1 INTRODUCTION

The excess mortality seen during the Coronavirus Disease 2019 (COVID-19) pandemic has been undeniably profound. In the United States and worldwide, rates of excess mortality due to COVID-19 have been especially elevated for adults aged 65 years or older, males, and racial/ethnic minority groups.^{1, 2} Importantly, the measurable total excess in mortality includes deaths attributed to several major non-COVID-19-specific causes.^{3, 4} A substantial proportion of these non-COVID-19-specific deaths include cardiovascular deaths due to ischemic heart disease.⁴ Before the pandemic, ischemic heart disease was the leading worldwide cause of death—but with steadily improving downward trends in year-to-year mortality rates, observed through 2019.^{5, 6} Unfortunately, analyses of mortality trends since 2020 indicate that the COVID-19-associated excess in ischemic heart disease mortality has effectively erased these prior gains.⁴

The exact drivers of excess ischemic heart disease mortality seen during the pandemic remain unclear, although many probable and potential causes have been identified. For instance, there is now abundant evidence of how patients with pre-existing or newly developed cardiovascular conditions have experienced gaps and delays in access to care, especially during periods of COVID-19 surge.^7-10 Additionally, mounting data demonstrates that a proportion of COVID-19-affected individuals are at increased risk for thrombogenic events, including acute coronary events, either during or following the acute infection phase.¹¹ It is also possible that the excess in cardiovascular risks observed during earlier phases of the pandemic was associated to generally more severe COVID-19 illness caused by more virulent variants of SARS-CoV-2. To further understand the nature and potential origins of excess mortality attributed to ischemic heart disease during the pandemic and their changes over time, we analyzed national vital statistics data to comprehensively examine and compare temporal trends in mortality rates across demographic groups and geographic regions in the United States.

2 MATERIALS AND METHODS

3 RESULTS

3.2 AMI-associated mortality by age and sex

Age-standardized AMI-associated mortality rates increased from prepandemic to postpandemic epochs across all age groups (Figure 1). Despite a reduction in the 4th epoch, the relative increase remained disproportionately pronounced in the youngest age group, with the percent difference between observed and predicted mortality rates ranging from 23% to 34% for the youngest compared to 13% to 18% for the oldest age groups (Table 1); this trend was evident even when accounting for seasonality of events (Figure 2A and Supporting Information: Figure S1). Moreover, the change in the proportion of COVID-19- and non-COVID-19-associated deaths varied substantially across all age groups. While the proportion of COVID-19 increased considerably from the 1st epoch to the 2nd epoch, followed by a decrease in the 3rd epoch for both the middle and oldest age groups, it gradually increased through the first three epochs in the youngest age group. Subsequently, there was a striking increase in the proportions of COVID-19-associated mortality observed in all age groups in the fourth epoch (Figure 1). The excess mortality rates are displayed as an absolute number of deaths in Table 2. More than three-quarters of deaths were aged 65 or above.

Table 1. Age-standardized mortality rate in US adults with acute myocardial infarction (AMI), categorized by age and sex between 2012 and 2022

	Age-standardized mortality rate (per 100 000 persons)
	Prepandemic 4/2012–9/2012	Pandemic epoch 1 4/2020–9/2020			Pandemic epoch 2 10/2020–3/2021			Pandemic epoch 3 4/2021–9/2021			Pandemic epoch 4 10/2021–3/2022
	Observed	Observed	Predicted [95% CI]	% Differencea	Observed	Predicted [95% CI]	% Differencea	Observed	Predicted [95% CI]	% Differencea	Observed	Predicted [95% CI]	% Differencea
Age
25–44	1.91	2.08	1.69 [1.48–1.91]	+23.37%	2.16	1.71 [1.56–1.86]	+26.34%	2.25	1.68 [1.45–1.91]	+34.22%	2.07	1.68 [1.52–1.85]	+23.02%
45–64	19.47	20.31	17.39 [16.58–18.20]	+16.80%	21.92	18.10 [17.53–18.66]	+21.11%	20.56	17.14 [16.28–18.01]	+19.93%	20.66	17.70 [17.10–18.31]	+16.70%
≥65	129.16	108.48	96.12 [92.70–99.53]	+12.86%	125.24	106.22 [95.67–116.77]	+17.90%	105.54	92.21 [88.54–95.87]	+14.46%	119.11	101.30 [90.00–112.61]	+17.58%
Sex
Female	25.45	21.62	18.84 [18.00–19.69]	+14.74%	24.12	20.65 [18.69–22.62]	+16.82%	21.31	18.05 [17.15–18.96]	+18.05%	23.05	19.68 [17.58–21.79]	+17.12%
Male	44.77	40.88	35.93 [34.66–37.20]	+13.77%	46.83	38.84 [36.12–41.56]	+20.58%	40.32	34.9 [33.54–36.26]	+15.52%	44.24	37.43 [34.51–40.35]	+18.21%

• ^a % Difference between predicted and observed values.

