Premature mortality from acute myocardial infarction and psychoactive substance use in the United States: An observational trend (1999–2023) and forecasting (2035) study

Abstract

Background

Acute myocardial infarction (AMI) remains a leading cause of premature death (<65 years) in the United States, with a significant association with psychoactive substance use (PSU). However, long-term trends and future projections of this combined burden remain unclear.

Methods

We conducted a retrospective observational study using the Centers for Disease Control and Prevention Wide-Ranging Online Data for Epidemiologic Research Multiple Cause of Death database to assess premature mortality (ages 25–65 years) related to AMI and PSU from 1999 to 2023. Age-adjusted mortality rates (AAMRs) were derived, and annual percentage changes (APCs) were calculated using joinpoint regression. An autoregressive integrated moving average (ARIMA) model was used to forecast mortality through 2035.

Results

From 1999 to 2023, 176,641 premature deaths were attributed to AMI and PSU. The AAMR rose from 1.4 per 100,000 in 1999 to 4.16 in 2023. Mortality increased sharply from 1999 to 2005 (APC 19.04), continued rising through 2021 (APC 2.63), and declined between 2021 and 2023 (APC −9.13). Middle-aged adults (45–65 years) had substantially higher mortality than younger adults (25–45 years) (AAMR 9.63 vs 0.91, 2018–2023). Men had higher mortality than women (6.81 vs 2.53). American Indian/Alaska Native and White populations showed the higher racial burden. Nonmetropolitan areas had markedly higher mortality than metropolitan regions. South Dakota, Kentucky, and Arkansas had the highest state-level mortality. Forecasting predicts only a modest decline to 3.80 per 100,000 by 2035.

Conclusion

Despite recent declines, premature mortality from AMI and PSU is projected to remain high through 2035, underscoring the need for targeted, evidence-based prevention strategies.

Graphical abstract.

This is a visual representation of the abstract.

Introduction

Cardiovascular disease (CVD) is the leading cause of death in the United States (US), posing a major public health challenge. Each year, millions of people and their families feel its impact. ¹ Among the various forms of cardiovascular disease, acute myocardial infarction (AMI) is of major concern, particularly important because of its contribution to premature mortality, defined here as death before 65 years of age. This early loss brings about significant personal, social, and economic difficulties. ¹ While there have been strides in reducing overall CVD deaths due to advancements in prevention, early detection, and treatment, these benefits have been unequally distributed, along with emerging risk factors.^1,2 National studies show varying mortality trends influenced by race, ethnicity, age, and location with younger adults experiencing stagnation or even increases in premature CVD deaths, particularly from ischemic heart disease (IHD).^1,3 Compared to other high-income countries, the U.S. has also experienced less favorable trends in cardiovascular mortality, with a widening gap driven in part by slower declines in IHD deaths. ²

Psychoactive substance use (PSU) such as alcohol, tobacco, cannabis, opioids, sedatives, stimulants and hallucinogens can lead to serious cardiovascular risks.^4,5 Studies show cocaine and methamphetamines in particular have a significantly elevated risk of acute events like IHD, heart failure and pulmonary hypertension.^4,5 National data also highlight the growing burden of cardiovascular mortality associated with stimulant use, with IHD representing a major component of that burden. ⁵ Cigarette smoking has a well-established association with IHD with threefold increased risk. ⁶ Similarly, alcohol use disorder can increase the risk of IHD up to twofold compared to nondrinkers.^7,8

Evidence shows that men have a higher mortality rate from AMI than women and reflects factors beyond biology. The gap in mortality rates between urban and rural areas has widened—rural areas demonstrate significantly and consistently higher mortality rates and slower improvement in AMI outcomes.^9,10 Several factors, including differences in access to healthcare, social conditions, lifestyle, and economy, largely contribute to this disparity.

