Introduction
Preserved ratio impaired spirometry (PRISm), defined as a post-bronchodilator forced expiratory volume in one second (FEV1)/forced vital capacity (FVC) ratio ≥0.70 with FEV1 <80% predicted, does not meet the spirometric criterion for airflow obstruction used to define chronic obstructive pulmonary disease (COPD).1 Based on the presence or absence of a restrictive spirometric abnormality, PRISm is further classified into restrictive PRISm (FEV1/FVC ≥0.70, FEV1 <80% predicted, and FVC <80% predicted) and non-restrictive PRISm (FEV1/FVC ≥0.70, FEV1 <80% predicted, and FVC ≥80% predicted).2 Several longitudinal studies have shown that individuals with PRISm may follow heterogeneous spirometric trajectories, including recovery to normal spirometry, persistence of PRISm, or progression to COPD.1,3,4 Accordingly, PRISm has increasingly been recognized as an important pre-COPD phenotype, characterized by respiratory symptoms, structural lung abnormalities, and impaired lung function overlapping with COPD before the development of airflow obstruction.5 Individuals with greater symptom burden and more pronounced radiographic abnormalities are more likely to follow adverse disease trajectories.6
Chronic obstructive pulmonary disease frequently coexists with extrapulmonary comorbidities, among which cardiovascular disease (CVD) is one of the most important. COPD and CVD are closely interconnected through shared risk factors and overlapping pathophysiological mechanisms that may influence disease progression and overall prognosis.7 In COPD, comorbid CVD is associated with higher risks of acute exacerbations, hospitalization, and mortality.8 Conversely, COPD and acute exacerbations may also increase the risk of cardiovascular events,9,10 particularly in the early period after an exacerbation, when the risks of myocardial infarction, heart failure, and stroke are substantially elevated.11
Similarly, PRISm is also commonly accompanied by a substantial burden of comorbidities. Large cohort studies have shown that PRISm is associated with more comorbidities than normal spirometry and, in some studies, even more than COPD.12 Cardiovascular and metabolic diseases are particularly prevalent in this population. Compared with normal spirometry, PRISm has been established as linked to higher risks of cardiovascular events and mortality,13 with these risks appearing more pronounced in restrictive PRISm.14 Furthermore, longitudinal spirometric transitions from PRISm are closely related to prognosis.
Persistent PRISm and progression to COPD are associated with elevated risks of cardiopulmonary disease and mortality, whereas recovery to normal spirometry suggests a more favorable prognosis.15
Although previous studies have primarily focused on the cardiovascular consequences of PRISm, growing evidence indicates a close relationship between impaired lung function and cardiovascular disease. While PRISm has been linked to an increased risk of cardiovascular events, the presence of comorbid CVD may identify individuals with a greater risk of subsequent respiratory deterioration. Several mechanisms, including cardiac dysfunction, hemodynamic alterations, reduced exercise capacity, and medication-related effects, may contribute to this association.
Therefore, assessment of CVD may provide additional information for risk stratification among individuals with PRISm.
To date, only a limited number of studies using composite cardiovascular health measures have suggested that poorer cardiovascular status may be associated with increased PRISm development, reduced recovery, and higher mortality.15 Furthermore, the attenuation of the association between PRISm and incident airflow obstruction after adjustment for CVD-related factors suggests a potential role of cardiovascular comorbidity in subsequent spirometric transitions.16 However, direct longitudinal evidence remains limited regarding whether baseline CVD can identify PRISm individuals at higher risk of adverse respiratory outcomes, including exacerbations, accelerated lung function decline, and progression to COPD.
Accordingly, we included individuals with PRISm at baseline and examined the associations of CVD with acute exacerbations, longitudinal lung function changes, and progression to COPD. CVD was defined as a composite measure of coronary heart disease, congestive heart failure, and stroke to reflect overall cardiovascular comorbidity burden, although individual CVD subtypes may influence respiratory outcomes through distinct pathophysiological pathways. We further performed analyses stratified by restrictive and non-restrictive PRISm to determine whether these associations varied by spirometric subtypes.
This study aimed to clarify the prognostic relevance of CVD in PRISm and its potential value in clinical risk assessment.
Methods
Study Population
This retrospective cohort study enrolled individuals who underwent pulmonary function testing at a hospital in Beijing between 2021 and 2023. Eligible participants were aged ≥40 years, had complete post-bronchodilator spirometric data, and met the Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria for PRISm, defined as a post-bronchodilator FEV1/FVC ratio ≥0.70 with FEV1 <80% predicted.16 Baseline was defined as the date of the first pulmonary function test meeting the PRISm criteria. After completion of the follow-up period, clinical records, follow-up spirometric data, and outcome events from baseline through December 31, 2025 were retrospectively retrieved from the electronic medical record system.
To ensure reliable longitudinal assessment, participants were required to have at least two pulmonary function tests performed ≥6 months apart and a clinical follow-up duration of ≥1 year. To evaluate potential selection bias related to the longitudinal inclusion criteria, baseline characteristics were compared between included PRISm participants and those who met baseline eligibility criteria but were excluded due to insufficient longitudinal follow-up (Table S1). Individuals with missing baseline information on cardiovascular disease or key covariates were also excluded.
CVD status was determined through medical record review of previous diagnoses, laboratory results, and relevant examination findings. Two physicians independently assessed CVD status according to established clinical guidelines, with disagreements resolved by discussion. A composite CVD measure was defined as the presence of coronary heart disease, congestive heart failure, or stroke.
The composite CVD definition was used to capture the overall burden of major cardiovascular comorbidities in relation to respiratory outcomes among individuals with PRISm. Participants with any of these conditions were classified as having CVD. Detailed information regarding CVD severity, duration, and treatment was unavailable due to incomplete documentation.
Among 8187 individuals with available spirometry data between 2021 and 2023, 7544 were excluded due to age <40 years (n=2333), not meeting the spirometric criteria for PRISm (n=4924), or missing baseline CVD status or key covariate information (n=287). Given the relatively small proportion of participants with missing baseline information (287/8187, 3.5%), a complete-case approach was used for the primary analysis. Among the remaining 643 participants with PRISm and complete baseline data, 122 were further excluded because of insufficient longitudinal follow-up, including fewer than two spirometric measurements (n=71) or a follow-up duration shorter than 1 year (n=51).
Finally, 521 participants with PRISm were included in the final analysis (Figure 1). The study protocol was approved by the Ethics Committee of Beijing Chao-Yang Hospital, Capital Medical University (Ethics No. 2021-ke-34).
