Ischemic heart disease and cerebral infarction associated mortality in the United States

Abstract

Ischemic Heart Disease (IHD) and cerebral infarction share risk factors and are leading cardiovascular causes of death. We analyzed national mortality trends and disparities for co-occurring IHD and cerebral infarction in older United States adults. Using the CDC WONDER database, we analyzed 1999 to 2023 death records for adults aged ≥55 listing both conditions as underlying or contributing causes. Age-adjusted mortality rates (AAMRs) per 1,00,000 and annual percent changes (APCs) were calculated via Joinpoint regression. Of 1,06,423 deaths, the overall AAMR declined from 1999 to 2011 (APC: −11.42), then rose sharply through 2023 (APC: 9.64). Females comprised 50.20% of absolute deaths, but males had a higher 2023 AAMR (7.2 vs 4.9). Non-Hispanic African Americans had the highest AAMR (6.5). The West (6.0) and nonmetropolitan areas (6.0) faced the heaviest mortality burdens. Following a decade of decline, mortality for comorbid IHD and cerebral infarction increased significantly after 2011. Disparities among males, Non-Hispanic African Americans, and rural populations underscore the need for targeted healthcare interventions and equitable resource allocation.

1. Introduction

Ischemic heart disease (IHD) continues to be one of the leading causes of death in the United States (US), as approximately 3,71,506 deaths were recorded from coronary artery disease in 2022, which makes up nearly 40% of all cardiovascular deaths in this country.^[1] However, the IHD-associated mortality rates have been stabilized across some populations, representing gains in treatment and prevention in past decades, yet significant disparities exist, revealing a disproportionate burden among populations.^[2] Cerebral infarction, on the other hand, has frequently been reported with IHD, both of which have many risk factors in common, like atherosclerosis, hypertension, diabetes, hyperlipidemia, and inflammation.^[3,4] Moreover, IHD increases the risk of cerebral infarction due to low cardiac output, arrhythmias, the presence of thrombus in the left ventricle, or cardioembolism.^[5,6] Patients with both conditions are at increased risk of in-hospital mortality, adverse vascular events, and experience poor long-term survival when compared to those having either of the conditions alone.^[7,8] Further, cerebral infarction patients without any prior cardiac history have shown significant narrowing of the coronary artery.^[9]

Current literature has explored the association of IHD and cerebral infarction, but has been limited to smaller, regional, or hospital-based registries, which limits the wide applicability of the data.^[10] Further, no mortality analysis has been done at the national level to assess long-term mortality trends and disparities across different demographic and geographic subgroups in a country. To fill this void, we looked at the US national death certificate data listing both cerebral infarction and IHD from the years 1999 to 2023 using the CDC WONDER (Centers for Disease Control and Prevention Wide-ranging Online Data for Epidemiologic Research) database. By analyzing trends across sex, age, race, and region, we can inform healthcare policy to develop targeted interventions and allocate equitable resources to the populations with a high mortality burden. We charted trends over time and assessed which demographics were most common in the years IHD and cerebral infarction appeared on death certificates.

2. Materials and methods

2.1. Study design and data source

A retrospective population-based analysis was done using death records from 1999 to 2023 from the CDC WONDER database.^[11] Anonymized death certificate data for all 50 states and the District of Columbia were included in the database from the National Vital Statistics System. The de-identified datasets can be used publicly and do not require patient consent or institutional review board approval. Our main query logic used was the multiple cause of death files. Only those records with both IHD (International Classification of Diseases, Tenth Revision [ICD-10] codes I20-I25) and cerebral infarction (ICD-10 code I63) co-mentioned anywhere on the death certificate were retrieved, and these may either be the underlying cause of death or a contributing cause. Following the guidelines for the use of CDC WONDER data to protect the privacy of the data and to provide statistical reliability, any demographic or geographic cross-tabulations that yielded fewer than 20 deaths were suppressed. We acknowledge that the exclusion of these suppressed cells in our join point analysis may limit the stability and interpretability of some of the subgroup trends, especially across rural populations. The study was reported using the Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines.^[12]