The main goal of this study is to fill the gap in our understanding by analyzing the trends of premature death due to AMI and PSU in the US from 1999 to 2023. Researchers used the data from Centers for Disease Control and Prevention Wide-Ranging Online Data for Epidemiologic Research (CDC WONDER) database to trend the mortality, compare demographic and geographic differences, and forecast the future estimates for mortality rate using machine learning models. Our key objectives are to (i) explore the trends of mortality due to AMI and PSU over the 25 years period; (ii) identify gaps based on age, sex, race, ethnicity, urbanicity, and states; (iii) analyze points of critical changes using joinpoint regression analysis; and (iv) project the mortality rate to 2035 based on current trends with autoregressive integrated moving average (ARIMA) modeling. This CDC WONDER trend analysis will provide targeted preventive strategies and guide on resource allocations, as well as provide evidence for policy changes aimed at reducing preventable death from this dual burden.

Methodology

Data availability

Researchers used the data sets that are publicly accessible in CDC WONDER. The study's outcomes can be easily replicated using the methodology discussed in the following section.

Data source

A retrospective observational study was conducted, in which we used the mortality data from death certificates through CDC WONDER database. It offers mortality data across various demographics. Data on premature death from AMI and PSU among adults aged 25 to 65 years between 1999 and 2023 in the US were utilized. Fatalities associated with AMI were based on International Statistical Classification of Diseases and Related Health Problems-10th Revision (ICD-10) codes: I21.0, I21.1, I21.2, I21.3, I21.4, and I21.9,^11–13 whereas for PSU, codes F10–F19 were employed which included substances such as alcohol, tobacco, cannabis, opioids, sedatives, stimulants, and hallucinogen.^10,13 Code I22 was intentionally excluded as we intend to reflect only the initial episodes of AMI. Multiple cause of Death dataset was implemented. This case definition identifies the co-occurrence of AMI and documented polysubstance use on the death certificate and should not be interpreted as evidence of a causal relationship or that PSU directly precipitated the fatal AMI. Our study utilized publicly available government data that doesn’t include any identifiable information and adhered to the STROBE guidelines for reporting observational studies; thus institutional review board (IRB) approval was not required. ¹⁴

Data abstraction

Researchers included various demographic factors—age, sex, race, ethnicity, urbanicity, and state for data collection. Ethnicity and race were determined as per the death certificates, which are based on self-identification. Ethnicity (Hispanic origin) and race were outlined separately in accordance with the Office of Management and Budget (OMB). ¹⁵ Between 1999 and 2020, races were classified as White, Black or African American, American Indian or Alaskan Native, and Asian. However, from 2018 to 2023, a six-race classification was employed (American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Other Pacific Islander, White, and More than one race) following the CDC WONDER classification based on the 1997 OMB's standards. ¹⁶ National Center for Health Statistics 2013 urban–rural classification was used for geographical classification and for simplification, we further divided them into: (i) large metro (large central and large fringe); (ii) medium/small metro; and (iii) nonmetro (micropolitan and noncore).

Statistical analysis

AAMRs per 100,000 individuals standardized to the 2000 US population were calculated along with 95% confidence intervals (CIs). These data was explored to assess the trends in premature mortality due to AMI and PSU from 1999 to 2023 across various demographics. ¹⁷ We used Joinpoint Regression Program (Version 5.4.0, National Cancer Institute) to study trends in mortality over time that involved log-linear regression models to calculate the annual percent change (APC).^18,19 The analyst adopted grid search method (22,0), between one and three joinpoints, with permutation test and parametric method to estimate the APCs with 95% CIs. This methodology helped to identify significant trends and inflection points over time by applying log-linear regression models to capture temporal fluctuations. APCs were classified as increasing or decreasing when the slope of regression line significantly differed from zero, based on two-tailed t-tests. A p < 0.05 was considered statistically significant.

We implemented ARIMA model to forecast the mortality rates to 2035. The optimal model specification was selected using the Akaike Information Criterion (AIC). An ARIMA (1, 1, 0) model was selected (AIC = 11.79, BIC = 14.15). Multiple diagnostic tests assessed the adequacy of the model. The Augmented Dickey Fuller test assessed stationarity (test statistic: −2.6, p: 0.092), the Ljung-Box test confirmed the absence of significant residual autocorrection (p: 0.850), and the Shapiro–Wilk test evaluated the normality of residuals (p: 0.001). Model fit was quantified using root mean squared error (RMSE: 0.395) and mean absolute error (MAE: 0.2464) (Supplemental Figure 1 and Supplemental Table 1). For age-wise categories, historical data were modeled using piecewise exponential trends based on APCs. Model performance was evaluated using a holdout test period comprising the most recent 3 years, with accuracy assessed using RMSE, MAE, mean absolute percentage error (MAPE), and overall accuracy. All analyses were conducted using Python 3.x with the statsmodel package. Statistical significance was set at α = 0.05 (Supplemental Table 1).

