Triggering of cardiovascular disease by infection type: The Atherosclerosis Risk in Communities study (ARIC)

Abstract

Introduction:

Acute infections are known triggers of cardiovascular disease (CVD) but how this association varies across infection types is unknown. We hypothesized while acute infections increase CVD risk, the strength of this association varies across infection types.

Method:

Acute coronary heart disease (CHD) and ischemic stroke cases were identified in the Atherosclerosis Risk in Communities Study (ARIC). ICD-9 codes from Medicare claims were used to identify cellulitis, pneumonia, urinary tract infections (UTI), and bloodstream infections. A case-crossover design and conditional logistic regression were used to compare infection types among acute CHD and stroke cases 14, 30, 42, and 90 days before the event with two corresponding control periods (1 and 2 years prior).

Results:

Of the 1312 acute CHD cases, 116 had a UTI, 102 had pneumonia, 43 had cellulitis, and 28 had a bloodstream infection 90 days before the CHD event. Pneumonia (OR = 25.53 (9.21,70.78)), UTI (OR = 3.32 (1.93, 5.71)), bloodstream infections (OR = 5.93 (2.07, 17.00)), and cellulitis (OR = 2.58 (1.09, 6.13)) were associated with higher acute CHD risk within 14 days of infection. Of the 727 ischemic stroke cases, 12 had cellulitis, 27 had pneumonia, 56 had a UTI, and 5 had a bloodstream infection within 90 days of the stroke. Pneumonia (OR = 5.59 (1.77, 17.67)) and UTI (OR = 3.16 (1.68, 5.94)) were associated with higher stroke risk within 14 days of infection.

Conclusions:

Patients with pneumonia, UTI, or bloodstream infection appear to be at a 2.5 to 25.5 fold elevated CVD risk following infection. Preventive therapies during this high-risk period should be considered.

1. Introduction

Population-based cohort studies have identified many chronic risk factors for cardiovascular disease (CVD) that are both modifiable, including high blood pressure, elevated serum cholesterol, and smoking, and non-modifiable, such as male sex, non-white race, family history, and greater age [1,2]. Acute precipitants – or triggers – of CVD are less studied. Prior research provided evidence that acute infections trigger CVD events, including acute coronary heart disease (CHD) [3–8] and ischemic stroke [3,9,10]. The absolute risk of CVD events following community-acquired pneumonia has been explored. A meta-analysis on the incidence of CVD events within 30 days of receiving a pneumonia diagnosis reported incidence of new or worsening heart failure, new or worsening arrhythmias, myocardial infarction (MI) or unstable angina, and overall cardiac complications (any of the aforementioned cardiac events) in 14.1%, 5.3%, 4.7%, and 17.7% of patients admitted to the hospital, respectively [11]. A more recent study on hospitalization for pneumonia and acute CVD (i.e., MI, stroke, and fatal CHD) reported that 10.6% of exposed participants experienced an acute CVD event within 30 days of infection [7].

Studies also concluded increased CVD risk is highest immediately following infection and decreases with time since infection for up to 3 months [12–15]. Infection-induced inflammation may be the mechanism underlying the increased time-dependent risk of acute CHD and ischemic stroke after acute infection [13,16–18]. Specifically, infection-induced inflammation may lead to the triggering of cardiovascular events by increasing inflammation markers in the bloodstream, promoting the development of platelet thrombi, increasing vasoconstriction, disrupting coronary plaques, and decreasing endothelial function [12,13,18]. Prior studies on the effects of acute infections on cardiovascular risk have examined infections collectively or assessed the impact of a single infection on cardiovascular risk. A review by Corrales-Medina et al. [12] and a meta-analysis by Mei et al. [19] identified the most studied infection triggers in CVD studies were acute respiratory infections, such as pneumonia. Meanwhile other common infections, including urinary tract infections (UTI), were not thoroughly studied. Furthermore, associations between increased CVD risk and common infections such as bloodstream infections and cellulitis lack evidence.

Differential associations between infection types and increased CVD risk may inform the utility of CVD preventive therapies immediately following high-risk infections. The aim of this study was to examine the associations between the four most common types of infections (cellulitis, pneumonia, UTI, and bloodstream infections) [20] and the risk of acute CHD events and ischemic stroke. We hypothesized all acute infections increase acute CHD and ischemic stroke risk, but the level of increased risk may vary by infection type.