2.2. Study population and case definition

The study population included only adults ≥55 years of age and excluded younger populations. This particular threshold was chosen for 2 main reasons. Pathophysiologically, the age of 55 years is a turning point clinically, as the burden of multi-vessel atherosclerosis increases, resulting in a multiple-fold increase in the incidence, mortality, and disability-adjusted life years of IHD and ischemic stroke when compared to younger age groups.^[13,14] From an epidemiological point of view, registry data from the globe and the region over the years have clearly shown that the behavior of cardio-cerebrovascular disease changes at this very age boundary. Over the past decades, there has been stable or increasing prevalence or incidence of cardiovascular disease (CVD) in people under 55 years of age, but there has been a clear downward or plateauing trend in older people ≥55..^[13,15] By limiting the age range of our cohort to this older age group, we can avoid the bias of using unreliable rate estimates and suppressing data (fewer than 20 deaths) associated with lower incidence and younger age groups, and thus maximize the statistical stability and clinical relevance of the data.

In addition, differences were evaluated by stratifying by demographic and geographic factors. Demographic variables included age groups (55–64, 65–74, 75–84, and 85+ years), race/ethnicity (Non-Hispanic [NH] White, NH African American, Hispanic, NH Asian, and NH American Indian), and gender, and geographic variables included urbanization level (metropolitan vs nonmetropolitan), census regions (Northeast, Midwest, South, and West), and states and places of death. Urbanization was categorized as either metropolitan regions or nonmetropolitan regions based on the National Center for Health Statistics Urban-Rural Classification Scheme for Counties, 2013.^[16] Given the data availability limitations associated with this urbanization classification scheme in the CDC WONDER database, all analyses that compared any level of urbanization were limited to the years 1999 and 2020.

2.3. Statistical analysis

Crude mortality rates (CMR) were only used for age-stratified analyses, as age-specific cohorts are already limited in terms of age, which means they do not require population standardization. Thus, CMR is an accurate reflection of the true, unadjusted mortality load in those specific age groups. For all other demographic and geographic comparisons, however, age-adjusted mortality rates (AAMR) were computed per 1,00,000 population, assuming the 2000 US standard population to adjust for the changing age structure of the population during the 25-year trend.^[17] The Joinpoint regression program (V 5.0.2) was used to analyze trends over time.^[18] This program fits log-linear models to determine statistically significant changes in mortality trends over time and calculates the annual percent change (APC) over each segment. A maximum of four joinpoints (inflection points) were allowed, and models were selected based on the weighted Bayesian information criterion. Using a 2-sided P-value < .05, the difference in rates and estimates of trends were considered statistically significant, and 95% confidence intervals (CIs) were computed for all rates and estimates of trends. Importantly, all APCs with a 95% confidence interval (CI) crossing the null value were explicitly identified and reported as being statistically nonsignificant to avoid the misinterpretation of a stable or fluctuating trend. A sensitivity analysis was conducted to confirm our main results and to mitigate the possibility of overestimation due to multiple causes of death coding. The following 2 restricted cohorts were analyzed from the main data set: those where IHD was listed as the primary/underlying cause of death (UCD) and cerebral infarction as a contributing cause of death, and those where IHD was listed as a contributing cause of death and cerebral infarction as the UCD. This enabled us to assess whether the overall temporal trends and inflection points remain consistent regardless of the condition that triggered the sequence of events leading to death. In addition, mortality rate ratios (RRs) and 95% CIs were formally computed to compare the mean AAMRs over the 25-year study period between specified subgroups of the population. These comparative analyses were done using a prospectively determined reference group, namely females (for gender disparities), NH Whites (for racial and ethnic disparities), and metros (for urbanization disparities). Rate comparisons were considered statistically significant if the 2-sided P-value was <.05. This formal statistical comparison ensures that any differences between subgroups that are observed are stable and not due to chance. Lastly, detailed query parameters, age-adjustment metrics, and Joinpoint regression settings to ensure full reproducibility are provided in the supplementary material (Appendix 1, Supplemental Digital Content 1).