Results

Between 1999 and 2023, a total of 176,641 premature deaths occurred in a population of age group 25 to 65 years due to AMI and PSU in the USA. The average age-adjusted mortality rate (AAMR) for the recent 6 years (2018–2023; total death: 58,275) was 4.63 per year (95% CI, 4.59–4.67) per 100,000. Overall, the AAMR increased rapidly from 1.4 (95% CI, 1.33–1.46) per 100,000 in 1999 to 3.3 (95% CI, 3.21–3.38) per 100,000 in 2005, with an APC of 19.04 per year (95% CI, 13.84–24.48). Between 2005 and 2021, there was a steady growth with the APC in AAMR of 2.63 per year (95% CI, 1.90–3.37), followed by a decline between 2021 and 2023 [–9.13 per year (95% CI, −23.38–−7.79)] (Figure 1) (Supplemental Table 2).

Figure 1.

Historical trends and ARIMA forecast of AAMRs per 100,000 due to AMI and PSU among adults 25 to 65 years of age in the United States, 1999 to 2035. AAMR: age-adjusted mortality rate; AMI: acute myocardial infarction; ARIMA: autoregressive integrated moving average; PSU: psychoactive substance use.

Forecast to 2035: The ARIMA (1, 1, 0) model projected an AAMR from AMI and PSU of 3.80 per year (95% CI: 0.69–6.91) per 100,000 population in 2035, representing an 8.6% decrease from the 2023 rate of 4.16 per 100,000 (Figure 1). The forecast assumes continuation of the mortality dynamics observed during the 1999 to 2023 period. CIs widened progressively with forecast horizon, reflecting increasing uncertainty in long-term projections (CI width in 2025: 1.92; 2035: 6.22). The projected trends from 2023 to 2035 are included in Supplemental Table 2.

Trends in premature mortality due to AMI and PSU by age

The AAMR were significantly higher for middle-aged (45–65 years) adults than young adults (25–45 years) throughout 1999 to 2023 (Figure 2). Between 2018 and 2023, the AAMR among young and middle-aged adults were 0.91(95% CI, 0.88–0.94) and 9.63 (95% CI, 9.55–9.71), respectively.

Figure 2.

Historical trends and ARIMA forecast of age-adjusted mortality rates per 100,000 due to AMI and PSU among adults 25 to 65 years of age, stratified by age in the United States, 1999 to 2035. AMI: acute myocardial infarction; ARIMA: autoregressive integrated moving average; PSU: psychoactive substance use.

Table 1 reports age-adjusted annual AMI mortality rates among young and middle-aged adults from 1999 to 2023. Among young adults the AAMR significantly increased with APC 13.41 between 1999 and 2005, with steady growth (APC, 2.24) between 2005 and 2021 and declined (APC, −5.49) between 2021 and 2023. The middle-aged adults showed similar pattern with APC of 16.9, 2.29, −8.69 between 1999–2006, 2006–2021, and 2021–2023, respectively.

Table 1.

Historical trends and forecast in age-adjusted mortality rates per 100,000 and annual percentage change in mortality due to AMI and PSU among adults, stratified by age categories in the United States, 1999 to 2035.