2. Methods

2.1. Study population

The ARIC study is a multi-center, population-based, prospective cohort study designed to investigate the etiology and natural history of atherosclerosis in middle-aged Americans [21]. At baseline in 1987–1989 (visit 1), 15,792 mostly white and black men and women were selected from four United States communities: Forsyth County, North Carolina; Jackson, Mississippi; suburbs of Minneapolis, Minnesota; and Washington County, Maryland [21]. Subsequent exams took place during 1990 to 1992 (visit 2), 1993 to 1995 (visit 3), 1996 to 1998 (visit 4), 2011 to 2013 (visit 5), and 2016–2017 (visit 6), and a 7th visit is currently underway. This study conforms to the ethical guidelines of the 1975 Declaration of Helsinki and informed consent was obtained from each participant. Details of the ARIC study design and objectives are provided elsewhere [21].

2.2. Study design

A case-crossover study design was used, in which ARIC participants with CHD and ischemic stroke served as both cases and their own controls. The case-crossover design afforded the ability to isolate exposures that varied over time within a person, as well as to provide better control for potential confounding factors, both measured and unmeasured. The occurrence of infection immediately prior to acute CHD events and ischemic stroke events was compared with preceding time intervals, 1 year and 2 years prior to the CVD event, to account for the seasonality of infection. The crossover study design was summarized in Fig. 1.

Fig. 1.

Case-crossover study design employed to study acute CHD and ischemic stroke in relation to triggering infection, ARIC.

All ARIC participants with acute CHD or ischemic stroke events during follow-up were included. Events were identified using information obtained at study visits, during telephone questionnaires, and through review and abstraction of hospital and death records [22]. Hospitalizations were identified by surveillance of local hospital discharge lists for cohort members. Centers for Medicare and Medicaid Services (CMS) claims data, which were linked to ARIC cohort data, were used to identify infections in ARIC study participants. CMS claims data for inpatient and outpatient services were available since 1991. Individuals who were younger than 67 years of age at the time of their CVD event were excluded since they were not Medicare eligible for both the case and control periods. We also excluded participants whose CVD events occurred prior to 1993 to ensure that CMS data were available for both case and control periods. Data were available through 2013. Only ICD-9 codes were considered since the conversion to ICD-10 codes occurred after our study timeline.

2.3. Infection ascertainment

ICD-9 codes were used to identify four infection types as our exposures of interest based on infection frequency: cellulitis, pneumonia, UTI, and bloodstream infections. Both inpatient hospitalization and outpatient visit codes were used to identify incident infections. Table 1 contains the infection types and corresponding ICD-9 codes included as exposures of interest. ICD-9 codes in any chart position were counted.

Table 1.

Infection type and corresponding ICD-9 codes included in the exposure of interest.

Infection type	ICD-9 codes
Pneumonia	480*–486*
Urinary tract infection	590*, 595.0, 595.1, 595.2, 595.3, 595.4, 597*, 598.0*, 599.0, 601*, 604*, 607.1, 607.2, 608.0, 608.4
Bloodstream infections	038*, 790.7
Cellulitis	681*, 682*

2.4. Cardiovascular events

The outcomes of interest were acute CHD (MI and fatal CHD) and ischemic stroke. The ARIC methods used for outcome ascertainment included: (1) participants were contacted annually by phone and interviewed about interim hospitalizations; (2) local hospitals provided lists of hospital discharges with cardiovascular diagnoses that were reviewed to identify cohort hospitalizations; and (3) health department death certificate files were regularly surveyed. All discharge codes for cohort hospitalizations and listed causes of death from death certificates were recorded. CVD events were classified by a combination of computer algorithm and adjudicated physician review; disagreements were adjudicated by the ARIC Mortality and Morbidity Classification Committee using standardized ARIC criteria [22].

Ischemic stroke was identified and classified as thrombotic or cardioembolic stroke based on discharge codes, signs, symptoms, neuroimaging (computerized tomography/magnetic resonance imaging), and other diagnostic reports [23]. CHD and stroke event dates were based on death dates for fatal events and hospital admission dates for hospitalized events [22]. CHD and ischemic stroke events between study enrollment and end of year 2013 were included in the analyses.

