting evidence regarding the impact of COVID-19 on COPD patients. Some studies suggested that COPD might confer a protective effect against contracting COVID-19 because of the isolation and quarantine policies during the outbreak,^{43 44} while other research demonstrated that COVID-19 infection was associated with greater severity and mortality rates in COPD patients.⁴⁵ As a result, the potential impact of the absence of mortality data during this period was uncertain. Lastly, our findings were derived from a regional database in Yinzhou, a developed district of East China, where COPD burden and clinical features may differ from other regions in China. Regional variations in diagnosis pathways, population characteristics and local policies should be considered when applying these results beyond Yinzhou. Internationally, differences in smoking patterns, environmental exposures, genetics and health-system resources may also shape the magnitude of risks. Nonetheless, the temporal pattern observed in the current study had been consistently reported in studies from China and other EXACOS-CV countries, supporting the broader relevance of our finding.²⁴⁶,⁵² In light of these limitations, extrapolating our findings to broader populations requires caution, particularly regarding diagnostic misclassification, ACS case definitions, transient patient inclusion and regional disparities.

Conclusion

In this large Chinese cohort, COPD exacerbations acutely increased severe CV risk, with residual effects persisting for 1 year. Cardiometabolic and respiratory risks were intertwined, emphasising the need for integrated care. Early post-exacerbation monitoring and dual cardiopulmonary management are critical to improving outcomes.

Data availability statement

Data may be obtained from a third party and are not publicly available.