Trends in Psychotropic-Drug-Implicated Cardiovascular Mortality: Patterns in U.S. Mortality 1999–2020

Abstract

Introduction:

Psychotropic-drug-implicated (PDI) mortality – deaths in which psychotropic drugs were a contributing but not underlying cause of death – increased over two decades, with circulatory mortality the primary cause leading such deaths. Trends in PDI circulatory mortality over a 22-year period and its patterning in U.S. deaths are described.

Methods:

Deaths extracted from the CDC WONDER Multiple Causes of Death database from 1999–2020 were analyzed to generate annual counts and rates for drug-implicated deaths due to diseases of the circulatory system, including by specific drug, gender, race/ethnicity, age, and state.

Results:

During a period when overall age-adjusted circulatory mortality rates declined, PDI circulatory mortality more than doubled, from 0.22 per 100,000 in 1999 to 0.57 per 100,000 by 2020, now representing 1 in 444 circulatory deaths. While PDI deaths from ischaemic heart diseases are proportionally aligned with overall circulatory deaths (50.0% vs. 48.5%), PDI deaths from hypertensive diseases represent a larger proportion (19.8% vs. 8.0%). Psychostimulants generated the greatest escalation for PDI circulatory deaths (0.029 to 0.332 per 100,000). The sex gap in PDI mortality rates widened (0.291 females; 0.861 males). PDI circulatory mortality is particularly notable for Black Americans and midlife Americans, with considerable geographic variability.

Conclusions:

Circulatory mortality with psychotropic drugs as a contributing cause escalated over two decades. Trends in PDI mortality are not evenly distributed across the population. Greater engagement with patients about their substance use is needed to intervene on cardiovascular deaths. Prevention and clinical intervention could contribute to reinvigorating prior trends of declining cardiovascular mortality.

Introduction

Since 1999, over 1 million overdose deaths occurred in the U.S., which has contributed to reductions in life expectancy.^1–3 While such mortality cannot be overlooked, attention has also centered on increases in psychotropic-drug-implicated (PDI) mortality – deaths in which psychotropic drugs are a contributing cause but not the underlying cause of death.⁴ In particular, cardiovascular deaths in which psychoactive drugs are implicated remain an especially important area of further study, as they represent considerable components in PDI death trends⁴ and cardiovascular mortality remains the leading cause of death.⁵ Examinations of PDI cardiovascular death trends may identify foci for improving population health, especially within areas in which clinicians have avenues for intervention.

Over 600,000 cardiovascular-related deaths occur in the U.S. each year.⁵ Cardiovascular mortality has declined within the U.S. for several decades (from 520.0 per 100,000 in 1969 to an estimated 145.3 per 100,000 in 2020),^6,7 although a deceleration of this decline began in 2011.^8,9 Such declines resulted from improvements in clinical care – including prevention, diagnosis, and treatment – and behavioral factors such as smoking and diet.¹⁰ Yet, psychotropic drug use and related patterns of mortality may potentially offset gains made in reducing cardiovascular deaths. As such, examining how patterns of PDI cardiovascular mortality are situated within general cardiovascular mortality trends remains important.

Psychotropic drugs have wide-ranging effects on the cardiovascular system. Although findings are mixed, some studies indicate opioids affect the cardiovascular system through both acute (hypotension, bradycardia) or chronic use (vascular, valvular, and arrhythmic complications),¹¹ and withdrawal has implications for myocardial infarction, via stress-induced cardiomyopathy.¹² Opioid use in patient samples elevates risks of adverse cardiovascular events.^13–15 Via sympathetic nervous system stimulation, cocaine has varied effects on the cardiovascular system, including vasoconstriction, inhibition of sodium and potassium channels, and increased myocardial contractibility, which can produce adverse outcomes such as arrhythmias, myocardial infarction, cardiomyopathy, and coronary artery disease,¹⁶ resulting in higher mortality rates among cocaine users.¹⁷ Methamphetamine can produce tachycardia, hypertension, myocardial infarction, and aortic dissection.^18–20 While benzodiazepines alleviate problems associated with poor cardiovascular outcomes, such as anxiety and sleep disorders, population-based research identifies benzodiazepine use with increased cardiovascular mortality.²¹ Importantly, concerns about cardiovascular effects extend beyond additional drug use risks, such as injection, which may introduce pathogens leading to endocarditis.²² Thus, psychotropic drugs shape cardiovascular health and mortality in varied ways.