2.5. Statistical analysis

Acute CHD and ischemic stroke were analyzed separately since the impact of infection may differ between acute CHD and ischemic stroke events. The prevalence of each infection type 14, 30, 42, and 90 days before each event was compared with the corresponding time periods exactly one year and two years before the event to account for the seasonality of infection. A washout period of two days between the CVD admission date and preceding infection date was used to exclude infections that may have been diagnosed secondarily at a CVD hospitalization. Conditional logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for each time period (14, 30, 42, and 90 days). Separate models were run for each infection type of interest. Differences between the 90-day ORs for each infection type for each outcome were compared using multinomial logistic regression.

Potential confounders including CVD risk factors stable within an individual were controlled through the case-crossover study design by having cases serve as their own controls. Confounding by overall health status related to age was possible because deteriorating health could be a common cause of both infection and CVD. As participants age and their health status decreases, their CVD risk and risk of infection may increase. To reduce potential confounding by age and deteriorating health status, only time periods proximal to the CVD event (1 year and 2 years before) were included. Further, the total number of hospitalizations for any cause in the nine months preceding the start of each of the three exposure periods (case period and 2 control periods) was also controlled for to account for potential decline in overall health status.

3. Results

Among the 15,792 ARIC study participants, 2356 (14.9%) experienced an acute CHD event. Those participants who were younger than 67 years of age at the time of their event (n = 1017) and those events that occurred prior to 1993 (n = 27) were excluded to ensure CMS data were available for both case and control periods. Our final sample size included 1312 acute CHD cases. A combined 1150 (7.3%) ARIC participants experienced an ischemic stroke event. We excluded those participants who were younger than 67 years of age at the time of their event (n = 410) and those events that occurred prior to 1993 (n = 13) to ensure CMS data were available for both case and control periods. Our final sample size included 727 ischemic stroke cases.

At-event characteristics of ARIC participants who developed acute CHD and ischemic stroke are provided in Table 2. The mean age at CVD event was 75 years with a standard deviation of five years. Acute CHD was more common among men (57.4%) while ischemic stroke (54.1%) was more common among women. The majority of events occurred in white participants consistent with baseline enrollment.

Table 2.

At-event characteristics of ARIC participants who developed acute CHD events and ischemic stroke, 1987–2013.

Characteristic*	Acute CHD (n = 1312)	Ischemic stroke (n = 727)
Age, years, mean ± SD	75.0 ± 5.3	75.1 ± 5.1
Sex, count (%)
Male	753 (57.4)	334 (45.9)
Female	559 (42.6)	393 (54.1)
Race, count (%)
White	949 (72.3)	492 (67.7)
Black	359 (27.4)	234 (32.2)
Other	4 (0.3)	1 (0.1)

^*At baseline, in 1987–1989, participants were on average 54.2 ± 5.8 years old, 55.2% were female, 27.0% black and 72.7% white.

3.1. Acute CHD

Of the 1312 acute CHD cases, 116 (8.8%) had a UTI, 102 (7.8%) had pneumonia, 43 (3.3%) had cellulitis, and 28 (2.1%) had a bloodstream infection within 90 days of the acute CHD event. Table 3 contains the conditional logistic regression model results for acute CHD. Cellulitis (14-day: 2.58 (1.09, 6.13), 90-Day: 1.41 (0.93, 2.15)), pneumonia (14-day: 25.53 (9.21, 70.78), 90-day: 5.60 (3.72, 8.43), UTI (14-day: 3.32 (1.93, 5.71), 90-day: 2.62 (1.92, 3.57), and bloodstream infections (14-day: 5.93 (2.07, 17.00), 90-day: 4.77 (2.34, 9.71) were more common in all case periods compared with equivalent control periods. Associations were generally strongest in exposure periods closest to the acute CHD event and decreased as the time window before the acute CHD event increased. Across infection types, controlling for the number of hospitalizations in the nine-month period preceding each exposure period, mostly attenuated the associations between infection and acute CHD. In the 90-day period, pneumonia, UTI, and bloodstream infections were stronger acute CHD triggers compared to cellulitis (p < 0.05). Pneumonia and bloodstream infections were stronger CHD triggers compared to UTI, but only pneumonia reached statistical significance (p < 0.05).

Table 3.

Association between infection and acute CHD events by infection type among 1312 ARIC cohort participants who experienced an incident CHD event, OR (95%CI).