3. Results

3.1. Absolute mortality distribution across subgroups

Both IHD and cerebral infarction were recorded in a total of 1,06,423 deaths, with females (50.20%) being higher in proportion compared to males (49.80%). The majority of the deaths occurred in NH Whites (81.72%), followed by NH African Americans (9.49%), Hispanics (5.50%), and NH Asians (2.55%). Mortalities were documented as the highest in medical facilities (45.29%), followed by nursing homes (25.73%), decedents’ homes (19.15%), Hospice facilities (6.20%), and other (3.52%), respectively. Deaths were more common in metropolitan areas (80.30%) compared to nonmetropolitan areas (19.70%). Deaths were more predominant in persons aged 85+ years (39.32%), followed by persons aged 75 to 84 years (34.14%), 65 to 74 years (18.28%), and 55 to 64 years (8.27%). Deaths were more prevalent in the South region (37.04%), followed by the Midwest (22.54%), the West (21.54%), and the Northeast (18.87%). Mortalities are documented highest in California (11.50%) and lowest in Alaska (0.10%; Figs. S1 and S2, Supplemental Digital Content 2; Table S1, Supplemental Digital Content 3).

3.2. Overall and gender-related IHD and cerebral infarction mortality trends

A significant reduction was noted from 1999 (11.6) to 2023 (7.9) when considering overall mortality rates. A sharp decrease in AAMR was recorded from 1999 to 2011, followed by a significant incline from 2011 to 2023 with associated APCs of −11.4* (95% CI: −13.3 to −10.2) and 9.6* (95% CI: 7.8–12.5), respectively. The mean AAMR was significantly higher in males (7.2) compared to females (4.9), demonstrating a marked gender disparity (RR 1.48; 95% CI: 1.39–1.57, P < .001). After significant reductions in AAMR in females until 2011, a sharp rise was observed through 2023 with associated APCs of −12.0* (95% CI: −14 to −10) and 8.8* (95% CI: 7.0–11.7), respectively. Males’ AAMR declined significantly from 1999 to 2008, followed by a gradual nonsignificant decline from 2008 to 2014 with associated APCs of −12.3* (95% CI: −17.4 to −9.8) and −3.2 (95% CI: −13.7 to 4.9), respectively. After 2014, a steady nonsignificant rise was noted from 2014 to 2017, followed by a significant increase through 2023, with associated APCs of 21.2 (95% CI: −1.8 to 27.7) and 8.6* (95% CI: 1.3–11.6), respectively (Figs. 1, 2; Tables 1, 2 and S2, Supplemental Digital Content 4).

Figure 1.

Overall trends in ischemic heart disease and cerebral infarction-related age-adjusted mortality rates per 1,00,000 among older adults in the United States, 1999 to 2023. AAMR = age-adjusted mortality rate.

Figure 2.

Trends in ischemic heart disease and cerebral infarction-related age-adjusted mortality rates per 1,00,000, stratified by sex among older adults in the United States, 1999 to 2023. AAMR = age-adjusted mortality rate.

Table 1.

Annual percent change (APC) of ischemic heart disease and cerebral infarction-related age-adjusted mortality rates per 1,00,000 among older adults in the United States, 1999 to 2023.