Given the retrospective design and the use of anonymized data, the requirement for written informed consent was waived. The study was conducted in accordance with the Declaration of Helsinki. Figure 1Flowchart of participant selection.Flowchart sequence: 8187 participants with available spirometry records from 2021 to 2023.
Excluded left parenthesis n equals 7544 right parenthesis: Age less than 40 years old left parenthesis n equals 2333 right parenthesis; Not meeting PRISm spirometric criteria left parenthesis n equals 4924 right parenthesis; Missing baseline CVD status or key covariate information left parenthesis n equals 287 right parenthesis. Then 643 PRISm participants with complete baseline data. Excluded left parenthesis n equals 122 right parenthesis: less than 2 spirometry measurements left parenthesis n equals 71 right parenthesis; less than 1 year of follow up left parenthesis n equals 51 right parenthesis.
Then 521 participants were included in the final analysis. A vertical timeline at left marks 2021 left parenthesis Baseline right parenthesis, 2022, 2023 and 2025 left parenthesis End of follow up right parenthesis.A flowchart of participant selection from 8187 records to 521 included in the final analysis.Abbreviations: PRISm, preserved ratio impaired spirometry; CVD, cardiovascular disease.
Baseline Characteristics and Study Variables
Baseline characteristics, pulmonary function results, and other study variables were extracted from the hospital electronic medical record system, pulmonary function database, and related information systems. Collected variables included demographic characteristics (age, sex, body mass index [BMI] and smoking history), respiratory symptoms (cough, sputum production, dyspnea, and wheezing), comorbidities (cardiovascular disease, hypertension, diabetes, and hyperlipidemia), laboratory measures (blood cell count, hemoglobin, albumin, and high-density lipoprotein [HDL]), chest computed tomography (CT) findings (emphysema, interstitial lung abnormalities, and bronchiectasis), and spirometric indices. Smoking status was obtained from electronic medical records and categorized as ever smoking or never smoking based on available documentation.
Detailed quantitative smoking exposure data, including smoking duration, daily cigarette consumption, and pack-years, were unavailable. Medication information, including cardiovascular and respiratory therapies, was not included because available records lacked consistent information regarding drug type, initiation time, and treatment duration across participants, precluding reliable assessment of medication effects.
Spirometry at baseline and during follow-up was performed by trained technicians according to standardized procedures and quality-control protocols. Unless otherwise indicated, all spirometric indices analyzed in this study were post-bronchodilator values, including FEV1, FVC, FEV1% predicted, FVC% predicted, and FEV1/FVC.
Outcomes
The outcomes of interest included acute exacerbations, longitudinal lung function changes, and incident COPD during follow-up. Acute exacerbations were defined as clinically significant worsening of respiratory symptoms requiring additional treatment and were classified as moderate or severe. Potential exacerbation events identified from the electronic medical records were independently reviewed by two physicians according to guideline-based criteria, and discrepancies in event identification or severity classification were resolved through discussion.
Moderate exacerbation referred to an event requiring systemic antibiotics and/or corticosteroids, whereas severe exacerbation referred to an event requiring an emergency department visit or hospitalization. Mild exacerbations that did not require additional treatment were not included since they could not be reliably identified from retrospective medical records. Annualized exacerbation rate was calculated as the number of events divided by the corresponding follow-up time in years.
Frequent exacerbation was defined as ≥2 moderate or severe exacerbations per year. As exacerbation events were based on clinical documentation, their ascertainment may have been influenced by healthcare utilization patterns.
Longitudinal lung function changes were assessed using at least two post-bronchodilator spirometric measurements and expressed as annualized changes in spirometric parameters. Incident COPD was identified by either a post-bronchodilator FEV1/FVC ratio <0.70 during follow-up or a physician diagnosis of COPD recorded in the electronic medical record. Time to incident COPD was calculated from baseline to the first qualifying event.
Participants without incident COPD were censored at the date of the last available follow-up.
Statistical Analysis
Participants were categorized by baseline CVD status, and baseline characteristics were compared between groups. Continuous variables are presented as mean ± standard deviation (SD) or median [interquartile range (IQR)] according to distribution, and categorical variables as counts and percentages, n (%). Comparisons between groups were performed using independent-samples t tests, Wilcoxon rank-sum tests, χ2-tests, or Fisher’s exact tests, as appropriate.
Logistic regression was used to identify factors related to CVD and evaluate the association between CVD and frequent acute exacerbations, with results reported as odds ratios (ORs) and 95% confidence intervals (CIs). Longitudinal lung function change was assessed using linear mixed-effects models with years from baseline as the time variable. Each model included CVD status, follow-up time, and a CVD × time interaction term to estimate between-group differences in annualized lung function changes.
Models were adjusted for age, sex, body mass index, smoking status, and the corresponding baseline spirometric value to account for potential confounding and differences in baseline lung function between participants with and without CVD. A random intercept was included for each participant to account for individual variability and correlations among repeated measurements. The random-effects structure was determined based on the longitudinal study design and model convergence.
All available post-bronchodilator spirometric measurements were incorporated into the mixed-effects models, allowing for unequal numbers of repeated assessments across participants. Missing spirometric observations were handled within the mixed-effects framework without imputation. The linearity of lung function trajectories over time was assessed by comparing linear and quadratic time models, and model assumptions were evaluated using residual plots.
Spirometric transitions during follow-up were categorized as recovery to normal spirometry, persistent PRISm, or progression to COPD and were visualized using Sankey diagrams. Cox proportional hazards models were used to evaluate the association between CVD and incident COPD, with results reported as hazard ratios (HRs) and 95% CIs. Univariable and multivariable analyses were constructed for both logistic and Cox regression models.
Covariates included in multivariable models were selected based on univariable results, previous evidence, and clinical relevance. Multicollinearity was assessed using variance inflation factors (VIFs), and the proportional hazards assumption was evaluated using Schoenfeld residuals. No adjustment for multiple comparisons was performed, as the primary analyses focused on predefined clinically relevant respiratory outcomes, including frequent exacerbations and incident COPD.
Subgroup analyses were considered exploratory.
Subgroup and Interaction Analyses
To determine whether the associations of cardiovascular disease with acute exacerbations and incident COPD varied by PRISm subtype, subgroup analyses were performed in participants with restrictive and non-restrictive PRISm, and effect modification was evaluated using interaction tests.