Crude	Cellulitis	Pneumonia	UTI	Blood infections
14-Day	2.89 (1.24, 6.76)	25.76 (9.32, 71.24)	3.40 (1.99, 5.82)	5.60 (2.02, 15.55)
30-Day	2.19 (1.18, 4.06)	8.69 (4.96, 15.22)	3.21 (2.12, 4.86)	7.60 (2.84, 20.35)
42-Day	1.69 (0.97, 2.94)	7.48 (4.57, 12.24)	3.36 (2.30, 4.91)	5.08 (2.24, 11.50)
90-Day	1.51 (1.00, 2.28)	5.83 (3.88, 8.75)	2.69 (1.97, 3.66)	4.84 (2.40, 9.76)
Model 1*	Cellulitis	Pneumonia	UTI	Blood infections
14-Day	2.58 (1.09, 6.13)	25.53 (9.21, 70.78)	3.32 (1.93, 5.71)	5.93 (2.07, 17.00)
30-Day	2.02 (1.08, 3.78)	8.58 (4.88, 15.08)	3.12 (2.06, 4.72)	8.29 (3.01, 22.82)
42-Day	1.62 (0.93, 2.85)	7.27 (4.43, 11.94)	3.28 (2.24, 4.80)	5.22 (2.25, 12.12)
90-Day	1.41 (0.93, 2.15)	5.60 (3.72, 8.43)	2.62 (1.92, 3.57)	4.77 (2.34, 9.71)

^*Model 1 adjusted for total hospitalizations in the nine months preceding each exposure period.

3.2. Ischemic stroke

Of the 727 ischemic stroke cases, 56 (7.7%) had a UTI, 27 (3.7%) had pneumonia, 12 (1.7%) had cellulitis, and 5 (0.7%) had a bloodstream infection within 90 days of the stroke event. Table 4 contains the conditional logistic regression model results for ischemic stroke. Pneumonia (14-day: 5.59 (1.77, 17.67), 90-day: 3.07 (1.64, 5.75)) and UTI (14-day: 3.16 (1.68, 5.94), 90-day: 1.78 (1.18, 2.69)) were more common in all case periods compared with equivalent control periods. Bloodstream infections were more common in all case periods compared to control periods and cellulitis was more common in the 14-day period, but these associations failed to reach statistical significance. The association between infection and stroke was generally stronger in the exposure periods closest to the stroke event and decreased as the time window before ischemic stroke increased. Like acute CHD, controlling for the number of hospitalizations in the nine-month period preceding each exposure period mostly attenuated the association between infection and ischemic stroke. In the 90-day period, pneumonia and UTI were stronger stroke triggers compared to cellulitis (p < 0.05). Pneumonia was a stronger stroke trigger compared to UTI and bloodstream infections but failed to reach statistical significance. The results of each infection type for both acute CHD and ischemic stroke are presented side by side in Fig. 2.

Table 4.

Association between infection and ischemic stroke by infection type among 727 ARIC cohort participants who experienced an incident ischemic stroke event, OR (95% CI).

Crude	Cellulitis	Pneumonia	UTI	Blood infections
14-Day	1.71 (0.42, 7.05)	5.50 (1.75, 17.27)	3.17 (1.70, 5.92)	2.00 (0.13, 31.98)
30-Day	0.80 (0.26, 2.51)	4.00 (1.71, 9.35)	2.31 (1.35, 3.96)	3.00 (0.50, 17.95)
42-Day	0.79 (0.31, 2.07)	4.75 (2.08, 10.85)	2.02 (1.26, 3.26)	2.67 (0.60, 11.92)
90-Day	0.66 (0.32, 1.35)	3.13 (1.68, 5.85)	1.80 (1.20, 2.71)	1.48 (0.44, 4.95)
Model 1*	Cellulitis	Pneumonia	UTI	Blood infections
14-Day	1.64 (0.40, 6.74)	5.59 (1.77, 17.67)	3.16 (1.68, 5.94)	1.73 (0.11, 28.04)
30-Day	0.83 (0.27, 2.58)	3.82 (1.62, 8.96)	2.21 (1.28, 3.80)	3.23 (0.52, 19.91)
42-Day	0.84 (0.32, 2.19)	4.50 (1.96, 10.33)	1.99 (1.23, 3.22)	3.28 (0.69, 15.57)
90-Day	0.67 (0.33, 1.39)	3.07 (1.64, 5.75)	1.78 (1.18, 2.69)	1.64 (0.48, 5.63)

^*Model 1 adjusted for total hospitalizations in the nine months preceding each exposure period.