Year interval	APC (95% CI)	P-value
Overall
1999–2011	−11.4* (−13.3 to −10.2)	<.000001
2011–2023	9.6* (7.8–12.5)	<.000001
Gender
Female
1999–2011	−12* (−14 to −10)	<.000001
2011–2023	8.8* (7–11.7)	<.000001
Male
1999–2008	−12.3* (−17.4 to −9.8)	.013597
2008–2014	−3.2 (−13.7 to 4.9)	.216357
2014–2017	21.2 (−1.8 to 27.7)	.067986
2017–2023	8.6* (1.3–11.6)	.041592
Race
NH Asian or Pacific Islander
1999–2010	−14.6* (−18.9 to −11.5)	<.000001
2010–2023	7.6* (5.3–11.4)	<.000001
NH African American
1999–2013	−10.6* (−12.6 to −9.1)	<.000001
2013–2023	12.6* (9.8–16.6)	<.000001
NH White
1999–2011	−11.5* (−13.7 to −10.2)	<.000001
2011–2023	10.1* (8.3–13)	<.000001
Hispanic
1999–2011	−10.3* (−13.1 to −8)	<.000001
2011–2023	9.7* (7.8–12.2)	<.000001
Urbanization
Metropolitan
1999–2011	−11.8* (−14.6 to −10.1)	<.000001
2011–2020	9.6* (6–17)	<.000001
Nonmetropolitan
1999–2003	−8.4* (−11 to −3)	.018796
2003–2006	−18.5* (−21.8 to −12.9)	.007598
2006–2014	−3.6 (−7.1 to 2)	.117576
2014–2018	20.1* (0.9–29.5)	.047590
2018–2020	5.8 (−0.6 to 14.3)	.059988
Age groups
55–64 yr
1999–2011	−6.5* (−9.0 to −4.7)	<.000001
2011–2023	10.3* (8.7–12.9)	<.000001
65–74 yr
1999–2011	−9.6* (−12.2 to −7.6)	<.000001
2011–2023	9.5* (7.7–12.4)	<.000001
75–84 yr
1999–2011	−11.5* (−14.0 to −9.6)	<.000001
2011–2023	9.0* (6.5–12.5)	<.000001
85+ yr
1999–2003	−9.5* (−12.2 to −3.5)	.015197
2003–2006	−22.4* (−26.0 to −16.3)	.002000
2006–2014	−5.1 (−8.0 to 0.4)	.057988
2014–2018	18.7 (−4.6 to 27.6)	.060788
2018–2023	8.8* (2.7–12.2)	.016797
Sensitivity analysis
IHD as UCD + cerebral infarction as MCD/contributing cause
1999–2004	−10.4* (−12.3 to −7.5)	<.000001
2004–2007	−20.6* (−24.9 to −13.2)	.008798
2007–2013	−4.7 (−12.3 to 7.3)	.337932
2013–2023	8.6* (6.3–15.0)	.018796
Cerebral infarction as UCD + IHD as MCD/contributing cause
1999–2006	−11.1* (−16.8 to −8.7)	.001200
2006–2014	−4.1 (−8.8–11.2)	.205159
2014–2017	24.5* (1.7–32.7)	.037992
2017–2023	6.3* (1.5–9.9)	.031594

APC = annual percent change, CI = confidence interval, IHD = ischemic heart disease, MCD = multiple cause of death, NH = non-Hispanic, UCD = underlying cause of death.

Table 2.

Formal statistical comparison of mean age-adjusted mortality rates (AAMR) across demographic and geographic subgroups (1999–2023).

Disparity category	Subgroup	Mean AAMR (95% CI)	Reference group	Mean AAMR (95% CI)	Rate ratio (95% CI)	P-value
Gender	Male	7.19 (6.88–7.50)	Female	4.86 (4.67–5.08)	1.48 (1.39–1.57)	<.001
Race/ethnicity	NH African American	6.49 (5.84–7.12)	NH White	5.92 (5.72–6.13)	1.10 (0.99–1.22)	.0824
	Hispanic	4.57 (3.93–5.20)	NH White	5.92 (5.72–6.13)	0.77 (0.67–0.89)	<.001
	NH Asian	4.39 (3.54–5.29)	NH White	5.92 (5.72–6.13)	0.74 (0.61–0.91)	.0038
Urbanization	Nonmetropolitan	6.03 (5.62–6.45)	Metropolitan	5.49 (5.31–5.68)	1.10 (1.02–1.19)	.0161

AAMR = age-adjusted mortality rate, CI = confidence interval.

3.3. Sensitivity analysis by underlying cause of death

To validate the primary temporal trends and isolate the primary drivers of mortality, a sensitivity analysis was conducted evaluating 2 restricted cohorts: one where IHD was designated as the UCD with cerebral infarction as a contributing cause, and another where cerebral infarction was the UCD with IHD as a contributing cause. Among the total co-occurring deaths, IHD was identified as the UCD in nearly half of the cases (49.15%; n = 52,310), while cerebral infarction was the UCD in one-quarter of the cohort (25.31%; n = 26,939).