Sensitivity Analyses
To test the robustness of the findings, several sensitivity analyses were performed. First, baseline characteristics were compared between included and excluded PRISm participants to evaluate potential selection bias related to follow-up requirements. Second, inverse probability of treatment weighting (IPTW) was applied to balance baseline covariates between participants with and without CVD.
Propensity scores were estimated using logistic regression, including age, sex, BMI, smoking status, baseline pulmonary function parameters, respiratory comorbidities, and cardiometabolic comorbidities (hypertension, diabetes, and hyperlipidemia) as covariates. These covariates were selected based on their potential confounding effects and clinical relevance to CVD and respiratory outcomes. Stabilized weights were calculated and truncated at the 1st and 99th percentiles to reduce the influence of extreme values.
Covariate balance after weighting was evaluated using standardized mean differences (SMDs), with an SMD <0.1 indicating adequate balance. Weighted multivariable logistic and Cox proportional hazards models were used to examine the associations of CVD with frequent acute exacerbations and incident COPD, respectively. Third, sensitivity analyses were conducted using severe exacerbations requiring hospitalization as an alternative outcome to reduce the potential impact of differences in healthcare utilization.
In addition, negative binomial regression models were used to analyze the total number of moderate-to-severe exacerbation events during follow-up, with follow-up duration included as an offset term. Finally, the robustness of longitudinal lung function findings was evaluated using linear mixed-effects models without adjustment for baseline spirometric values and additional models incorporating exacerbations as a time-updated covariate to assess whether exacerbation burden during follow-up affected the observed association between CVD and lung function decline.
Software and Statistical Significance
All statistical analyses were performed using R software, version 4.6.0. All tests were two-sided, and a P value <0.05 was considered statistically significant.
Results
Baseline Characteristics
A total of 521 participants with PRISm were included in the analysis, comprising 166 (31.9%) with CVD and 355 (68.1%) without CVD. Baseline characteristics according to CVD status are presented in Table 1.Table 1Baseline Characteristics of Participants with PRISm Categorized According to CVDWith CVD(n=166)Without CVD(n=355)P valueDemographicsAge, years61.8 ± 12.652.4 ± 15.7<0.001Male sex117 (70.5)191 (53.8)<0.001Body mass index, kg/m226.9 ± 4.925.7 ± 5.20.014ExposuresSmoking history0.004Smoker92 (55.4)147 (41.4)Non-smoker74 (44.6)208 (58.6)Lung functionFEV1, L1.81 ± 0.491.92 ± 0.590.040FEV1, %predicted69.9 ± 9.670.0 ± 8.30.860FVC, L2.48 ± 0.642.60 ± 0.800.089FVC, %predicted71.0 ± 13.272.3 ± 16.20.391FEV1/FVC, %75.9 ± 5.477.0 ± 6.10.056Respiratory symptomsCough37 (22.3)105 (29.6)0.102Sputum production15 (9.0)44 (12.4)0.328Dyspnea39 (23.5)56 (15.8)0.045Wheeze26 (15.7)46 (13.0)0.486ComorbiditiesHypertension113 (68.1)153 (43.1)<0.001Diabetes36 (21.7)25 (7.0)<0.001Hyperlipidemia108 (65.1)131 (36.9)<0.001Asthma29 (17.5)49 (13.8)0.336Radiologic findingsEmphysema63 (38.0)100 (28.2)0.032Interstitial lung abnormalities38 (22.9)58 (16.3)0.094Bronchiectasis11 (6.6)43 (12.1)0.078Laboratory findingsNeutrophil-to-lymphocyte ratio2.46 [1.64, 3.30]2.14 [1.62, 3.18]0.109Eosinophil counts, × 109/L0.14 [0.06, 0.28]0.16 [0.09, 0.29]0.086Platelet counts, ×109/L205.00 [177.50, 258.00]252.00 [207.50, 302.50]<0.001Hemoglobin, g/L139.50 [128.25, 152.00]138.00 [126.00, 150.00]0.250Albumin, g/L41.80 [38.80, 44.45]41.85 [38.80, 45.23]0.389HDL cholesterol, mmol/L1.00 [0.87, 1.29]1.15 [0.94, 1.40]0.015Notes: Data are presented as mean ± SD for normally distributed continuous variables, median [IQR] for non-normally distributed continuous variables, and n (%) for categorical variables. Respiratory symptoms were assessed individually as binary variables (presence or absence of each symptom).
All spirometric indices are post-bronchodilator values unless otherwise indicated.Abbreviations: PRISm, preserved ratio impaired spirometry; CVD, cardiovascular disease; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; HDL, high-density lipoprotein; SD, standard deviation; IQR, interquartile range.
Compared with those without CVD, participants with CVD were older (61.8 ± 12.6 vs 52.4 ± 15.7 years, P <0.001), more likely to be male (70.5% vs 53.8%, P <0.001) and smokers (55.4% vs 41.4%, P = 0.004), with a higher BMI (26.9 ± 4.9 vs 25.7 ± 5.2 kg/m2, P = 0.014). Hypertension, diabetes, and hyperlipidemia were also more prevalent in the CVD group (all P <0.01). Regarding respiratory characteristics, emphysema was more common among individuals with CVD (38.0% vs 28.2%, P = 0.032), whereas physician-diagnosed asthma as well as CT findings of interstitial lung abnormalities and bronchiectasis were comparable between groups.
Post-bronchodilator FEV1 was lower in participants with CVD (1.81 ± 0.49 vs 1.92 ± 0.59 L, P = 0.040), while FVC and FEV1/FVC also tended to be lower, although these differences did not reach statistical significance. In laboratory testing, the CVD group showed lower HDL cholesterol levels and platelet counts, along with a slightly higher neutrophil-to-lymphocyte ratio. Respiratory symptoms and protein nutritional indicators did not differ significantly between groups.
Clinical Factors Associated with CVD
Exploratory logistic regression was performed to identify clinical correlates of CVD in participants with PRISm (Table S2). In univariable analyses, age, sex, BMI, smoking status, post-bronchodilator FEV1, platelet count, and several comorbidities were related to CVD. In the multivariable model, older age (adjusted OR, 1.04; 95% CI, 1.02–1.07; P = 0.002), hypertension (adjusted OR, 1.70; 95% CI, 1.04–2.78; P = 0.034), and hyperlipidemia (adjusted OR, 2.16; 95% CI, 1.34–3.49; P = 0.002) remained independently associated with CVD, whereas platelet count showed an inverse association.