Fig. 2.

Associations between infection type within 90 days and acute CHD and ischemic stroke, ARIC.

Population attributable fractions (PAFs) were computed to estimate the proportion of acute CHD and ischemic stroke events that could be avoided in the absence of each acute infection at the 90-day period. Adjusted PAFs were calculated using the formula PAF = pd((RR-1)/RR), where pd is the proportion of cases exposed to the risk factor [24] and the risk-ratio (RR) was estimated with the OR due to the rare disease assumption being met. The 90-day PAFs of acute CHD for cellulitis, pneumonia, UTI, and bloodstream infections were 0.95%, 6.39%, 5.47%, and 1.69%, respectively. The 90-day PAFs of ischemic stroke for cellulitis, pneumonia, UTI, and bloodstream infections were −0.81%, 2.50%, 3.38%, and 0.27%, respectively. Among the four infections, participants with pneumonia and UTI presented the greatest 90-day PAFs for acute CHD and ischemic stroke.

4. Discussion

This case-crossover study nested within a population-based cohort study suggests acute CHD and ischemic stroke risk are higher in the time period shortly after an acute infection and that the magnitude of increased risk varies by infection type. Pneumonia presented the strongest association with both acute CHD and ischemic stroke, followed by bloodstream infections and UTI for acute CHD and UTI and bloodstream infections for ischemic stroke. Cellulitis demonstrated the weakest association with both acute CHD and ischemic stroke and did not reach statistical significance for the 42-day and 90-day analysis for acute CHD – and for the 30-day, 42-day, and 90-day analysis for ischemic stroke. While the CIs for pneumonia and bloodstream infections did overlap for acute CHD for all four periods, they did not overlap with UTI and cellulitis on the 14-day, 30-day, and 90-day analysis, suggesting pneumonia and bloodstream infections are significantly stronger triggers than UTI and cellulitis. For all infection types, the magnitude of association was generally stronger for acute CHD. The association between all infection types and both acute CHD and ischemic stroke was graded such that the triggering was strongest in the exposure periods most proximal to the event and generally decreased as the time window before the event increased.

These findings support our hypothesis that acute infections can trigger acute CHD and ischemic stroke events and the risk of acute CHD and ischemic stroke varies by infection type. Previous studies found similar results when assessing acute infections and increased risk of CVD, but they differ from this study since multiple acute infections were assessed collectively [10,25–27] or the relationship between a specific acute infection type (i.e. respiratory infection) and CVD was the primary focus [4,7,13–15]. Our results were similar to findings by Corrales-Medina et al. [7] who found the hazard ratio for incident CVD (MI, stroke, and fatal CHD) between 31 and 90 days after hospitalization for pneumonia was 2.94 (95% CI, 2.18–3.70). Our findings reinforce associations found by the aforementioned studies while expanding on how these associations vary by infection type.

Previous work indicated the transient triggering impact of infection on CVD was likely due to a link between inflammation and coagulation activation [12]. Acute infection may trigger CVD through short-term changes in endothelial function, plaque composition and white-cell activation – or from mechanisms that are not biophysiological responses to infection including dehydration or bed rest [13]. Corrales-Medina et al. [12] suggested these infection-triggered events could be from a multitude of biophysiological alterations including increased systemic inflammatory activity, disturbed hemodynamic homeostasis, dominant prothrombic conditions, increased biochemical stress on coronary arteries, variations in the coronary arterial tone, and altered myocardial metabolic balance.

Although associations between infections and acute cardiovascular events were previously reported, only pneumonia’s mechanistic effects on the cardiovascular system has been well characterized [28]. Summarizing the impact of pneumonia on the cardiovascular system, Corrales-Medina et al. [28] concluded pneumonia was found to affect vascular endothelium and peripheral vessels, cardiac rhythm, coronary arteries, pulmonary circulation, cardiac autonomic function, coagulation, and renal function and fluid and sodium balance. The authors proposed a pathophysiological model for cardiac complications (heart failure, myocardial ischemia or infarction, and cardiac arrhythmias) in patients with pneumonia which may explain the strong association between pneumonia and CVD observed in our study [28]. Future research should be conducted to understand how other infection types impact the cardiovasular system and lead to cardiovascular disease, which may explain the disparate impact on CVD between infection types found in our study.

