Association Between Vaccination and Acute Myocardial Infarction and Ischemic Stroke After COVID-19 Infection

Abstract

This retrospective cohort study examines the incidence of acute myocardial infarction and ischemic stroke after COVID-19 infection among vaccinated vs unvaccinated adults in Korea.

Studies have suggested an increased incidence of acute myocardial infarction (AMI) and ischemic stroke after COVID-19 infection related to an increased risk of thrombosis.^1,2 Vaccines against SARS-CoV-2 are effective against COVID-19 and its progression to severe disease.³ However, it is unclear if vaccines also prevent secondary complications. We examined the association between vaccination and AMI and ischemic stroke after COVID-19 infection.

Methods

We conducted a retrospective cohort study to compare the incidence of AMI and ischemic stroke after COVID-19 infection between patients who were never vaccinated and those who were fully vaccinated (2 doses of mRNA vaccines or viral vector vaccine) against SARS-CoV-2. The Korean nationwide COVID-19 registry (on infection and vaccination) and the Korean National Health Insurance Service database were used. COVID-19 reporting is mandated, and Korea has universal health care coverage. Adults aged 18 years or older who were diagnosed with COVID-19, including asymptomatic infections, between July 2020 and December 2021 were included. Exclusion criteria included (1) outcome events less than 3 months before COVID-19 diagnosis; (2) reinfection; (3) hospitalization for COVID-19 for 30 or more days and, among vaccinated patients, (4) single dose of vaccine; and (5) COVID-19 diagnosis before or within 7 days after the second vaccination. Patients were observed until March 31, 2022.

The primary outcome was a composite of hospitalizations for AMI and ischemic stroke that occurred 31 to 120 days after COVID-19 diagnosis; these were identified by the diagnosis codes and relevant imaging (eMethods in the Supplement). The first 30 days were excluded because of the difficulty of differentiating cardiovascular events occurring as complications of COVID-19 vs acute phase treatment. Secondary outcomes included the components of the composite outcome. Inverse probability of treatment weighting (IPTW) was used to control for differences in patient characteristics between the 2 groups,⁴ with standardized differences used to assess the balance of covariates. Logistic regression was performed for IPTW with full vaccination as an independent variable and age, sex, Charlson Comorbidity Index, hypertension, and insurance type as covariates. A Cox proportional hazards model with IPTW was constructed for the outcome events, with sex, age, comorbidities, previous history of outcome events, and the severity of COVID-19 (need for supplementary oxygen [severe], high-flow nasal cannula or higher respiratory support [critical] vs no respiratory support needed) as covariates. The proportionality assumption was tested (zph tests) and met. SAS Enterprise Guide 7.1 (SAS Institute) was used for statistical analysis. A 2-tailed P < .05 was considered significant. This study was approved by the institutional review board of the Gil Medical Center with a waiver of informed consent.

Results

Of 592 719 patients with COVID-19 during the study period, 231 037 patients were included, of whom 62 727 were never vaccinated and 168 310 were fully vaccinated. Patients who were fully vaccinated were older and had more comorbidities (Table 1). In contrast, severe or critical COVID-19 was less common in the fully vaccinated group. The differences in age and comorbidities were reduced after weighting, while the severity of COVID-19 became less balanced. The median follow-up duration starting 30 days after COVID-19 was 90 days in the unvaccinated group and 84 days in the fully vaccinated group.

Table 1. Baseline Characteristics of Study Population by Vaccination Status^a.

	Unweighted population, No. (%)b			Weighted population, %b
	Not vaccinated (n = 62 727)	Fully vaccinated (n = 168 310)	Absolute standardized differencec	Not vaccinated	Fully vaccinated	Absolute standardized differencec
Sexd			0.029			0.042
Male	30 407 (48.48)	79 176 (47.04)		45.11	47.21
Female	32 320 (51.52)	89 134 (52.96)		54.89	52.79
Age, median (IQR), yd	42 (31 to 58)	57 (42 to 66)	0.504	52 (37-68)	54 (38-65)	0.087
18-39	28 467 (45.38)	36 444 (21.65)		30.39	26.80
40-64	24 183 (38.55)	80 647 (47.92)		39.71	46.66
≥65	10 077 (16.06)	51 219 (30.43)		29.90	26.54
Insurance plan for low incomed	3308 (5.27)	6310 (3.75)	0.074	4.47	4.24	0.011
Comorbidities
Charlson Comorbidity Index, median (IQR)e	0 (0 to 2)	1 (0 to 2)
Charlson Comorbidity Index ≥5d	4001 (6.38)	11 792 (7.01)	0.025	7.26	6.87	0.015
Diabetes	4479 (7.14)	19 929 (11.84)	0.161	9.17	11.06	0.063
Hypertensiond	6782 (10.81)	37 166 (22.08)	0.308	20.07	19.03	0.029
Dyslipidemia	2254 (3.59)	13 618 (8.09)	0.193	4.25	7.57	0.141
Previous history of outcome events	909 (1.45)	2704 (1.61)	0.013	2.29	1.46	0.062
Severity of COVID-19f
Severe	6136 (9.78)	5298 (3.15)	0.289	12.45	2.84	0.399
Critical	3514 (5.60)	1772 (1.05)	0.276	8.52	0.95	0.397

^a The Cox proportional hazard model was constructed for the outcome events with sex, age, previous history of outcome events, diabetes, hypertension, hyperlipidemia, CCI, and the severity of COVID-19 as covariates, weighted by the inverse probability of full vaccination.