When restricting to cases where IHD was the UCD (and cerebral infarction contributed), the AAMR exhibited a significant decline from 1999 to 2004 (APC −10.4*; 95% CI: −12.3 to −7.5) and from 2004 to 2007 (APC −20.6*; 95% CI: −24.9 to −13.2). This was followed by a period of nonsignificant decline through 2013 (APC −4.7; 95% CI: −12.3 to 7.3), before transitioning into a significant upward trajectory through 2023 (APC 8.6*; 95% CI: 6.3–15.0).

Similarly, when restricting to cases where cerebral infarction was the UCD (and IHD contributed), a significant early decline was observed from 1999 to 2006 (APC −11.1*; 95% CI: −16.8 to −8.7), followed by a stable, nonsignificant trend through 2014 (APC −4.1; 95% CI: −8.8 to 11.2). Subsequently, mortality rates surged significantly from 2014 to 2017 (APC 24.5*; 95% CI: 1.7–32.7) and maintained a significant incline through 2023 (APC 6.3*; 95% CI: 1.5–9.9). These subgroup analyses demonstrate that the overall temporal pattern of an initial substantial decline followed by a recent significant increase remains robust regardless of which specific condition was the UCD (Fig. S3, Supplemental Digital Content 5; Tables 1 and S3, Supplemental Digital Content 6).

3.4. Racial and ethnic disparities

NH African Americans, though, exhibited a higher absolute mean AAMR (6.5) compared to NH Whites (5.9); this overall difference did not reach statistical significance across the 25-year study period (RR 1.10; 95% CI: 0.99–1.22, P = .0824). Conversely, mortality was significantly lower in Hispanics (RR 0.77; 95% CI: 0.67–0.89, P < .001) and NH Asians (RR 0.74; 95% CI: 0.61–0.91, P = .0038) compared to the NH White reference group.

NH African Americans exhibited the highest AAMR (6.5) compared to NH Whites (5.9), Hispanics (4.6), and NH Asians (4.4), respectively. In NH Asians, the AAMR declined greatly from 1999 to 2010 (APC −14.6*; 95% CI: −18.9 to −11.5), until 2011 in NH Whites (APC −11.5*; 95% CI: −13.7 to −10.2) and Hispanics (APC −10.3*; 95% CI: −13.1 to −8) and until 2013 in NH African Americans (APC −10.6*; 95% CI: −12.6 to −9.1). Following that, all races have shown a significant increase in AAMRs through 2023. The associated APCs were 7.6* (95% CI: 5.3–11.4) for NH Asians, 10.1* (95% CI: 8.3–13) for NH Whites, 9.7* (95% CI: 7.8–12.2) for Hispanics, and 12.6* (95% CI: 9.8–16.6) for NH African Americans (Fig. 3; Tables 1, 2 and S4, Supplemental Digital Content 7).

Figure 3.

Trends in ischemic heart disease and cerebral infarction-related age-adjusted mortality rates per 1,00,000, stratified by race and ethnicity among older adults in the United States, 1999 to 2023. AAMR = age-adjusted mortality rate.

3.5. Age group stratified

Among all the age groups, the highest mean CMR was observed in people in the age group 85+ years at 31.6 per 1,00,000 (95% CI: 30.2–33.1). At the same time, the lowest rate occurred in people in the age group 55 to 64 years at 1.0 per 1,00,000 (95% CI: 0.9–1.1). Individuals aged 55 to 64 (APC −6.5*; 95% CI: −9.0 to −4.7), 65-74 (APC −9.6*; 95% CI: −12.2 to −7.6), and 75 to 84 years (APC −11.5*; 95% CI: −14.0 to −9.6) showed a significant decline in mortality rates from 1999 to 2011 followed by sharp surges through 2023. The associated annual rates of increase were 10.3* (95% CI: 8.7–12.9) for the 55 to 64-year group, 9.5* (95% CI: 7.7–12.4) for the 65 to 74-year group, and 9.0* (95% CI: 6.5–12.5) for the 75 to 84-year group. However, populations aged 85+ experienced significant reductions from 1999 to 2003 (APC −9.5*; 95% CI: −12.2 to −3.5) and until 2006 (APC −22.4*; 95% CI: −26.0 to −16.3), which continued nonsignificantly through 2014 (APC −5.1; 95% CI: −8.0 to 0.4). A stable incline was then observed until 2018 (APC 18.7; 95% CI: −4.6 to 27.6), followed by a significant spike in rates, peaking in 2023 (APC 8.8*; 95% CI: 2.7 to 12.2; Fig. 4; Tables 1 and S5, Supplemental Digital Content 8).

