Pneumococcal pneumonia and the risk of acute coronary
syndrome: A populationbased
cohort study
Chun-Cheng Wang
a,b, Chiao-Ling Peng
c, Guei-JaneWang
b,d,e, Fung-Chang
Sung
c, Chia-Hung Kao
f,⁎
Previous research has proposed an association between acute bacterial infection and an increased incidence of acute coronary
syndrome (ACS) [1]. Several studies have proposed an association
between CAP and an increased incidence of ACS. However, relatively few studies have focused on the specifc bacterial agents that cause pneumonia. Streptococcus pneumonia is the most common pathogen
found in CAP [2]. We investigated whether pneumococcal
pneumonia is associated with an increased incidence of ACS in Asian populations.
We established a longitudinal cohort study based on data from the Taiwan National Health Institute Research Database (NHIRD). The accuracy and validity of diagnosis identifed in the NHIRD has
been strictly implemented and certifcated [3]. This study was
exempted by the Institutional Review Board of China Medical University (CMU-REC-101-012).We selected all adult pneumococcal pneumonia (ICD-9 code 481) patients (≥20 years old), identifying 20111patients with a frst-time diagnosis of pneumococcal pneumonia and with at least one health care admission from 1997 to
2010. The index date was the date of pneumococcal pneumonia registration. We excluded patients with a history of ACS before the index date, and those with incomplete age or sex information. We selected patients without a history of pneumonia (ICD-9 codes 480– 487) or ACS in the NHIRD as our control group. We created four comparison controls for each case, and frequency- matched the patients based on age, sex, and index year of enrollment. We linked study patients to the admission claims data to identify the frst episode of ACS (ICD-9-CM codes 410, 411.1) after pneumococcal pneumonia. We considered hypertension (ICD-9 codes 401–405), diabetes mellitus (DM, ICD-9 code 250), hyperlipidemia (ICD-9-CM 272), and chronic obstructive pulmonary disease (COPD) as potential confounders. We used the chi-square test to compare the distributions of the categorical characteristics between the two
groups.We used Poisson regressionmodels to evaluate the incidence rate ratios (IRR) of ACS between the two groups (relative risk) and 95% confdence intervals (CI). We also used Cox proportionalhazards analysis to investigate the hazard ratios (HR) of ACS over
time between the two groups, and adjusted for any potential confounders. A further analysis was performed to assess whether the association of ACS varied according to the length of the follow-up period after pneumococcal pneumonia was diagnosed. We performed all statistical analyses using the SAS package (Version 9.1 for Windows; SAS institute, Inc., Cary, NC, USA). We adopted a twotailed P value lower than .05 as the statistical signifcance level.
Table 1 shows the demographic characteristics of the study sample. More male patients were present in our study, and almost 75% of the patients were more than 55 years old (mean age 65.0 ± 17.8 years and 64.9 ± 17.6 years in controls, respectively). The prevalence of comorbidities was greater in the pneumococcal pneumonia group.
Table 2 shows the incidence rate ratio (IRR, or relative risk) and adjusted hazard ratio (aHR, or absolute risk) between the two groups. The IRR of ACS is 92% higher in the pneumococcal pneumonia group than in the controls (95% CI=1.70–2.17) with an aHR of 1.47 (95% CI=1.24–1.73) in the following 14 years.We found a 3.90-fold greater relative risk of developing ACS (the highest value) within the frst 3-months follow-up period (95% CI=2.46–6.18).
The result of our study has two conclusions. First, patients infected
with pneumococcal pneumonia have increased risk of ACS in the longtermfollowup. Second, patients infectedwith pneumococcal pneumonia
had the highest relative risk of ACS than patients without pneumococcal pneumonia in the frst 3 months. The relative risk of ACS gradually decreased after 3 months but the risk was persistently signifcantly higher in pneumococcal pneumonia patients. Our explanation is that after the antibiotics treatment of the acute infection, the streptococcus pneumonia may still reside within the nasopharynx and lead to
asymptomatic persistence of infection [4]. Johnstone et al. conducted a
retrospective cohort study to investigate long-term outcomes of patients hospitalized for CAP and concluded that up to 5.4 years of follow-up,
approximately 31% of patients died of cardiovascular disease [5]. The
result implied that CAP patients may have an increased risk of cardiovascular events even after an acute infectious stage.
Several studies have discussed the association between CAP and ACS [4,6]. However, few of these studies have focused on the
association between pneumococcal pneumonia and ACS [7], and
most of them are limited by a relatively small sample size. This prevents any defnite conclusion regarding the association between pneumococcal pneumonia and ACS. Therefore, we used a subset of the database from NHIRD in Taiwan, which encompasses a larger sample size, to investigate this association. The results of this study are justifable because of several reasons. Streptococcus is the most common pathogen in CAP. Second, several studies had pointed out that patients with severe form of CAP might have increased risk of
ACS [8]. Streptococcus pneumonia infection produces higher serum
procalcitonin levels and had a higher infammatory expression than
other atypical pathogen [9]. Therefore, streptococcus pneumonia is
likely associated with the severe form of CAP and with ACS. Our study result implied that pneumococcal vaccination may reduce the incidence of ACS.
This study has limitations. First, the NHI database did not disclose patients' personal histories and serum laboratory data such as smoking, and infammatory markers. Second, in approximately 10% CAP cases, the causative pathogens are mixed. Some patients in the pneumococcal pneumonia group may have a coexistent infuenza virus infection. In conclusion, our study demonstrates that pneumococcal pneumonia is associated with an increased risk of ACS in
the long-term follow up, and the relative risk is the highest within the frst 3 months after exposure to pneumococcal pneumonia.