^b Data are presented as No. (%) or percentage of patients unless stated otherwise.

^c Standardized mean difference <0.1 was considered a good balance between groups

^d Variables included in the calculation of propensity score.

^e Charlson Comorbidity Index (CCI) is calculated by the addition of points designated to each comorbid condition, identified from the International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD-10) codes entered within 3 years. One point is given for myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, connective disease, ulcer disease, mild liver disease, and diabetes. Two points for hemiplegia, moderate or severe kidney disease, diabetes with end organ damage, any tumor, leukemia, and lymphoma. Three points for moderate or severe liver disease. Six points for metastatic solid tumor and AIDS.

^f Severe COVID-19 was identified using electronic data interchange (EDI) procedure codes for supplementary oxygen. Critical COVID-19 was identified by the EDI codes for high-flow nasal cannula, intubation, tracheostomy, mechanical ventilation, and extracorporeal membrane oxygenation.

The composite outcome occurred in 31 unvaccinated patients and 74 fully vaccinated patients, with an incidence of 6.18 vs 5.49 per 1 000 000 person-days (Table 2). The adjusted risk was significantly lower in the fully vaccinated group (adjusted hazard ratio [aHR], 0.42; 95% CI, 0.29-0.62). The adjusted risk was significantly lower in fully vaccinated patients for both AMI (aHR, 0.48; 95% CI, 0.25-0.94) and ischemic stroke (aHR, 0.40; 95% CI, 0.26-0.63). A lower risk for outcome events in fully vaccinated patients was observed in all subgroups, although some did not reach statistical significance, including those with severe or critical infection (Table 2).

Table 2. Risk for Cardiovascular Events by Vaccination Status.

	No. of events		Incidence per 1 000 000 person-days		Adjusted HR (95% CI)	P value
	Not vaccinated (n = 62 727)	Fully vaccinated (n = 168 310)	Not vaccinated	Fully vaccinated
Composite outcome	31	74	6.18	5.49	0.42 (0.29-0.62)	<.001
Acute myocardial infarction	8	24	1.60	1.78	0.48 (0.25-0.94)	.03
Ischemic stroke	23	50	4.59	3.71	0.40 (0.26-0.63)	<.001
Subgroups
Male	17	48	6.98	7.59	0.41 (0.26-0.66)	<.001
Female	14	26	5.44	3.63	0.42 (0.23-0.76)	.004
Age, y
40-64	11	22	5.48	3.39	0.38 (0.20-0.74)	.004
≥65	20	51	33.99	12.42	0.41 (0.26-0.66)	<.001
Charlson Comorbidity Index
<5	25	56	5.22	4.45	0.40 (0.26-0.60)	<.001
≥5	6	18	25.04	19.79	0.54 (0.24-1.22)	.14
Diabetes
No	23	46	4.89	3.87	0.38 (0.24-0.61)	<.001
Yes	8	28	26.29	17.58	0.47 (0.25-0.91)	.03
Hypertension
No	20	46	4.41	4.39	0.50 (0.31-0.80)	.004
Yes	11	28	23.11	10.90	0.34 (0.18-0.62)	<.001
Dyslipidemia
No	27	70	5.58	5.65	0.54 (0.37-0.80)	.002
Yes	4	4	22.50	3.62	0.09 (0.03-0.34)a	<.001
Previous history of outcome events
No	26	67	5.24	5.05	0.44 (0.29-0.65)	<.001
Yes	5	7	97.55	33.26	0.33 (0.10-1.07)	.06
Severe or critical COVID-19
No	22	65	5.02	5.00	0.37 (0.25-0.55)	<.001
Yes	9	9	14.38	18.51	0.66 (0.20-2.23)	.51

Abbreviation: HR, hazard ratio.

^a The following covariates were excluded from this model due to separation: severity of COVID-19 and previous history of outcome events.

Discussion

This study found that full vaccination against COVID-19 was associated with a reduced risk of AMI and ischemic stroke after COVID-19. The findings support vaccination, especially for those with risk factors for cardiovascular diseases. Study limitations include that diagnosis codes for reimbursement were used to capture outcome events. Although the operational definition in this study has been widely used, some diagnostic inaccuracies may exist. Also, there were imbalances in patient characteristics by vaccination status. The decision to be vaccinated is affected by multiple factors that may also be associated with cardiovascular risk. A robust model was applied to mitigate the effect of such imbalances, but the possibility of unobserved bias remains.

Section Editors: Jody W. Zylke, MD, Deputy Editor; Kristin Walter, MD, Senior Editor.

Supplement.

eMethods

eReference