Year	25–45 years AAMR (95% CI)	APC % (95% CI)	45–65 years AAMR (95% CI)	APC % (95% CI)
1999	0.32 (0.28–0.36)	13.41 (8.52–18.52)	2.84 (2.71–2.98)	16.90 (12.66–21.31)
2000	0.37 (0.33–0.41)		2.98 (2.85–3.12)
2001	0.37 (0.33–0.41)		2.93 (2.79–3.06)
2002	0.43 (0.38–0.47)		2.96 (2.83–3.09)
2003	0.54 (0.49–0.59)		5.01 (4.84–5.17)
2004	0.64 (0.58–0.69)		5.88 (5.71–6.06)
2005	0.69 (0.63–0.75)	2.24 (1.42–3.07)	6.80 (6.61–6.98)
2006	0.75 (0.69–0.80)		6.84 (6.65–7.02)	2.29 (1.49–3.10)
2007	0.69 (0.63–0.75)		7.31 (7.13–7.50)
2008	0.69 (0.63–0.75)		8.05 (7.85–8.25)
2009	0.65 (0.59–0.70)		7.49 (7.31–7.68)
2010	0.69 (0.63–0.75)		8.03 (7.84–8.22)
2011	0.80 (0.74–0.86)		8.37 (8.17–8.56)
2012	0.80 (0.74–0.86)		8.80 (8.61–9.00)
2013	0.80 (0.74–0.86)		9.12 (8.92–9.32)
2014	0.75 (0.68–0.81)		9.04 (8.84–9.24)
2015	0.80 (0.74–0.86)		9.30 (9.10–9.50)
2016	0.75 (0.69–0.81)		9.59 (9.39–9.80)
2017	0.85 (0.79–0.92)		9.57 (9.37–9.77)
2018	0.85 (0.79–0.92)		9.49 (9.29–9.70)
2019	0.85 (0.79–0.92)		9.37 (9.17–9.57)
2020	1.01 (0.94–1.08)		10.29 (10.07–10.50)
2021	1.06 (0.99–1.13)	−5.49 (−21.56–13.88)	10.46 (10.25–10.68)	−8.68 (−21.86–6.72)
2022	1.01 (0.94–1.08)		9.59 (9.38–9.79)
2023	0.80 (0.74–0.86)		8.68 (8.48–8.87)
Projected Rates 2030−2035
2030	0.82 (0.46–1.19)		7.72 (6.41–9.03)
2031	0.81 (0.42–1.21)		7.59 (6.19–9.0)
2032	0.81 (0.39–1.22)		7.49 (6.01–8.98)
2033	0.8 (0.36–1.24)		7.41 (5.85–8.98)
2034	0.8 (0.34–8.98)		7.34 (5.7–8.98)
2035	0.79 (0.31–1.27)		7.29 (5.58–9.0)

AMI: acute myocardial infarction; APC: annual percentage changes; PSU: psychoactive substance use.

Forecast to 2035: The ARIMA (1, 1, 0) model projected an AAMR from AMI and PSU of 0.79 (95% CI, 0.3–1.3) and 7.29 (95% CI 5.6 −9.0) per 100,000 population among 25 to 45 years adults and 45 to 65 years adults respectively in 2035. For individuals aged 25 to 45 years, the projected AAMR remains approximately unchanged from 2023 [0.80 (95% CI, 0.74–0.86)], whereas for those aged 45 to 65 years, the projected AAMR shows a slight decline relative to 2023 [8.68 (95% CI, 8.48–8.87)] (Table 1).

Trends in premature mortality due to AMI and PSU by sex

Overall, age-adjusted AMI mortality rates were consistently higher in male than female throughout 1999 to 2023 (Figure 3, Supplemental Table 3). Between 2018 and 2023 the AAMR among male and female were 6.81 per year (95% CI, 6.74–6.88) and 2.53 per year (95% CI, 2.49–2.57), respectively. Among males the AAMR significantly increased with APC 18.76 between 1999 and 2005, with steady growth (APC, 2.19) between 2005 and 2021 and has been trending down (APC, −10.22) between 2021 and 2023. The females showed similar pattern with APC of 21.53, 3.51, −8.01 between 1999–2005, 2005–2021, and 2021–2023, respectively (Supplemental Table 3).

Figure 3.

Historical trends of age-adjusted mortality rates per 100,000 due to AMI and PSU among adults 25 to 65 years of age, stratified by sex in the United States, 1999 to 2023. AMI: acute myocardial infarction; PSU: psychoactive substance use.