Social determinants of psychotropic drug use and cardiovascular health also may influence how PDI cardiovascular mortality becomes patterned across the population.^23–24 For example, sex impacts the detrimental and ameliorating influences for both men and women.²⁵ While men’s cardiovascular mortality rates remain notably higher than women’s,^10,26 scholars have also highlighted growth in cardiovascular mortality among midlife women.²⁷ Racial disparities in cardiovascular mortality are well established, and although declines in cardiovascular mortality occurred broadly, racial disparities persist.^26,28 Additionally, while cardiovascular mortality primarily accumulates in later life, overdose trends heavily affect midlife adults,²⁹ and PDI cardiovascular mortality may follow similar patterns. Identifying whether and how PDI cardiovascular mortality is differentially shaped across the population may contribute to prevention of adverse outcomes.

The objective was to examine PDI cardiovascular deaths within a context of declines in overall cardiovascular mortality rates. The paper provides an examination of specific psychoactive drugs that impact PDI cardiovascular deaths and specific forms of cardiovascular mortality involved. Lastly, the paper details differentiation across the population by sex, race/ethnicity, and age differences in PDI cardiovascular mortality, as well as across U.S. states.

Methods

Study Population

CDC WONDER Multiple Cause of Death data was utilized for a twenty-two-year period, 1999–2020.³⁰ Coroners and medical examiners list a single primary cause of death, the underlying cause, and may list up to 10 contributing causes of death, “multiple” factors that contributed to the death but were not the primary cause. Deaths in which poisoning from a psychotropic drug was deemed the underlying cause of death (ICD-10 codes X40–X44, X60–X64, X85, and Y10–Y14) were excluded to focus on deaths in which psychotropic drugs were implicated as contributing to deaths in which the underlying cause of death was of circulatory origin. Specifically, these deaths have ICD-10 classifications of I00–I99, Diseases of the Circulatory System, as the underlying cause of death. More specific classifications of circulatory deaths include: acute rheumatic fever (I00–I02), chronic rheumatic heart diseases (I05–I09), hypertensive diseases (I10–I15), ischaemic heart diseases (I20–25), pulmonary health disease and diseases of pulmonary circulation (I26–I28), other forms of heart disease (I30–I51), cerebrovascular diseases (I60–I69), diseases of arteries, arterioles and capillaries (I70–I78), diseases of the veins, lymphatic vessels, and lymph nodes, not elsewhere classified (I80–I89), and other unspecified disorders of the circulatory system (I95–I99). Although most deaths are cardiovascular in nature, the term “circulatory” is used to adhere to the ICD-classification of “Diseases of the Circulatory System.”

Measures

Analyses focused on deaths for which the underlying cause of death was related to the circulatory system combined with multiple cause of death codes in any of the following drug categories: narcotics and psychodysleptics (T40.0–T40.9); antiepileptic, sedative-hypnotic and antiparkinsonism drugs (T42.0–T42.8); and psychotropic drugs, not elsewhere classified (T43.0–T43.9). These are drug categories in which psychoactive controlled substances fall, not general medications. To be clear, deaths examined had an underlying cause of death of circulatory system origin, not overdose, but in which these multiple cause of death codes for psychotropic drugs were included as contributing causes to circulatory deaths. Such deaths are not counted with overdose mortality, but a psychotropic drug was present in the decedent at the time of death and identified as contributing to the death. To investigate specific psychotropic substances implicated in these deaths, deaths implicating opioids (T40.0–T40.4; T40.6), cocaine (T40.5), cannabis (T40.7), benzodiazepines (T42.4), and psychostimulants with abuse potential (T43.6) were examined. In recognition of the complexity of the opioid crisis, opioids were also disaggregated into heroin (T40.1), natural and semi-synthetic (T40.2), methadone (T40.3), and synthetic other than methadone (T40.4). When a substance is completely unknown, it may potentially get coded as “unspecified drugs, medicaments, and biological substances” (T50.9). However, given that this category can include non-psychotropic substances, it was not included (see Vuolo, Frizzell, and Kelly⁴ for more detailed information).

Statistical Analyses

As these analyses provide an overview of trends in PDI cardiovascular deaths and the data include a complete census of deaths, descriptive statistics are provided, with data shown graphically. Additionally, CDC standard joinpoint regression methodology was used for assessing significant changes (p<.05) in aggregate rates over time³¹ using the National Cancer Institute’s Joinpoint software.³² As is standard for age-based rates, analyses for mortality rates stratified by age are crude rates (death divided by population without age-adjustment), while most other analyses are age-adjusted. The year 2000 is the default population selection for the calculation of age-adjusted rates with standard 10-year age groupings. For age figures, the categories for <1 year old, 1–4 years old, and 5–14 years old are not shown, as very few PDI deaths occur for these groups (11 deaths total). Finally, for numbers by state and District of Columbia (DC), numbers for the entirety of 1999 to 2020 are shown, rather than by year, to depict age-adjusted rates. Data analyses occurred in 2021 and 2022. The research received IRB approval from The Ohio State University and Purdue University.

