Association between hospitalization for pneumonia and subsequent risk of cardiovascular disease

Abstract

Importance: The risk of cardiovascular disease (CVD) after infection is poorly understood.

Objective: To determine whether hospitalization for pneumonia is associated with an increased short-term and long-term risk of CVD.

Design, settings, and participants: We examined 2 community-based cohorts: the Cardiovascular Health Study (CHS, n = 5888; enrollment age, ≥65 years; enrollment period, 1989-1994) and the Atherosclerosis Risk in Communities study (ARIC, n = 15,792; enrollment age, 45-64 years; enrollment period, 1987-1989). Participants were followed up through December 31, 2010. We matched each participant hospitalized with pneumonia to 2 controls. Pneumonia cases and controls were followed for occurrence of CVD over 10 years after matching. We estimated hazard ratios (HRs) for CVD at different time intervals, adjusting for demographics, CVD risk factors, subclinical CVD, comorbidities, and functional status.

Exposures: Hospitalization for pneumonia.

Main outcomes and measures: Incident CVD (myocardial infarction, stroke, and fatal coronary heart disease).

Results: Of 591 pneumonia cases in CHS, 206 had CVD events over 10 years after pneumonia hospitalization. CVD risk after pneumonia was highest in the first year. CVD occurred in 54 cases and 6 controls in the first 30 days (HR, 4.07; 95% CI, 2.86-5.27); 11 cases and 9 controls between 31 and 90 days (HR, 2.94; 95% CI, 2.18-3.70); and 22 cases and 55 controls between 91 days and 1 year (HR, 2.10; 95% CI, 1.59-2.60). Additional CVD risk remained elevated into the tenth year, when 4 cases and 12 controls developed CVD (HR, 1.86; 95% CI, 1.18-2.55). In ARIC, of 680 pneumonia cases, 112 had CVD over 10 years after hospitalization. CVD occurred in 4 cases and 3 controls in the first 30 days (HR, 2.38; 95% CI, 1.12-3.63); 4 cases and 0 controls between 31 and 90 days (HR, 2.40; 95% CI, 1.23-3.47); 11 cases and 8 controls between 91 days and 1 year (HR, 2.19; 95% CI, 1.20-3.19); and 8 cases and 7 controls during the second year (HR, 1.88; 95% CI, 1.10-2.66). After the second year, the HRs were no longer statistically significant.

Conclusions and relevance: Hospitalization for pneumonia was associated with increased short-term and long-term risk of CVD, suggesting that pneumonia may be a risk factor for CVD.

Figure 1. Selection of the Nested Analysis Cohorts From the Cardiovascular Health Study and the Atherosclerosis Risk in Communities Study

ARIC indicates Atherosclerosis Risk in Communities study; CHS, Cardiovascular Health Study; CVD, cardiovascular disease. We used an incidence density approach to match pneumonia cases to controls. Controls were not hospitalized with pneumonia when matched or prior to matching but could develop pneumonia at a later time during follow-up; thus, 54 controls in CHS and 22 controls in ARIC developed pneumonia after matching.

Figure 2. Risk of Cardiovascular Disease Events After Hospitalization for Pneumonia in the Cardiovascular Health Study

CVD indicates cardiovascular disease. The analysis included 1773 participants (591 pneumonia cases and 1182 controls). Stratified analyses by severity of pneumonia included 633 participants (211 pneumonia cases and 422 controls) for pneumonia with organ dysfunction and 1140 participants (380 pneumonia cases and 422 controls) for pneumonia without organ dysfunction. The number of participants at risk and those who developed an event over each time interval were estimated using a complete case approach and participants with missing data for covariates were excluded. The estimates were adjusted for age, sex, race, hypertension, diabetes mellitus, serum total, high-density lipoprotein and low-density lipoprotein cholesterol, smoking, alcohol abuse, atrial fibrillation, chronic kidney disease, serum C-reactive protein, presence of subclinical cardiovascular disease, percentage of predicted forced expiratory volume in first second of expiration (FEV₁) measured by spirometry, trajectories of activities of daily living and independent activities of daily living over time, and trajectories of modified mini-mental status examination scores over time. Adjusted hazard ratios were calculated using the most recently available measurements before inclusion in the nested analysis cohort.

Figure 3. Adjusted Failure Plots to Show the Magnitude of Risk Increase for Cardiovascular Disease That Was Associated With Hospitalization for Pneumonia in CHS and ARIC

ARIC indicates Atherosclerosis Risk in Communities study; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CHS, Cardiovascular Health Study; CVD, cardiovascular disease; LDL, low-density lipoprotein. The failure plots are adjusted for demographics and cardiovascular risk factors (see Methods for additional details). The plots in brown represent participants with pneumonia and the plots in blue represent controls with a covariate pattern in the 25th, 50th, and 75th percentiles in CHS and ARIC. The shaded areas around each curve represent 95% CIs.

Figure 4. Risk of Cardiovascular Disease (CVD) Events After Hospitalization for All Pneumonia in the Atherosclerosis Risk in Communities Study

The analysis included 2040 participants (680 pneumonia cases and 1360 controls). The number of participants at risk and those who developed an event over each time interval were estimated using a complete case approach and participants with missing data for covariates were excluded. The estimates were adjusted for age, sex, race, hypertension, diabetes mellitus, plasma total, high-density lipoprotein and low-density lipoprotein cholesterol, smoking, alcohol abuse, atrial fibrillation, chronic kidney disease, presence of diagnostic Q waves in electrocardiogram, peripheral arterial disease (defined by ankle brachial index <0.9), carotid artery wall thickness, presence of carotid atherosclerotic plaque by ultrasound, and percentage of predicted forced expiratory volume in first second of expiration (FEV₁) measured by spirometry. Adjusted hazard ratios were calculated using baseline (at study entry) covariates measurements.