Trends in premature mortality due to AMI and PSU by ethnicity and race

Overall, AAMR due to AMI and PSU were higher in non-Hispanic population than in Hispanic/Latino populations throughout 1999 to 2023 (Figure 4A). Between 2018 and 2023, with ethnicity reported in 94.75% of total death, the AAMR among non-Hispanic and Hispanic/Latino population were 5.19 per year [95% CI, 5.14–5.23]) and 1.54 per year (95% CI, 1.49–1.6), respectively. Among the Hispanic or Latino population, the AAMR initially increased with an APC of 17.71 between 1999 and 2005, and afterward has been steady (APC, 0.55) between 2005 and 2023. The non-Hispanic populations showed a similar pattern with APC of 19.55 and 3.02 between 1999–2005 and 2005–2021, respectively, with a recent drop (APC, −9.34) between 2021 and 2023.

Figure 4.

Trends in age-adjusted mortality rates per 100,000 due to AMI and PSU among adults 25 to 65 years, stratified by (a) Ethnicity/Hispanic origin. (b) Race (four classifications) in the United States, 1999 to 2023. AMI: acute myocardial infarction; PSU: psychoactive substance use.

Regarding race, between 1999 and 2020, AAMR was highest among American Indian or Alaska Native [4.17 (95% CI, 3.98–4.36)] followed by White [3.85 (95% CI, 3.83–3.87)], Black or African American [3.45 (95% CI, 3.4–3.5)] and significantly lower among Asian or Pacific Islander [0.76 (95% CI, 0.73–0.8)] (Figure 4B). APCs showed a consistent inclination in AMMR rates for all races, except sudden drop between 2006 and 2009 among American Indian or Alaska Native ( Supplemental Figure 2 ). Between 2018 and 2023, the updated six race classifications in order of decreasing AAMRs were White (AAMR, 5.00), American Indian or Alaska Native (AAMR, 4.95), Black or African American (AAMR, 4.50), Native Hawaiian or Other Pacific Islander (AAMR, 3.52), more than one race (AAMR, 1.79), and Asian (AAMR, 0.86) (Supplemental Table 4).

Trends in premature mortality due to AMI and PSU by urbanicity

Age-adjusted AMI mortality rates were significantly higher in the nonmetro region than in the metro region (Figure 5). Between 1999 and 2020, AAMR among nonmetro, small/medium metro, and large metro were 6.74 per year (95% CI, 6.68–6.81), 4.2 per year (95% CI, 4.17–4.24), and 2.43 per year (95% CI, 2.41–2.45), respectively.

Figure 5.

Trends in age-adjusted mortality rates per 100,000 due to AMI and PSU among adults 25 to 65 years, stratified by urbanicity in the United States, 1999 to 2020. AMI: acute myocardial infarction; PSU: psychoactive substance use.

Trend analysis showed that large metropolitan areas experienced a brief, decline from 1999 to 2001 (APC −1.26), followed by a sharp significant increase from 2001 to 2004 (APC 37.06%) and a continued rise from 2004 to 2020 (APC 1.47%). Medium/small metropolitan areas had similar pattern with a gradual increase from 1999 to 2002 (APC 4.19), a significant rise from 2002 to 2005 (APC 26.59%), and a sustained increase from 2005 to 2020 (APC 2.79%). Nonmetropolitan areas showed consistently significant rise throughout the period, with the steepest rise from 1999 to 2005 (APC 17.63%), followed by slower but continued increases from 2005 to 2013 (APC 5.93%) and 2013 to 2020 (APC 2.63%). Urbanization classification data after 2020 were not available in CDC WONDER database.

Trends in premature mortality due to AMI and PSU by the US states

Figure 6 illustrates the distribution of AAMR due to AMI and PSU across the US states from 2018 to 2023. High burden States (AAMR >90th percentile) were South Dakota (15.26), Kentucky (14.13), Arkansas (13.98), Tennessee (10.57), Mississippi (10.22), and Idaho (9.32) and low burden states (AAMR <10th percentile) were California (1.03), Connecticut (1.81), Hawaii (2.51), District of Columbia (2.58), Nevada (2.65), and Massachusetts (2.68). Supplemental Table 5 documents trends in AAMR across the US states at different points from 1999 to 2023. Overall, APC trends show early periods of rapid increase in several states, followed by slower growth, stabilization, or declines in more recent years, with many later APC estimates becoming smaller and often nonsignificant (Supplemental Table 6).

Figure 6.