Results

Trends in overall circulatory deaths (primary axis) and PDI circulatory deaths (secondary axis) are depicted in Figure 1, where the lines represent age-adjusted death rates per 100,000. A contrast is notable that while overall circulatory mortality declined, PDI circulatory mortality increased. Specifically, PDI deaths escalated from 0.22 deaths per 100,000 during 1999 to 0.57 deaths per 100,000 during 2020. Joinpoint regression analyses revealed that periods of statistically significant increases occurred from 1999 to 2004 and again from 2011 to 2020. In 1999, 0.64 per 1000 circulatory deaths were PDI, whereas by 2020 2.25 per 1000 circulatory deaths were PDI (Figure 2). These numbers are the equivalent of 1 of every 1,564 circulatory deaths in 1999 increasing to 1 of every 444 in 2020.

Figure 1: Trends in Overall Circulatory Mortality and Psychotropic-Drug-Implicated Circulatory Mortality: 1999–2020.

Notes: Data are from the Center for Disease Control and Prevention Wide-ranging ONline Data for Epidemiologic Research for 1999–2020. Corresponding to the primary y-axis, the gray line with circles shows the decreasing trend in age-adjusted death rate for all deaths due to underlying cause of diseases of the circulatory system. Corresponding to the secondary y-axis, the black line with triangles shows the increasing trend in age-adjusted death rate for deaths due to underlying cause of diseases of the circulatory system that were psychotropic-drug-implicated (PDI).

Figure 2: Trends in the Proportion of Circulatory Deaths per 1000 that are Psychotropic-Drug-Implicated.

Notes: Data are from the Center for Disease Control and Prevention Wide-ranging ONline Data for Epidemiologic Research for 1999–2020. The line shows the increasing trend in the number of deaths due to underlying cause of diseases of the circulatory system per 1000 that were psychotropic-drug-implicated.

The most significant proportion of PDI circulatory deaths were classified as ischaemic heart diseases (I20–I25), constituting half of all PDI circulatory deaths (see Table 1). This was followed by hypertensive diseases (I10–I15; 19.8%), other forms of heart disease (I30–I51; 13.8%), cerebrovascular diseases (I60–I69; 9.6%), and diseases of arteries, arterioles and capillaries (I70–I78; 4.9%). Within these categories, one ICD-10 sub-chapter tended to dominate. Respectively, these were chronic ischemic heart disease (I25; 41.0% of all PDI deaths), hypertensive heart disease (I11; 18.6%), cardiomyopathy (I42, 7.2%), nontraumatic intracerebral (I61; 3.8%) or subarachnoid (I60; 3.3%) hemorrhage, and aortic aneurysm and dissection (I71; 4.5%). A full breakdown of sub-chapters is shown in Appendix Table 1. All other classifications constituted negligible proportions of PDI deaths. While PDI deaths from ischaemic heart diseases are proportional to overall circulatory deaths (50.0% vs. 48.5%), PDI deaths from hypertensive diseases constitute a larger proportion (19.8% vs. 8.0%).

Table 1:

Psychotropic-Drug-Implicated Circulatory Deaths by ICD-10 specific classification

	Psychotropic-drug-implicated		All
ICD Classification	Total Deaths	Percent	Total Deaths	Percent
I20-I25. Ischaemic heart diseases	13,651	50.0%	9,108,644	48.5%
I10-I15. Hypertensive diseases	5,391	19.8%	1,497,884	8.0%
I30-I51. Other forms of heart disease	3,776	13.8%	3,829,500	20.4%
I60-I69. Cerebrovascular diseases	2,615	9.6%	3,184,602	17.0%
I70-I78. Diseases of arteries, arterioles and capillaries	1,350	4.9%	662,201	3.5%
I80-I89. Diseases of veins, lymphatic vessels and lymph nodes, not elsewhere classified	332	1.2%	86,998	0.5%
I26-I28. Pulmonary heart disease and diseases of pulmonary circulation	131	0.5%	326,932	1.7%
I05-I09. Chronic rheumatic heart diseases	34	0.1%	73,743	0.4%
I95-I99. Other and unspecified disorders of the circulatory system	1	0.0%	12,579	0.1%
I00-I02. Acute rheumatic fever	0	0.0%	708	0.0%

Notes: Data are from the Center for Disease Control and Prevention Wide-ranging ONline Data for Epidemiologic Research for 1999–2020. The numbers depict the total deaths due to an underlying cause of diseases of the circulatory system by sub-chapter of the International Classification of Diseases, Tenth Revision (ICD-10) that are psychotropic-drug-implicated and the total. The percent corresponds to the column percentage. For the next sub-level of ICD-10 PDI circulatory death numbers and percentage, see Appendix 1.