Figure 4.

Trends in ischemic heart disease and cerebral infarction-related crude mortality rates, stratified by age groups among older adults in the United States, 1999 to 2023.

3.6. Regional death rate variations

Significant disparities were observed across geographies, with states (Ohio, Texas, Maryland, Washington, Vermont, and the District of Columbia) that belong to the top 90th percentile having triple the AAMR of states (Utah, Nevada, Georgia, Massachusetts, Montana, and Kansas) in the lower 10th percentile. Furthermore, the West region (6.0) showed the highest mean AAMR, followed by the South (5.9), Midwest (5.8), and Northeast (5.6). Additionally, the mean AAMR was significantly higher in nonmetropolitan regions (6) compared with metropolitan regions (5.3; RR: 1.10; 95% CI: 1.02–1.19, P = .0161). Metropolitan areas showed an initial significant fall in AAMRs from 1000 to 2011 (APC −11.8*; 95% CI: −14.6 to −10.1), after which they began to rise sharply through 2023 (APC 9.6*; 95% CI: 6.0–17.0). However, nonmetropolitan regions showed significant declines from 1999 to 2003 (APC −8.4*; 95% CI: −11 to −3) and until 2006 (APC −18.5*; 95% CI: −21.8 to −12.9), which gained stability through 2014 (APC −3.6; 95% CI: −7.1 to 2) before rising sharply until 2018 (APC 20.1*; 95% CI: 0.9–29.5). The rates then maintained a stable, nonsignificant incline in AAMR through 2020 (APC 5.8; 95% CI: −0.6 to 14.3; Fig. 5; Tables 1, 2 and S6–S8, Supplemental Digital Content 9).

Figure 5.

Trends in ischemic heart disease and cerebral infarction-related age-adjusted mortality rates per 1,00,000, stratified by urbanization among older adults in the United States, 1999 to 2020. Data stratification by urbanization is limited to the years 1999–2020 due to the availability constraints of the NCHS Urban-Rural Classification Scheme within the CDC WONDER dataset. AAMR = age-adjusted mortality rate, CDC WONDER = Centers for Disease Control and Prevention Wide-ranging Online Data for Epidemiologic Research, NCHS = National Center for Health Statistics.

4. Discussion

CVD is the major cause of death in the US In 2021, about 1 in every 5 deaths was associated with CVD.^[19] Our study reported that over the past 20 years, there has been a steady decline in the mortality rates from IHD and cerebral infarction across annual, demographic, gender, and geographic categories, showing a downward trend nationwide. However, our Joinpoint analysis identified a critical epidemiological inflection point around 2011, demarcating 2 distinct phases of rising mortality.

The AAMR decreased by 11.42% per year until 2011 and then rose dramatically by 9.64% per year. Therefore, the turnaround from the downward trend (2011–2019) is not accounted for by the COVID-19 pandemic. This pre-pandemic period is rather a reflection of the leveling off in the cardiovascular mortality gains made by the widespread use of statins and antihypertensives in the late ‘90s and 2000s. At the same time, it is associated with the growing public health burden of obesity, metabolic syndrome, and type 2 diabetes, all of which are upstream determinants of both IHD and cerebral infarction, frequently exacerbated by the poor social determinants of health.^[20] This further mirrors the national picture, as Shah et al found that in 1999, there were 7,52,192 deaths due to heart disease, but in 2011 this had reduced to 5,96,577, before rising to 6,55,381 by 2018, offset by decreases in IHD, yet increases in heart failure and hypertensive heart disease.^[21] Similarly, Woodruff et al discovered that CVD death rates declined 8.9% from 2010 to 2019, but then climbed 9.3% from 2019 to 2022.^[22]