Premature mortality burden from AMI and PSU across US states, 2018 to 2023. AMI: acute myocardial infarction; PSU: psychoactive substance use.

Discussion

This comprehensive study of 176,641 premature deaths from co-occurrence of AMI and PSU over a 25-year study period demonstrates a persistent and significant public health hurdle in the US. Though some optimism is observed in mortality rate reduction from 2021 to 2023, our ARIMA forecast projected the premature mortality linked with this dual burden to continue to be high through 2035 with anticipated AAMR of 3.80 per 100,000 (with a projected decline of only 8.6%). This persistent high mortality rate in a high-income country with an advanced healthcare system emphasizes the critical need for extensive multilevel interventions addressing both cardiovascular health and PSU disorders.

Our study reveals a remarkable disparity in genders, with males facing nearly three-fold higher AAMR than females (6.81 vs 2.53; 2018–2023) which represent complex interplay of biological, behavioral, and social factors. Previous analysis of national data has documented higher premature AMI mortality among men ⁹ and this difference persisted in our analysis even when investigating premature mortality with co-occurrence of AMI and PSU. Biologically, the sex hormone estrogen provides protective benefits against atherosclerosis and IHD in women. ²⁰ Behavioral aspects are particularly significant, as males with substance use disorders exhibit higher rates of concurrent tobacco consumption, alcohol misuse, and depression—all recognized cardiovascular risk factors. ⁴ Addressing this gender disparity necessitates targeted screening and preventive strategies prioritizing males with substance use disorders, integrated care models addressing substance use treatment along with cardiovascular risk management, and interventions targeting social determinants like education and obesity in high-risk populations.

The higher AAMR in individuals aged 45 to 65 compared with 25 to 44 is likely multifactorial. Decades of substance use lead to cumulative organ damage, raising mortality risk. ²¹ Premature mortality in this group reflects the combined effects of long-term substance exposure, established atherosclerotic risk factors, higher comorbidity burden, and the additive cardiotoxic impact of substances on an already vulnerable cardiovascular system.

Similarly, nonmetropolitan areas demonstrated nearly threefold higher mortality rates than large metropolitan areas in our analysis. Over the course of the study period, this disparity has further widened, with nonmetropolitan regions consistently displaying higher mortality rates and slower rates of improvement. There are multiple elements contributing to this trend. Healthcare professionals’ shortage, delays in specialist consultations, and limited access to emergent cardiac service creates a challenging scenario for healthcare access in rural settings. Although telemedicine has been shown to improve access, reduce costs, and be clinically effective in rural US settings, there are considerable limitations. These include insufficient broadband infrastructure, inability to perform physical examinations and patient-level challenges such as low digital literacy. ²² Moreover, in acute events like AMI, telemedicine can be of limited help where quick access to emergency services is not available and long travel time is needed for cardiovascular imaging. Yet, telemedicine can be utilized for preventive healthcare and the reduction of cardiovascular risk, thereby assisting in mitigating healthcare shortages in rural setting. ²³ Socioeconomic factors further compound these access barriers, due to elevated poverty levels, lower educational profile and awareness, and higher prevalence of cardiovascular risk factors like obesity, smoking and high-risk lifestyle. Obesity and low educational background particularly have demonstrated highest likelihood of premature CVD death.^1,24 Additionally, higher incidences of substance use disorders, specifically methamphetamine and prescribed opioids are observed in rural settings which indicate both restricted access to treatment options and the effect of socioeconomic burden.^25,26 Furthermore, urban–rural analyses are limited to data through 2020. This constraint must be considered when interpreting contemporary rural–urban disparities.