Platelet count was included as an exploratory hematological marker, and its inverse association with CVD should be interpreted cautiously because it may be influenced by other clinical factors. No substantial multicollinearity was observed among the covariates.
Association of CVD with Acute Exacerbation Burden
During a mean follow-up of 2.59 ± 1.58 years, participants with CVD had a greater acute exacerbation burden than those without CVD (Table 2A). Annualized rates of severe exacerbations (0.39 ± 0.70 vs 0.14 ± 0.40 events/year, P <0.001) and moderate-to-severe exacerbations (1.03 ± 2.95 vs 0.47 ± 0.64 events/year, P = 0.017) were higher in the CVD group. The annualized rate of moderate exacerbations also increased numerically, although the difference was not statistically significant.
Frequent exacerbators were also more common among participants with CVD, with the prevalence nearly three times that in those without CVD (10.8% vs 3.9%, P = 0.004).Table 2Acute Exacerbation Burden and Association Between Baseline CVD and Frequent Exacerbation in Participants with PRISmA. Acute exacerbation burden according to CVDWith CVD(n=166)Without CVD(n=355)P valueAnnualized rates of AE, events/year Moderate0.64 ± 2.480.34 ± 0.520.114 Severe0.39 ± 0.700.14 ± 0.40<0.001 Moderate-to-severe1.03 ± 2.950.47 ± 0.640.017Frequent exacerbator, n (%)18 (10.8%)14 (3.9%)0.004B. Association between baseline CVD and frequent exacerbationCrude modelModel 1Model 2Model 3OR (95% CI)P valueOR (95% CI)P valueOR (95% CI)P valueOR (95% CI)P valueWithout CVDReferenceReferenceReferenceReferenceWith CVD2.96(1.44,6.11)0.0032.61(1.20,5.68)0.0162.61(1.20,5.68)0.0162.48(1.13,5.47)0.024Notes: In Panel A, annualized exacerbation rates were calculated as exacerbation events divided by follow-up time in years and are presented as mean ± SD (mean follow-up duration, 2.59 ± 1.58 years).
Frequent exacerbators were defined as ≥2 moderate or severe exacerbations per year and are presented as n (%). In Panel B, the crude model was unadjusted; Model 1 was adjusted for age, sex, body mass index, and smoking history; Model 2 was additionally adjusted for baseline lung function (post-bronchodilator FEV1); and Model 3 was further adjusted for emphysema and diabetes.Abbreviations: CVD, cardiovascular disease; PRISm, preserved ratio impaired spirometry; AE, acute exacerbation; OR, odds ratio; CI, confidence interval; SD, standard deviation; FEV1, forced expiratory volume in 1 second.
The association between CVD and frequent exacerbation was further evaluated using logistic regression models (Table 2B). In the crude model, CVD was associated with higher odds of frequent exacerbation (OR, 2.96; 95% CI, 1.44–6.11; P = 0.003). The relationship remained statistically significant after sequential adjustment for demographic characteristics, baseline lung function, and other potential confounders, with similar estimates in Model 1 and Model 2 (adjusted OR, 2.61; 95% CI, 1.20–5.68; P = 0.016).
In the fully adjusted model, CVD remained independently related to frequent exacerbation, corresponding to a 148% increased likelihood of frequent exacerbation (adjusted OR, 2.48; 95% CI, 1.13–5.47; P=0.024).
Consistent with these findings, negative binomial regression models incorporating follow-up duration as an offset term showed that participants with CVD had a higher rate of moderate-to-severe exacerbation events during follow-up (IRR, 1.68; 95% CI, 1.31–2.14; P<0.001; Table S3).
Association of CVD with Longitudinal Spirometric Change
During follow-up, participants contributed different numbers of repeated post-bronchodilator spirometric measurements, ranging from 2 to 8 assessments per participant, with a median of 2 assessments. Table 3 summarizes annualized spirometric changes according to CVD status. In the unadjusted model, participants with CVD showed a faster annual decline in FVC than those without CVD (−5.11 vs 46.56 mL/year, respectively), with a between-group difference of −51.68 mL/year (95% CI, −100.48 to −2.88; P=0.038).
After adjustment for age, sex, BMI, smoking status, and baseline lung function, this difference became more evident (adjusted difference, −58.03 mL/year; 95% CI, −105.63 to −10.43; P=0.017).Table 3Annualized Changes in Lung Function Parameters According to CVD Status in Participants with PRISmUnadjusted ModelAdjusted Model*EstimateP valueEstimateP valueFEV1, mL/year With CVD0.05 ± 18.370.998−15.50 ± 16.740.356 Without CVD33.50 ± 17.730.06437.91 ± 16.810.025 Between-group difference−33.45 (−84.03,17.13)0.193−53.41 (−100.44,−6.38)0.026FEV1% pred, %/year With CVD1.88 ± 0.870.0340.88 ± 0.890.323 Without CVD3.43 ± 0.83<0.0013.39 ± 0.87<0.001 Between-group difference−1.55 (−3.93,0.83)0.199−2.51 (−4.97,−0.05)0.046FVC, mL/year With CVD−5.11 ± 17.730.774−12.75 ± 17.030.455 Without CVD46.56 ± 17.100.00745.28 ± 16.910.008 Between-group difference−51.68 (−100.48,−2.88)0.038−58.03 (−105.63,−10.43)0.017FVC% pred, %/year With CVD1.39 ± 0.850.1030.90 ± 0.870.301 Without CVD3.61 ± 0.81<0.0013.45 ± 0.86<0.001 Between-group difference−2.22 (−4.54,1.00)0.060−2.55 (−4.96,−0.13)0.039FEV1/FVC, %/year With CVD0.07 ± 0.300.806−0.24 ± 0.270.384 Without CVD0.03 ± 0.280.9040.15 ± 0.270.588 Between-group difference0.04 (−0.77,0.85)0.925−0.39 (−1.15,0.38)0.320Notes: Values are expressed as mL/year for FEV1 and FVC and as %/year for FEV1% predicted and FVC% predicted. Annualized changes in each group are presented as estimate ± standard error (SE). Between-group differences are expressed as mean differences with 95% confidence intervals (CIs), calculated as the annualized change in participants with CVD minus that in participants without CVD. * The adjusted models included age, sex, body mass index, smoking history, and the corresponding baseline spirometric value.Abbreviations: PRISm, preserved ratio impaired spirometry; CVD, cardiovascular disease; FEV1, forced expiratory volume in one second; FVC, forced vital capacity; SE, standard error; CI, confidence interval.