Table 2. Number of excess AMI-associated mortality in US adults with acute myocardial infarction (AMI), categorized by COVID-19

Excess AMI-associated deaths (N)
	Pandemic epoch 1 4/2020–9/2020	Pandemic epoch 2 10/2020–3/2021	Pandemic epoch 3 4/2021–9/2021	Pandemic epoch 4 10/2021–3/2022
The number of excess AMI-associated mortality, overall
Age group
25–44 years	1427	1464	1441	1439
45–64 years	15 384	15 893	15 015	15 555
≥65 years	50 581	56 129	48 864	53 565
Sex
Female	27 114	29 860	25 957	28 426
Male	40 171	43 449	39 152	41 975
The number of excess AMI-associated mortality, SARS-CoV-2 infection group
Age group
25–44 years	66	110	180	187
45–64 years	675	1480	1229	1644
≥65 years	2848	7041	2750	5386
Sex
Female	1444	3195	1568	2736
Male	2145	5436	2591	4481
The number of excess AMI-associated mortality, non-SARS-CoV-2 infection group
Age group
25–44 years	1361	1354	1261	1252
45–64 years	14 709	14 413	13 786	13 911
≥65 years	47 733	49 088	46 114	48 179
Sex
Female	25 670	26 665	24 389	25 690
Male	38 026	38 013	36 561	37 494

Consistently, in temporal trends analyses, AMI-associated mortality rates significantly decreased during the prepandemic epochs across all age groups and for both sexes (Figures 2 and 3, Supporting Information: Table S2); then, from before the start of the pandemic to the most recent pandemic epoch, AMI-associated mortality rates were observed to reverse direction and overall increase—and this rise was especially pronounced and statistically significant for the youngest and middle-age group among whom the SAPC in AMI-associated mortality increased by 5.3% (95% confidence interval [CI]: 1.6%–9.1%) and 3.4% (95% CI: 0.1%–6.8%), respectively (Supporting Information: Table S2).

While the trends of excess death were similar for both sexes (Table 1 and Figure 2B), substantial variation was seen across the sex-specific, age-stratified groups. In age- and sex-stratified analyses (Table 3, Figure 3, and Supporting Information: Figures S2 and S3), the percent differences between observed and predicted mortality rates were highest among the youngest age groups for females during the first (+26.3%) and third (+42.3%) epochs of the pandemic, although highest for the youngest age male group (+30.2%) during the second epoch of the pandemic (i.e., Fall/Winter of 2020). In the fourth epoch, the youngest females demonstrated no significant increase in mortality (+1.28%), whereas their male counterparts experienced a 26.2% excess death rate. In the analyses of temporal trends, AMI-associated mortality rates decreased during the prepandemic epochs across all subgroups with significant declines seen for older females (−2.0% [95% CI: −3.1 to −1.0%]) and middle-aged (−0.7% [95% CI: −1.4 to −0.1%]) to oldest (−1.5% [95% CI −2.7 to −0.4%]) males (Supporting Information: Table S3); then, from before the start of the pandemic to the most recent pandemic epoch, these trends reversed with AMI-associated mortality rates increasing and significant rises seen for the youngest-aged females (5.6% [95% CI: 1.1%–10.4%]) and for the youngest (5.1% [95% CI: 1.0%–9.4%]) and middle-aged (3.2% [95% CI: 0.0–6.4]) males (Supporting Information: Table S3).

Table 3. Age-standardized mortality rate in US adults with acute myocardial infarction (AMI), by age and further stratified by sex between 2012 and 2022