The second phase, which started in 2020, marked a significant acceleration of the following upward trend. COVID-19 is recognized to have several complications related to CVDs, such as inflammatory diseases like myocarditis, or ischemic events such as myocardial and cerebral infarctions, both during acute and post-acute infection. The pandemic had a profound impact as a disease multiplier, affecting the existing cardio-cerebrovascular vulnerability and contributing to mortality through disruption of routine care of chronic diseases, shortages of healthcare workers, and delays or avoidance of medical help-seeking behaviors.^[22,23]

Females had greater absolute mortality (50.20% of deaths) than males (49.80%), although males had higher AAMR (7.2 vs 4.9 per 1,00,000 by 2023). The temporal trends were comparable for both sexes, with decreases through 2011 followed by large increases, but differed in terms of timing and magnitude. These findings are similar to the national statistics, which indicate that while age-adjusted IHD mortality rates were higher among males, women had a larger percentage reduction (average yearly percent change −4.7% for White women vs −3.7% for White men).^[24]

Significant racial/ethnic differences were observed during the course of the study. NH African Americans had the highest AAMRs (6.5/1,00,000) compared to other groups. Initial decreases in mortality were observed for all racial/ethnic groups until 2010–2013, after which there were considerable increases through 2023, with NH African Americans seeing the greatest growth after 2013 (12.6% each year). These discrepancies are consistent with critical evidence that NH African American people in the US still have the highest rates of death due to CVD.^[25]

The mortality rates for the 85+ age group were the highest, while the 55 to 64 age group was the lowest. Initial decreases followed by increases were found in all age groups, although with varying time and magnitude. The age group with the most diverse trend was 85+, with a substantial decrease between 2003 and 2006 before a steady decline until 2014, a significant rise between 2014 and 2018, and then continued increases through to 2023. National data reveal that the relative mortality reductions were largest for older people (≥65 years) compared with younger groups.^[26] However, over the past few years, stroke deaths in younger adults (aged 35–64) have leveled off or increased, with an increase of 5.2% in deaths per year since 2018, and death rates in older adults have slightly stabilized.^[27,28] Factors thought to be associated with the developments in the literature include better acute treatment of CVDs and better risk factor management in older people, while the socioeconomic determinants of health have deteriorated in the same period, and obesity, diabetes, and hypertension rates have increased.^[29]

Due to the high mortality rates in the oldest age groups, including octogenarians, clinical care has to go beyond classical, disease-oriented secondary prevention and incorporate comprehensive geriatric domains. High-risk profiles should be defined on the basis of both traditional risk factors (previous stroke/transient ischemic attack, heart failure, especially with reduced ejection fraction, atrial fibrillation/device-detected atrial high-rate episodes, uncontrolled cardiometabolic disease, chronic kidney disease, and anemia) and geriatric vulnerability factors (falls, delirium, polypharmacy, cognitive impairment, and functional dependence). To model competing risks, determine the intensity of follow-up, and optimize resource allocation, particularly in access-restricted nonmetropolitan areas, prognostic tools and stratification by score in elderly populations are essential.^[30] In addition, malnutrition and sarcopenia are very modifiable risk amplifiers in older adults. Regular clinical assessments using a tool like the Mini Nutritional Assessment-Short Form or the Geriatric Nutritional Risk Index to identify unintentional weight loss are important. The pathways of multidisciplinary nutritional optimization and rehabilitation are plausible ways to reduce arrhythmic burden, prevent deconditioning, and ultimately reduce mortality.^[31]