For FEV1, the unadjusted between-group difference was not statistically significant (difference, −33.45 mL/year; 95% CI, −84.03 to 17.13; P = 0.193); however, after multivariable adjustment, participants with CVD showed an additional annual decline of 53.41 mL/year compared with those without CVD (95% CI, −100.44 to −6.38; P=0.026). Adjusted annualized changes in FEV1% predicted and FVC% predicted showed similar trends by CVD status (P=0.046 and P=0.039, respectively), whereas changes in FEV1/FVC remained comparable between groups (−0.24 vs 0.15%/year, respectively) after adjustment (adjusted difference, −0.39%/year; 95% CI, −1.15 to −0.38; P=0.320). Overall, the association between CVD and lung function change was more apparent in spirometric measures reflecting lung volumes than in the FEV1/FVC ratio.
Association of CVD with Spirometric Transitions and Incident COPD
Figure 2 illustrates spirometric transitions during follow-up among participants with PRISm at baseline. Of the 521 participants, 117 (22.5%) progressed to COPD, 270 (51.8%) had persistent PRISm, and 134 (25.7%) recovered to normal spirometry. Transition patterns differed by CVD status.
Participants with CVD were more likely to progress to COPD than those without CVD (31.9% vs 18.0%), corresponding to an absolute risk difference of 13.9%, and less likely to recover to normal spirometry (8.4% vs 33.8%). Persistent PRISm was also more frequent in the CVD group (59.6% vs 48.2%). In summary, these findings indicate a more adverse spirometric course among participants with CVD.
Figure 2Alluvial plot of spirometric transitions during follow-up according to baseline CVD status in participants with PRISm. All participants had PRISm at baseline and were grouped by baseline CVD status (CVD group, n=166; no CVD group, n=355). Flows represent transitions from baseline PRISm to follow-up spirometric categories.
Among participants with CVD, 53 (31.9%) progressed to COPD, 99 (59.6%) had persistent PRISm, and 14 (8.4%) recovered to normal spirometry. Among participants without CVD, 64 (18.0%) progressed to COPD, 171 (48.2%) had persistent PRISm, and 120 (33.8%) recovered to normal spirometry.An alluvial plot with y axis label, Number of Patients, unit patients and x axis categories, Baseline, CVD status and Follow up. Baseline shows PRISm, n equals 521.
CVD status splits into CVD, n equals 166 and No CVD, n equals 355. Follow up categories are COPD, n equals 117, PRISm, n equals 270 and Normal, n equals 134. Flow bands connect baseline PRISm through each CVD status group to the three follow up categories.An alluvial plot of PRISm baseline to follow up transitions by cardiovascular disease status.Abbreviations: PRISm, preserved ratio impaired spirometry; CVD, cardiovascular disease; COPD, chronic obstructive pulmonary disease.
Cox proportional hazards models were used to examine the association between CVD and incident COPD in PRISm (Table 4). In the unadjusted model, participants with CVD had an increased risk of developing COPD compared with those without CVD (HR, 1.90; 95% CI, 1.32–2.74; P < 0.01). This association was only modestly attenuated after multivariable adjustment and remained significant both the partially adjusted Model 2 and the fully adjusted Model 3, corresponding to 65% and 63% increased hazards of incident COPD, respectively (Model 2: HR, 1.65; 95% CI, 1.12–2.43; P = 0.012; Model 3: HR, 1.63; 95% CI, 1.07–2.48; P = 0.024).
The proportional hazards assumption was not violated according to Schoenfeld residual testing (all P values >0.05).Table 4Association Between Baseline CVD and Incident COPD in Participants with PRISmCrude ModelModel 1Model 2Model 3HR (95% CI)P valueHR (95% CI)P valueHR (95% CI)P valueHR (95% CI)P valueWithout CVDReferenceReferenceReferenceReferenceWith CVD1.90(1.32,2.74)<0.0011.65(1.12,2.43)0.0121.65(1.12,2.43)0.0121.63(1.07–2.48)0.024Notes: The crude model was unadjusted; Model 1 was adjusted for age, sex, body mass index, and smoking history; Model 2 was additionally adjusted for baseline lung function (post-bronchodilator FEV1 and FEV1/FVC). Model 3 was further adjusted for asthma, emphysema, interstitial lung abnormalities, hypertension, diabetes, and hyperlipidemia. Participants without CVD were used as the reference group.Abbreviations: CVD, cardiovascular disease; PRISm, preserved ratio impaired spirometry; HR, hazard ratio; CI, confidence interval; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity.Table 5IPTW-Weighted Acute Exacerbation Burden and Association Between Baseline CVD and Frequent Exacerbation in Participants with PRISmA: IPTW-weighted acute exacerbation burden according to CVDWith CVDWithout CVDP valueAnnualized rates of AE, events/year Moderate0.81 ± 2.890.33 ± 0.520.104 Severe0.34 ± 0.680.14 ± 0.410.002 Moderate-to-severe1.15 ± 3.350.47 ± 0.660.038Frequent exacerbator, n (%)18 (10.4%)15 (4.2%)0.022B.
Association between baseline CVD and frequent exacerbation after IPTWCrude modelModel 1Model 2Model 3OR (95% CI)P valueOR (95% CI)P valueOR (95% CI)P valueOR (95% CI)P valueWithout CVDReferenceReferenceReferenceReferenceWith CVD2.66(1.12,6.31)0.0262.75(1.16,6.49)0.0212.75(1.16,6.49)0.0212.76(1.16,6.56)0.022Notes: In Panel A, annualized exacerbation rates are presented as weighted mean ± SD, and frequent exacerbators as weighted n (%). In Panel B, the crude model was unadjusted; Model 1 was adjusted for age, sex, body mass index, and smoking history; Model 2 was additionally adjusted for baseline lung function (post-bronchodilator FEV1); and Model 3 was further adjusted for emphysema and diabetes. IPTW was performed using propensity score-derived stabilized weights.
The effective sample size after weighting was 421 participants.Abbreviations: IPTW, inverse probability of treatment weighting; CVD, cardiovascular disease; PRISm, preserved ratio impaired spirometry; AE, acute exacerbation; OR, odds ratio; CI, confidence interval; SD, standard deviation; FEV1, forced expiratory volume in 1 second.
Subgroup and Interaction Analyses
Based on baseline spirometry, participants were classified into non-restrictive PRISm (n = 212, 40.7%) and restrictive PRISm (n = 309, 59.3%). We then explored whether the prognostic implications of CVD differed across PRISm subtypes.