	Age-standardized Mortality Rate (Per 100 000 Persons)
	Prepandemic 4/2012–9/2012	Pandemic epoch 1 4/2020–9/2020			Pandemic epoch 2 10/2020–3/2021			Pandemic epoch 3 4/2021–9/2021			Pandemic epoch 4 10/2021–3/2022
	Observed	Observed	Predicted [95% CI]	% Differencea	Observed	Predicted [95% CI]	% Differencea	Observed	Predicted [95% CI]	% Differencea	Observed	Predicted [95% CI]	% Differencea
Females
25–44	1.11	1.23	0.97 [0.81–1.13]	+26.33%	1.30	1.20 [0.94–1.45]	+8.12%	1.37	0.96 [0.79–1.13]	+42.25%	1.27	1.25 [0.90–1.60]	+1.28%
45–64	11.01	11.71	9.42 [8.19–10.65]	+24.31%	12.53	10.12 [8.99–11.24]	+23.77%	11.96	9.01 [7.33–10.70]	+32.70%	12.10	9.61 [8.07–11.16]	+25.88%
≥65	104.33	83.26	73.57 [70.40–76.73]	+13.16%	94.15	80.95 [71.67–90.22]	+16.31%	80.91	69.89 [66.50–73.28]	+15.77%	89.64	76.47 [66.54–86.41]	+17.23%
Males
25–44	2.71	2.94	2.42 [2.11–2.72]	+21.60%	3.02	2.32 [2.02–2.62]	+30.21%	3.14	2.39 [2.07–2.72]	+31.42%	2.87	2.27 [1.95–2.59]	+26.24%
45–64	28.46	29.40	23.75 [21.40–26.11]	+23.79%	31.84	26.07 [25.31–26.82]	+22.12%	29.62	21.20 [16.51–25.89]	+39.74%	29.69	25.48 [24.67–26.28]	+16.52%
≥65	163.55	141.48	125.60 [121.11–130.09]	+12.64%	166.55	139.18 [126.16–152.20]	+19.66%	137.78	121.17 [116.36–125.98]	+13.71%	158.09	133.34 [119.39–147.29]	+18.56%

• ^a % Difference between predicted and observed values.

3.3 AMI-associated mortality by region

ASMRs increased from prepandemic to postpandemic epochs across the majority of states, particularly during the second and fourth epochs (Figure 4). The Southern United States experienced the highest rates of AMI-associated deaths during the COVID-19 pandemic between 4/2020–3/2022 (Figure 5). The individual states with the highest increase in the AMI-associated ASMR (per 100 000 persons) throughout the COVID-19 pandemic were Montana (12.5), Kentucky (11.0), West Virginia (10.5), and Mississippi (9.6) (Supporting Information: Table S4). In state-specific analyses during each epoch, Mississippi had the highest increase in AMI-associated death rate between 10/2019–3/2020 and 4/2020–9/2020. Tennessee, Kentucky, and South Dakota experienced the highest growth in AMI-associated death rates between 4/2020–9/2020 and 10/2020–3/2021 (Figure 4 and Supporting Information: Figure S4). Hawaii and North Dakota experienced the highest increase in ASMR for AMI-associated deaths between 10/2020–3/2021 and 4/2021–9/2021, while Montana and Rhode Island experienced the highest growth between 4/2021–9/2021 and 10/2021–3/2022.

4 DISCUSSION

From our temporal trends study of AMI-associated deaths across the United States, we observed several important findings. First, our results confirm that a substantial and significant increase in AMI-associated mortality across the population has emerged since the start of the COVID-19 pandemic. Importantly, this overall rise in AMI-associated mortality has effectively erased within 2 years what was previously a decade-long trend of steadily declining trend, with a majority contributed by non-COVID-19 conditions.⁶ Second, the excess in AMI-associated mortality has persisted throughout the pandemic, even during the most recent epoch marked by a surge of the presumed less-virulent Omicron variant, notwithstanding temporal variations in rates of non-COVID-19 versus COVID-19-associated mortality. Third, we found that although the increase in AMI-associated death rate during the pandemic was seen across all age groups, the relative rise was most significant among the youngest age group of 25–44 years old. The relative excess in AMI-associated deaths in youngest adults was seen initially in both sexes and then, during the fourth epoch marked by the Omicron surge, appeared to persist in younger males while somewhat improving in younger females.

Our findings extend from recent studies that have documented a substantial increase in cardiovascular morbidity experienced overall and particularly in traditionally high-risk subgroups such as older adults and males.^{4, 17} Our study expands from prior work by analyzing raw data collected over a longer period of time, comparing observed to predicted mortality rates, further stratifying sex groups by age, and accounting for known seasonal variation in AMI events. Furthermore, we focused on demographic comparisons to clarify how the accentuated and still widening disparities in excess mortality may be distinct from those relevant to morbidity. As shown in recent reports,¹⁸ the pandemic can augment absolute differences in outcomes in certain population subsets while augmenting relative differences in outcomes in others. Accordingly, we found that although older compared to younger adults continued to experience the greatest absolute magnitudes of excess AMI risk during the pandemic; younger adults experienced the greatest relative increase in excess AMI-associated mortality. Notably, the death rate increased more in males than females when considered in either absolute or relative terms, particularly during the fourth and most recent epoch. Importantly, even while AMI-associated mortality rates have recently overall improved along with trends in all-cause mortality during the pandemic, the relative disparity in excess AMI-associated deaths among certain at-risk subgroups such as younger-aged males has persisted.