The high burden states exceeding the 90^th percentile—South Dakota, Kentucky, Arkansas, Tennessee, Mississippi, and Idaho—further represents regional concentration and geographical clustering. These states exhibit parallel characteristics including rural settings, lower median incomes, higher uninsured rates, and limited healthcare infrastructure. This identification of states that have mortality rates nearly nine times higher than low burden states like California (1.03 per 100,000) highlights significant geographic variability and opportunities for focused interventions. Assuming high-burden states could theoretically achieve mortality profiles comparable to low-burden states, a substantial number of premature deaths might potentially be prevented annually. However, realizing such reductions presents complex challenges, as these disparities are deeply influenced by underlying variations in population demographics, socioeconomic conditions, regional healthcare infrastructure, substance-use patterns, and structural determinants. The nine-fold disparity between the states indicates that mortality reduction is challenging but achievable. Primary prevention is the most promising strategy for reduction of mortality, as both AMI and PSU are predominantly preventable conditions. Future state-level action plans and public health hypotheses should consider evaluating a broad spectrum of interventions, such as comprehensive tobacco control initiatives, the potential impact of Medicaid expansion on healthcare access, school-based health education, workplace wellness programs, enhanced prescription drug monitoring, expanded treatment availability for substance use disorders, and integrated public health campaigns addressing both cardiovascular health and substance use. Further large-scale research is required to determine the direct efficacy of these specific measures in mitigating the observed disparities.

As highlighted in our analysis, the projected premature mortality by 2035 signifies a considerable failure in public health and attracts significant economic burdens. Death occurring during the most productive years (ages 25–65) of life is especially critical, leading to diminished earning, reduced tax income, and workforce depletion, which is a loss for individuals, families, communities and the entire nation. What is alarming is the anticipated persistence through 2035, when it is already known that these are preventable conditions. This situation represents a public health emergency and indicates an inefficient use of resources, with major funds allocated to acute care and chronic disease management that could be utilized toward more efficient preventive strategies.

The recent fall in mortality rates from 2021 to 2023 (APC: −9.13%), indicates that rapid evolution is achievable, likely reflecting the effective strategies—increased naloxone access, medication-assisted treatment for opioids use disorder, and public health initiatives. ²⁷ However, for this swift decline to continue, steady commitment and resources will be required.

From a public health perspective, these findings generate hypotheses supporting the potential value of broader implementation and evaluation of evidence-based strategies, such as cardiovascular risk screening, tobacco cessation, substance-use treatment, integrated care pathways, and improved access to preventive services. To test these potential impacts, further research is needed to examine whether policies ensuring adequate coverage for preventive services and substance use treatment across private insurers, Medicare, and Medicaid could influence these trends. The observed overlap between cardiovascular disease patterns and substance use challenges highlights a critical area of public health concern, suggesting that coordinated, comprehensive, and sustained investigative and programmatic efforts warrant further study to better understand their role in mitigating premature mortality.

Limitations

There are few limitations in our study when interpreting the results. First, the mortality data derived from death certificates may be vulnerable to misclassification, especially substance use, which is likely to be underreported due to stigma, incomplete toxicologic evaluation or insufficient documentation—which might lead to an underestimation of actual burden. Second, while “multiple cause of death” databases represent the co-occurrence of AMI and PSU, it remains unclear whether substance use directly triggered the AMI, contributed to preexisting cardiovascular conditions, or was simply an incidental comorbidity. Third, our analysis doesn’t consider variation in coding practices, diagnostic criteria, or death certification methods over this 25-years study period, which could impact the observed trend. Fourth, data on urbanicity were not available after 2020, limiting the access to evaluate recent trends in rural–urban disparities. Fifth, ARIMA projection depends on the assumption that historical patterns will persist and cannot adjust potential future interventions or change in pattern of substance use behaviors. Sixth, interpretation of race-specific trends across the entire study period should be approached with caution due to changes in racial classification standards, which utilize four race categories through 2020 and transitions to six categories after 2018. This structural modification in data collection complicates direct temporal comparisons across specific racial groups over the full duration of the study. Lastly, our analysis focused on mortality and doesn’t account nonfatal AMI occurrences, which could offer clearer insights into actual AMI burden associated with substance use.

Conclusions

In the US, premature death due to co-occurring AMI and PSU continues to be a significant public health concern, with observed disparities between sex, region, and geography. Although our analysis presented some optimism due to recent mortality declines, the forecast suggests persistence of the high mortality rates until 2035, thereby continuing to place a significant burden on society. The preventable nature of both conditions highlights the crucial need for comprehensive, evidence-based interventions including primary prevention, healthcare system reforms, and policy revisions to address social determinants of health. High burden states and nonmetropolitan regions require focused attention and resources. Even small efforts in prevention strategies could lead to a significant decline in preventable premature mortality.