Baseline characteristics by subtype are shown in Table S4. Compared with restrictive PRISm, non-restrictive PRISm had a lower FEV1/FVC ratio and was more frequently accompanied by physician-diagnosed asthma and higher blood eosinophil counts. In contrast, restrictive PRISm was characterized by older age, higher proportions of men and smokers, and lower absolute and percent predicted values of FEV1 and FVC.
In addition, CT-defined interstitial lung abnormalities and cardiovascular comorbidities were more common in restrictive PRISm, which also showed a higher neutrophil-to-lymphocyte ratio and lower HDL cholesterol levels. These findings highlight the clinical heterogeneity of PRISm, with distinct spirometric, inflammatory and comorbidity profiles between restrictive and non-restrictive subtypes.
Acute exacerbation burden was also evaluated within restrictive and non-restrictive PRISm according to CVD status (Table S5). In restrictive PRISm, participants with CVD had higher annualized rates of severe and moderate-to-severe exacerbations than those without CVD. A similar pattern was observed in non-restrictive PRISm.
In logistic regression analyses, CVD was associated with a higher risk of frequent exacerbation in non-restrictive PRISm after full adjustment (OR, 5.41; 95% CI, 1.19–27.16; P = 0.030). In restrictive PRISm, the association was weakened after full adjustment (OR, 2.41; 95% CI, 0.93–6.43; P = 0.073). No significant interaction was observed between CVD status and PRISm subtype for frequent exacerbations (P for interaction =0.642), indicating no statistically significant evidence of interaction in this analysis.
Stratified Cox models were performed to assess COPD progression within each PRISm subtype (Table S6). After full adjustment, CVD remained independently associated with incident COPD in restrictive PRISm, but not in non-restrictive PRISm (HR, 2.60; 95% CI, 1.36–5.00; P = 0.004 vs HR, 1.56; 95% CI, 0.86–2.82; P = 0.143). The interaction by PRISm subtype was not statistically significant (P for interaction = 0.862).
Sensitivity Analyses
Sensitivity analyses showed that the main findings remained robust across different analytical approaches. Comparison of baseline characteristics between included and excluded PRISm participants showed no substantial differences in demographic characteristics, comorbidities, or baseline pulmonary function parameters (Table S1).
After IPTW, all 521 participants were retained in the weighted analysis. Baseline covariates were well balanced between participants with and without CVD, with all post-IPTW standardized mean differences below 0.1 (Table S7). The effective sample size of the weighted cohort was 421 participants.
In the weighted cohort, CVD remained related to frequent exacerbation (OR, 2.76; 95% CI, 1.16–6.56; P = 0.022) (Table 5) and incident COPD (HR, 1.68; 95% CI, 1.04–2.70; P = 0.033) (Table 6) in the fully adjusted model. These findings were consistent with the primary analyses.Table 6Association Between Baseline CVD and Incident COPD in Participants with PRISm After IPTWCrude ModelModel 1Model 2Model 3HR (95% CI)P valueHR (95% CI)P valueHR (95% CI)P valueHR (95% CI)P valueWithout CVDReferenceReferenceReferenceReferenceWith CVD1.40(0.90,2.18)0.01351.39(0.89,2.20)0.1561.63(1.02,2.60)0.0401.68(1.04–2.70)0.033Notes: The crude model was unadjusted; Model 1 was adjusted for age, sex, body mass index, and smoking history; Model 2 was additionally adjusted for baseline lung function (post-bronchodilator FEV1 and FEV1/FVC); and Model 3 was further adjusted for asthma, emphysema, interstitial lung abnormalities, hypertension, diabetes, and hyperlipidemia. IPTW was performed using propensity score-derived stabilized weights.
The effective sample size after weighting was 421 participants.Abbreviations: IPTW, inverse probability of treatment weighting; CVD, cardiovascular disease; PRISm, preserved ratio impaired spirometry; HR, hazard ratio; CI, confidence interval; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity.
Analyses using alternative exacerbation outcomes showed similar results. When severe exacerbations requiring hospitalization were analyzed to reduce the potential impact of differences in healthcare utilization, CVD remained associated with higher odds of severe exacerbations (OR, 3.84; 95% CI, 2.42–6.09; P <0.001; Table S8). In addition, negative binomial regression models incorporating follow-up duration as an offset term showed that participants with CVD had a higher rate of moderate-to-severe exacerbation events during follow-up (IRR, 1.68; 95% CI, 1.31–2.14; P <0.001; Table S3).
Additional linear mixed-effects models were used to evaluate the influence of exacerbation burden and baseline spirometric adjustment on longitudinal lung function analyses. After incorporating exacerbations as a time-updated covariate, the estimated differences in annualized lung function changes between participants with and without CVD were comparable to those from the primary models (Table S9). Models without adjustment for baseline spirometric values showed persistent between-group differences in annualized changes of FEV1 and FVC (Table S10).
Discussion
Using retrospective cohort data, we examine the prognostic relevance of CVD in participants with PRISm. Several key findings can be summarized as follows. First, PRISm participants with CVD represented a clinically high-risk subgroup with a heavier cardiometabolic profile and more marked structural and functional pulmonary abnormalities.
Second, CVD was linked to a more severe acute exacerbation burden in PRISm during follow-up and independently predicted frequent exacerbations. Third, longitudinal spirometric assessment showed accelerated decline in FEV1, FVC, and their percent predicted values in the presence of CVD, whereas changes in FEV1/FVC were comparable between groups. Fourth, participants with CVD were less likely to recover to normal spirometry and had a higher proportion of progression to COPD than those without CVD (31.9% vs 18.0%), corresponding to an absolute risk difference of 13.9%, which was consistent with the findings from the adjusted Cox regression model.
Given the absence of established treatment strategies specifically targeting PRISm,17 clinical management currently relies largely on risk stratification, longitudinal monitoring of lung function trajectories, and identification and management of associated comorbidities. Our findings suggest that CVD may provide clinically relevant information for identifying PRISm individuals at increased risk of exacerbation, impaired spirometric recovery, and COPD progression. Integrating cardiovascular assessment into longitudinal PRISm follow-up may therefore help refine risk stratification and optimize clinical management.