Significant geographic variations were seen among states and regions. States in the top 90th percentile had AAMRs nearly 3 times those in the bottom 10th percentile. Regionally, the West had the highest mortality rate (6.0), followed by the South (5.9), the Midwest (5.8), and the Northeast (5.6). Nonmetropolitan regions had considerably greater mortality rates than metropolitan areas (6.0 vs 5.3 per 1,00,000), with both showing comparable patterns of initial decrease followed by reversal. This is supported in the literature, with the South and nonmetropolitan areas continually having higher CVD mortality rates.^[32] Roth et al.’s study represented a distinct geographic belt of elevated cardiovascular mortality extending from southeastern Oklahoma through the Mississippi River Valley and reaching into eastern Kentucky. Counties that surrounded San Francisco, central Colorado, and northeastern Virginia, and the persisting rural-urban gap showed the lowest mortality, with nonmetropolitan regions having 18% to 21% greater CVD mortality than metropolitan areas, which could be caused by differences in healthcare access, delayed diagnosis, and the poor management of chronic cardiovascular conditions in marginalized and rural communities.^[33-35]

5. Limitations

This study has several limitations. First, the ecological design of this WONDER-based analysis necessitates cautious interpretation regarding causality. In this study, “co-occurring” refers strictly to a mortality-coding co-mention on the death certificate, rather than a clinically confirmed timeline of comorbidity onset or a definitively established physiological sequence. The accuracy of these trends is inherently reliant on the diagnostic precision of the certifying physicians and the administrative constraints of ICD-10 coding. Research has shown that deaths from CVD may be overreported on death certificates, while deaths from stroke are typically under-reported. Despite efforts, the 1999 switch from ICD-9 to ICD-10 coding may have artificially altered mortality trends. Second, the database suppresses data cells with fewer than twenty deaths to maintain confidentiality, which restricts county-level studies and may mask significant regional differences in the burden of CVD, especially in rural and sparsely populated areas. Third, county-level risk factor estimates do not represent individual-level exposures and cannot account for within-county heterogeneity in demographic characteristics, socioeconomic status, or health behaviors. Finally, age-adjustment methodology has inherent limitations, including the potential to obscure important age-specific trends, and changes in the standard population can alter both the magnitude of rates and comparisons between groups.

6. Conclusions

We found an overall decline in mortality trends related to IHD and cerebral infarction in the recent 2 decades, followed by a reversal beginning in 2011. The population groups with the greatest mortality burden are men, NH African Americans, the 85+ years old age group, nonmetropolitan areas, and the western states. Furthermore, from 2020 onward, a sharp acceleration in mortality was observed, which may reflect the compounded effects of the COVID-19 pandemic and a disrupted healthcare system, highlighting the importance of accessible and continuous care. Further studies utilizing patient-level data are needed to evaluate clinical interventions while providing an in-depth analysis of the behavioral, socioeconomic, and systemic elements that drive these disparities.

Author contributions

Conceptualization: Muhammad Shaheer Bin Faheem, Syed Tawassul Hassan.

Data curation: Muhammad Shaheer Bin Faheem, Syed Tawassul Hassan, Fahad Haider Shah, Syeda Lyba Onaiz, Muhammad Misam Raza.

Formal analysis: Muhammad Shaheer Bin Faheem, Syed Tawassul Hassan, Fahad Haider Shah, Syeda Lyba Onaiz.

Investigation: Fahad Haider Shah, Shaarif Rauf Khan.

Project administration: Muhammad Shaheer Bin Faheem, Syed Tawassul Hassan, Muhammad Idrees Khan.

Resources: Shaarif Rauf Khan, Muhammad Misam Raza, Muhammad Idrees Khan.

Software: Muhammad Shaheer Bin Faheem, Syed Tawassul Hassan, Fahad Haider Shah, Syeda Lyba Onaiz.

Supervision: Muhammad Shaheer Bin Faheem, Syed Tawassul Hassan.

Validation: Fahad Haider Shah, Syeda Lyba Onaiz, Shaarif Rauf Khan.

Visualization: Muhammad Shaheer Bin Faheem, Syed Tawassul Hassan, Fahad Haider Shah, Syeda Lyba Onaiz.

Writing – original draft: Muhammad Shaheer Bin Faheem, Syed Tawassul Hassan, Shaarif Rauf Khan, Muhammad Misam Raza.

Writing – review & editing: Muhammad Shaheer Bin Faheem, Syed Tawassul Hassan.