Treatment for acute infections such as pneumonia, UTI, blood stream infections, and cellulitis consist of prescribed antibiotic agents [29]. Recent studies have reported on the safety of antibiotic agents as it relates to CVD risk [30–35]. Findings from these studies suggests most antibiotics (e.g., non-macrolide antibiotics) do not increase CVD risk and are safe to use in high-risk populations [30,31]—while macrolide antibiotics, particularly azithromycin and clarithromycin, may marginally increase CVD risk in the short-term and should be weighed against their appreciable, immediate benefits [32–35]. Lastly, prior studies have shown that treatment of less severe infections in outpatient settings are associated with less CVD risk than their more severely infected inpatient counterparts [12,36].

Some important limitations to our study should be noted. Small cell sizes resulting from relatively few infections proximal to CVD events resulted in imprecise estimates and wide confidence intervals. Although comparisons at 14, 30, and 42 days before CVD event may not be as precise as the 90-day comparison, they were still informative and showed a graded trend demonstrated by other studies [4,7,9,10,12–15,25]. Like all case-crossover studies, our study may suffer from survival bias as we did not consider infections in participants who did not have a CVD event. While it is possible that infections and CVD could be diagnosed simultaneously, we used a 2-day washout period in an effort to reduce the possibility that the infection and CVD were diagnosed simultaneously or that the infection did not proceed the CVD event. Confounding by age was possible because as participants aged, their risk of both CVD and infection increased. However, the potential of confounding by age was mitigated by using control periods 1 and 2 years before the exposure period and by controlling for the total number of hospitalizations in the nine-month period preceding each exposure period. Other unmeasured confounders that may vary between the exposure and control periods were not included. We likely under-ascertained infections, especially minor ones that did not require care, since exposure data were collected using Medicare claims data. This data collection method would most likely lead to non-differential misclassification of the exposure that would typically bias ORs toward the null [37]. Other characteristics of infection including etiologic agent and severity of disease were not considered.

This study had a number of important strengths including a large population-based cohort, a stringent methodology to adjudicate CVD events, and a cross-over design that controlled for potential confounding factors stable within a person [38]. Prior hospitalizations for any cause in the nine months preceding exposure periods and seasonality of infections were also controlled for to account for confounding by overall health status.

Our findings have several implications pertaining to future research and care. Infection status and type should be considered in assessing CVD risk. Identification of specific infections as a CVD trigger may prompt more aggressive treatment with standard preventive strategies, including antiplatelet agents and statins, during and immediately after infection, to reduce the increased CVD risk. This time period immediately after infection has been referred to as a treatable moment [10]. Evidence-based vaccinations may also be considered because of its ability to not only reduce infection but also CVD [39].

Benefits of statin therapy, antiplatelet agents, and influenza vaccinations for acute CVD events have been demonstrated in RCTs, including decreasing cardiovascular mortality, stroke, MI, and composite cardiovascular outcomes [40–43]. A few RCTs have reported mixed results on ventilator-associated pneumonia survival and statins (N = 2) in intensive/critical care patients, with one finding that statins improve 30-day survival [44] and the other reporting null results [45]. Observational studies on statin treatment and pneumonia have reported an increase in patient survival after contracting pneumonia [46,47]. The effect of pneumococcal vaccinations on CVD has not been evaluated in large-scale RCTs and should be considered in future research endeavors [43].

The mixed results from RCTs and a lack of RCTs that evaluate the effect of statins and antiplatelet agents on acute CVD events immediately following acute infections—most notably pneumonia—in high-risk patients, warrants the need for large-scale RCTs that assess their effect on the transient elevated CVD risk observed immediately following acute infections. Finally, further research is needed to better understand the impact of specific infections on the cardiovascular system so that preventative therapies can be tailored to target the underlying pathophysiologic mechanisms leading to increased transient CVD risk.

5. Conclusion

Our study found participants who experienced an acute CHD event or ischemic stroke had a higher odds of having an acute infection up to 90 days before their CVD event compared to control periods. Generally, the magnitude of the associations was strongest at the 14-day analysis and decreased with time. Patients with pneumonia, bloodstream infections, and UTI have an elevated transient risk for developing these acute CVD events, especially immediately following infection. Of the infections explored, pneumonia may be the strongest trigger of both acute CHD and ischemic stroke events. Patients who have pneumonia, UTI, and bloodstream infections may be of particular interest for testing preventative strategies, including antiplatelet agents and statins, immediately following infection.