In this study, approximately one-third of participants at baseline had CVD, showing that cardiovascular comorbidity is common in the hospital-based PRISm population. This prevalence was comparable to that reported by Wallström et al in a similar real-world hospital cohort,18 but higher than estimates from PRISm studies conducted in general-population settings.1,19,20 This difference may relate to variations in participant source, underlying comorbidity burden, and the range of conditions included in the CVD definition.
At baseline, participants with CVD showed distinct clinical features, with more prominent metabolic risk factors, a higher prevalence of emphysema, and poorer lung function than those without CVD. This clustering of cardiovascular, metabolic, and pulmonary abnormalities may reflect a more complex systemic disease background. It may partly explain the subsequent excess in exacerbations and poorer spirometric course observed during follow-up.
Peripheral blood findings also raised the possibility of inflammatory heterogeneity, reflected by a higher neutrophil-to-lymphocyte ratio in the CVD group and higher eosinophil counts in the non-CVD group. However, definitive inflammatory phenotyping was not feasible because several inflammatory markers showed only nonsignificant trends, and detailed assessments, including CRP, IL-6, fibrinogen, FeNO, and sputum cell counts, were unavailable. Larger studies incorporating more specific inflammatory biomarkers are needed to confirm these observations.
During follow-up, participants with CVD experienced more clinically significant exacerbations, particularly severe and moderate-to-severe events. Previous studies have demonstrated that PRISm is associated with a considerable risk of exacerbations. In our cohort, participants without CVD showed exacerbation rates comparable with those reported in previous PRISm studies, with moderate, severe, and moderate-to-severe exacerbation rates of 0.34, 0.14, and 0.47 events/person-year, respectively.
These estimates were similar to those reported by Yoon et al21 (0.41, 0.16, and 0.56 events/person-year, respectively), and comparable exacerbation patterns have also been observed by Cheng et al22 among individuals with PRISm. In contrast, participants with CVD showed a higher exacerbation burden, including increased moderate-to-severe exacerbation rates (1.03 events/person-year) and a higher proportion of frequent exacerbators (10.8%). These findings suggest that CVD may identify a subgroup of PRISm individuals with increased susceptibility to clinically significant respiratory events.
The association between CVD and exacerbation burden may be particularly relevant for events requiring intensified treatment, emergency care, or hospitalization, rather than transient symptom fluctuations. Participants with CVD also had lower baseline FEV1 values compared with those without CVD, indicating poorer underlying pulmonary function in this group. However, direct assessments of cardiorespiratory reserve were unavailable in our study, and whether reduced physiological reserve contributes to the observed association between CVD and exacerbations remains unclear.
Although CVD is a well-recognized determinant of poor prognosis in established COPD, our findings extend this observation to PRISm, suggesting that its adverse respiratory impact may already be apparent before overt airflow obstruction develops.
Beyond the increased exacerbation burden, longitudinal assessment revealed heterogeneous spirometric trajectories among individuals with PRISm. In our cohort, 25.7% of participants recovered to normal spirometry during follow-up, whereas 22.5% transitioned to COPD. Previous studies have demonstrated substantial variability in PRISm trajectories, which may be attributable to differences in population characteristics, clinical settings, spirometric definitions, and follow-up duration.
The proportion of participants recovering to normal spirometry in our cohort was comparable with previous reports. For example, a Latin American population-based cohort reported a similar recovery proportion of 26.7%,23 whereas differences in recovery rates have been observed across studies with varying recruitment settings and clinical characteristics. Previous longitudinal studies have also demonstrated substantial variability in COPD transition rates among individuals with PRISm, ranging from approximately 6% to 53% across different cohorts.24 In our cohort, 22.5% of participants progressed to COPD over a mean follow-up of 2.6 years, which was comparable with previous reports despite the shorter observation period.
These differences may partly reflect variations in study populations, baseline clinical characteristics, and follow-up duration, given the hospital-based nature of our cohort and the potentially greater burden of comorbid conditions in this setting. Identifying factors associated with unfavorable disease trajectories remains important for further understanding PRISm progression.
In this context, our findings suggest that cardiovascular comorbidity may represent an important clinical factor associated with adverse spirometric trajectories in PRISm. Lung function analyses suggest that the observed association between CVD and functional deterioration may reflect differences in ventilatory capacity and functional reserve, rather than disproportionate worsening of airflow obstruction. Notably, in our study, several spirometric measures improved over time among participants without CVD, whereas those with CVD showed attenuated improvement or even decline.
Similar findings have been reported by Fan et al25 and in the Rotterdam Study,3 which likewise did not demonstrate a uniform relationship between PRISm and accelerated decline in FEV1 or FVC. This may reflect the heterogeneous clinical course of PRISm, in which some individuals recover to normal spirometry while only a minority progress to COPD. In keeping with this interpretation, in our study, participants without CVD in our study were more likely to recover to normal spirometry, consistent with the observed improvement in spirometric indices.
In contrast, participants with CVD showed a lower proportion of recovery to normal spirometry and a greater tendency toward unfavorable spirometric transitions. However, the markedly lower proportion of recovery to normal spirometry among participants with CVD should be interpreted cautiously, as it may reflect differences in baseline disease severity, treatment status, comorbidity burden, or other unmeasured factors that contributed to the observed differences in spirometric transitions, rather than a direct effect of CVD on spirometric recovery. In addition, differences in healthcare utilization and follow-up intensity between individuals with and without CVD may have influenced the detection of spirometric transitions.
Further longitudinal studies incorporating larger populations and more comprehensive longitudinal assessments are warranted to clarify the factors associated with PRISm recovery.
In this study, our results indicate an association between CVD and worse respiratory outcomes in PRISm; however, the observational design limits our ability to investigate the underlying mechanisms. The mechanisms linking CVD with adverse respiratory outcomes in PRISm remain uncertain, but several explanations are plausible based on previous literature. First, CVD in PRISm may reflect a state of clustered cardiometabolic risk factors and multisystem impairment rather than an isolated coexisting disease.
Factors such as aging, smoking, diabetes, and dyslipidemia may contribute to persistent PRISm and poorer spirometric change through their adverse effects on lung structure and function.1,5 Second, PRISm and CVD may involve overlapping pathophysiological pathways, including chronic inflammation,26 oxidative stress,27 and neurohormonal activation,28 mechanisms that have been widely discussed in COPD with cardiovascular comorbidity.7 These processes may compromise both cardiac and pulmonary function, promote airway inflammation and parenchymal injury,29,30 thereby increasing susceptibility to exacerbations while limiting spirometric recovery. However, whether these pathways specifically mediate the relationship between CVD and respiratory outcomes in PRISm remains unknown. Third, CVD may influence lung function through mechanical and hemodynamic pathways.