There are several potential explanations for the increased AMI-associated mortality excess in patients with COVID-19.^{11, 19, 20} Typically, AMI and cardiovascular death typically do not present until years to decades after de novo development or worsening of pre-existing risk factors. However, SARS-CoV-2 infection may activate or accelerate the development of pre-existing subclinical coronary artery disease,²¹ which may be prevalent, particularly among younger-aged men with clustering of cardiometabolic risk factors.²² While de novo rapid development of AMI during or after SARS-CoV-2 infection is likely to be uncommon; such cases have been reported among individuals without known pre-existing cardiovascular risk traits—especially males affected by more severe forms of COVID-19 illness.^{23, 24} Proposed mechanisms for the thrombogenicity associated with SARS-CoV-2 infection include downregulation of angiotensin-converting enzyme 2(ACE-2), leading to dysregulation of the renin-angiotensin-aldosterone system, oxidative stress damage, endothelial cell dysfunction, and complement-mediated coagulopathy and microangiopathy.^{11, 23} Studies have consistently shown that males, independent of age, are more likely to develop more severe forms of SARS-CoV-2 infection than females.²⁵ This male bias could have contributed to the predominance of excess AMI-associated mortality seen among males in our analyses. Psychological stressors associated with the pandemic may have also played a role in triggering events leading to excess AMI-associated mortality. Mental stress-induced ischemia has been shown to be associated with worse cardiovascular outcomes and disproportionately affects young women,^{25, 26} which could contribute to the excess of AMI-associated mortality seen among young females in our analyses.

Indeed, the effects of social distancing and stay-at-home mandates, reductions in outpatient visits and rehabilitation services, and deferment of elective procedures are all likely to have contributed to the overall excess of AMI-associated deaths during the COVID-19 pandemic.^27-30 Importantly, however, we observed improved and yet persistent relative excess in AMI-associated deaths even as health delivery services have adapted to manage care during the most recent surge of the generally less virulent Omicron variant of SARS-CoV-2.

Several limitations of our study merit consideration. Our analyses were based on vital statistics data wherein the cause of death is subject to inaccurate documentation, leading to misclassification of AMI versus non-AMI-associated deaths. For this reason, we prioritized analyses of temporal trends while recognizing potential residual effects of misclassification bias. For subgroup samples with fewer events, particularly over shorter periods, model fit was suboptimal, resulting in wide confidence limits for some estimates. For all analyses, data associated to clinical characteristics, including cardiovascular risk factors and access to cardiovascular care resources, were not available. Therefore, additional studies are needed to investigate correlates and potential determinants of the trends in excess AMI-associated mortality found in our analyses of vital statistics.

5 CONCLUSIONS

Our study revealed that the excess rise in AMI-associated death occurring in the United States during the COVID-19 pandemic had been especially pronounced in younger than older adults, with a sex disparity that has become especially evident during the most recent fourth epoch marked by the Omicron surge. While a recent reduction in non-COVID-19-associated deaths suggests the efficacy of policies designed to curtail the impact of the pandemic, AMI-associated death rates remain in excess. Taken together, our results highlight disparities in mortality that have developed during the COVID-19 pandemic and are persistent through March 2022.

AUTHOR CONTRIBUTIONS

Study design and data analysis: Yee Hui Yeo, Xinyuan He, Fan Lv, Susan Cheng, Fanpu Ji. Drafting of the manuscript: Yee Hui Yeo, Maggie Wang, Xinyuan He, Susan Cheng, Fanpu Ji. Critical review of the manuscript: Jian Zu, Joseph E. Ebinger, Jignesh K. Patel, Susan Cheng. Study conception and study supervision: Susan Cheng, Fanpu Ji. All authors contributed to data interpretation, critical revisions, and approval of the final manuscript.

CONFLICTS OF INTEREST

Dr. S. Cheng has received consulting fees from Zogenix outside of the submitted work. Dr. F. Ji has received speaker fees from Gilead Sciences, MSD, and Ascletis, in addition to consulting or advisory board fees from Gilead and MSD, all outside of the submitted work. The remaining authors declare no conflict of interest.