Conditions such as heart failure, cardiac enlargement, and pulmonary congestion have been reported to reduce lung volume and compliance and contribute to restrictive ventilatory impairment.31 These mechanisms may partly explain the greater decline in FVC and FVC% predicted observed among participants with CVD in our study. In addition, reduced exercise tolerance, loss of muscle strength, and the cumulative health impact of repeated medical visits or hospitalizations may further worsen dyspnea and functional limitation, lowering the likelihood of recovery to normal spirometry.32–34 Nevertheless, these explanations remain hypothetical in the context of PRISm because detailed cardiac function parameters and functional assessments were not available in our study. Treatment-related factors may also be relevant.
For example, some cardiovascular medications, particularly nonselective β-blockers, may affect airway reactivity or ventilatory function in susceptible individuals.35 However, since medication use was not systematically assessed in our study, these potential mechanisms remain speculative and require further investigation.
Given the marked heterogeneity of PRISm, we performed analyses stratified by PRISm subtype. At baseline, non-restrictive PRISm showed features more suggestive of airway-predominant involvement, including a lower FEV1/FVC ratio, whereas restrictive PRISm was characterized by a more restrictive ventilatory pattern, heavier cardiovascular burden, and more evident systemic inflammatory abnormalities. These observations further support the concept that PRISm does not represent a single spirometric category, but rather encompasses distinct clinical phenotypes with differing patterns of pulmonary dysfunction, comorbidity burden, and clinical risk.5 The associations between CVD and respiratory outcomes were observed in both PRISm subtypes, although effect estimates varied between restrictive and non-restrictive PRISm.
CVD showed a stronger association with incident COPD in restrictive PRISm, whereas the association with frequent exacerbations appeared more evident in non-restrictive PRISm. However, interaction analyses did not provide evidence of statistically significant effect modification by PRISm subtype. These findings should be interpreted cautiously, as the relatively small sample sizes within each PRISm subtype may have limited the statistical power to detect interaction effects.
Differences in baseline clinical characteristics between subtypes may also have contributed to the observed heterogeneity in effect estimates. Therefore, the absence of a statistically significant interaction does not preclude clinically meaningful differences between PRISm subtypes. Further studies with larger prospective cohorts are needed to confirm these observations.
This study adds longitudinal evidence that CVD has important prognostic implications and may provide useful information for phenotypic stratification and follow-up assessment in PRISm. Given the observational nature of this study, these findings should be interpreted as preliminary evidence for risk stratification rather than as support for routine CVD screening or specific management strategies in all individuals with PRISm. PRISm individuals with established CVD or accompanying cardiometabolic risk factors may represent a subgroup requiring closer monitoring of lung function and exacerbation risk.
In these high-risk individuals, attention to cardiovascular risk factor control and coordinated cardiopulmonary follow-up may warrant consideration. Further prospective studies are needed to determine whether targeted cardiopulmonary management strategies can improve long-term outcomes in PRISm with CVD.
Nevertheless, several limitations should be acknowledged in our study. First, this was a single-center retrospective observational cohort study, which limits causal inference between CVD and respiratory outcomes in PRISm. Participants were recruited from a tertiary hospital and required at least two spirometric assessments during follow-up, which may have introduced selection bias by preferentially including individuals with greater healthcare engagement, more complete clinical follow-up, or a potentially more severe underlying disease.
Therefore, the findings may not be directly generalizable to community-based PRISm populations or individuals with less frequent healthcare contact. In addition, although missing baseline data were limited, the complete-case analysis may have introduced additional selection bias if participants with missing information differed systematically from those included.
Second, CVD was analyzed as a composite exposure, including coronary heart disease, congestive heart failure, and stroke. Although this approach captured the overall burden of major cardiovascular comorbidity, different CVD subtypes may have distinct relationships with respiratory outcomes. Furthermore, detailed information regarding CVD severity, disease control, cardiac function, and treatment status was not systematically available.
The impact of incident CVD during follow-up could also not be fully assessed. Therefore, our findings should be interpreted as reflecting the overall association between cardiovascular comorbidity burden and respiratory outcomes rather than differential effects of specific CVD subtypes. Future studies with detailed cardiovascular phenotyping are warranted to clarify whether different CVD subtypes or clinical characteristics are associated with distinct respiratory outcomes in PRISm.
Third, although we adjusted for several important confounders and performed IPTW analyses, residual confounding cannot be completely excluded. Information on medication use, occupational and environmental exposures, physical activity, and socioeconomic factors was incomplete or unavailable. In addition, smoking exposure was assessed using smoking status rather than quantitative measures such as smoking duration, daily cigarette consumption, or cumulative pack-years.
Although smoking status was adjusted for in multivariable models, residual confounding related to smoking intensity and cumulative exposure cannot be completely excluded. Moreover, multiple statistical comparisons were performed, and secondary and subgroup analyses should be interpreted cautiously.
Fourth, acute exacerbations were mainly identified through hospital visit records. Consequently, exacerbations managed in other healthcare settings or milder events not requiring medical attention may have been underestimated. Furthermore, individuals with CVD may have more frequent healthcare encounters, which could increase the likelihood of detecting respiratory events and introduce potential surveillance bias.
Finally, the mechanisms underlying CVD and adverse respiratory outcomes in PRISm could not be directly evaluated in this study. Further multicenter prospective studies incorporating detailed cardiovascular phenotyping, medication data, biomarkers, lifestyle factors, and mechanistic assessments are needed to validate our findings and clarify the potential pathways linking cardiovascular comorbidity with respiratory outcomes in PRISm.
Conclusions
In this real-world PRISm cohort, CVD was associated with adverse respiratory outcomes, including frequent exacerbations (OR, 2.48), incident COPD (HR, 1.63), as well as impaired spirometric recovery after adjustment for potential confounders. CVD may represent a marker of increased cardiopulmonary disease burden in PRISm, although the underlying mechanisms remain unclear. Individuals with PRISm and CVD or cardiometabolic risk factors may represent a higher-risk subgroup requiring closer monitoring of respiratory outcomes and lung function decline.
However, given the retrospective observational design, these findings identify associations rather than establish causal relationships, and do not establish whether routine CVD screening or specific management strategies should be applied universally in PRISm. Further prospective studies are needed to determine the clinical utility of CVD assessment and whether integrated cardiopulmonary management strategies can improve long-term outcomes in PRISm.