With respect to types of drugs involved, crude death rates across the five most prevalent substances are shown (Figure 3; crude rates are used due to low numbers for some substances). In the top panel, psychostimulant drugs (such as methamphetamine) witnessed the sharpest increase (0.029 to 0.332), with a notable escalation from 2011 onward. Steady increases in opioid-implicated circulatory deaths occurred over the course of the period, growing from 0.059 in 1999 to 0.147 by 2020. Cocaine-implicated deaths peaked at 0.213 in 2006 followed by a decline to 0.115 in 2013 only to increase again to 0.180 by 2020. Benzodiazepine-implicated deaths were modest but increasing and cannabis-implicated deaths remained low across the period of observation. (For drug type by ICD-10 sub-chapters, see Appendix Table 2.) The bottom panel shows categories of opioid-implicated deaths. Up to 2010, the increase was largely attributed to natural and semi-synthetic opioids, which largely include pharmaceutical opioids, before leveling off for nearly a decade. Beginning in 2014, synthetic opioids other than methadone, which would largely represent fentanyl and its analogues, increased considerably to become the dominant opioid by 2020.

Figure 3: Trends in the Types of Psychotropic Drugs Implicated in Circulatory Deaths.

Notes: Data are from the Center for Disease Control and Prevention Wide-ranging ONline Data for Epidemiologic Research for 1999–2020. The top panel shows trends in the crude death rate per 100,000 due to psychotropic-drug-implicated (PDI) circulatory deaths for specific implicated controlled substance categories. The bottom panel shows the same for sub-categories of opioids. Definitions of psychotropic drug categories follow the International Classification of Diseases, Tenth Revision (ICD-10).

Deaths due to cocaine or psychostimulants without other drug classes experienced the greatest change over the period of observation (Appendix Figure 1). Thus, although polydrug use has been a notable factor in drug overdose mortality, the trends in PDI circulatory mortality do not appear to be driven by the use of multiple drug classes. Although cannabis-related PDI circulatory deaths are limited, 37.6% of cannabis-implicated deaths also implicated another psychotropic drug.

Sex differences in PDI circulatory mortality (Appendix Figure 2; middle) are characterized within a context of overall circulatory mortality (top) and overdose mortality (bottom). Although increases in PDI circulatory mortality occur for males and females, sharper escalations occur in male PDI deaths. By 2020, the rates are 0.291 for females and 0.861 for males. While rates for White non-Hispanic persons and Hispanic persons PDI mortality mirror each other over time (Appendix Figure 3), both remain below Black persons PDI mortality across the entire period. This contrasts with the concentration of overdose mortality among White persons until 2020. Appendix Figure 4 shows crude rates for PDI circulatory mortality by race to include Native American/Alaska Native persons and Asian persons, for whom small numbers prevent age-adjusted calculations. While caution is urged due to small numbers, crude rates for Native American/Alaska Native persons nonetheless increase dramatically. Similar to patterning for overdose deaths, PDI circulatory mortality is heavily concentrated in midlife (Appendix Figure 5), particularly between ages 45–54 and more recently those 55–64. PDI circulatory mortality has not elevated notably in younger adults ages 25–34, in contrast with overdose deaths.

Finally, age-adjusted rates by state (including DC; Appendix Table 3) were ranked by PDI circulatory mortality rates. Nevada led the nation by a considerable margin, with a rate of 2.164 per 100,000. The next state, Oklahoma, has a rate of 0.764. The comparison to overall circulatory death rates is informative. Some states ranking well on overall circulatory deaths have higher rates of PDI circulatory deaths, such as New Mexico (44 overall vs. 3 for PDI), Arizona (46 vs. 4), and Alaska (48 vs. 7). At the other end, some states ranking poorly on overall circulatory deaths have relatively low PDI deaths, such as Mississippi (1 vs. 47) and Alabama (3 vs. 36). Similarly, some states suffering high overdose rates have comparatively low rates of PDI circulatory deaths, such as Connecticut (17 for overdoses vs. 49 for PDI), Massachusetts (8 vs. 35), and Maryland (6 vs. 34). By contrast, states such as California (37 vs. 5) and Texas (43 vs. 10) have experienced low overdose mortality rates but high rates of PDI circulatory deaths. Supplemental analyses revealed relatively little change in how states ranked over time.

Discussion

Analyses identified trends in PDI cardiovascular mortality within the U.S. over a two-decade period. Although cardiovascular mortality declined overall,⁶ rates of PDI circulatory deaths increased, indicating that escalating problems associated with drugs may offset gains made towards reducing cardiovascular mortality. It is notable that declines in cardiovascular mortality have decelerated during the past decade,⁷ a period when PDI deaths increased in a statistically significant manner. Although this deceleration is likely multifactorial, further research is needed to identify the extent to which PDI mortality affects gains made towards improving cardiovascular health.

The psychotropic drugs most commonly implicated in circulatory deaths are those demonstrating considerable increases in overdose mortality. Stimulant drugs such as cocaine and methamphetamine appear particularly concerning, and these substances may stress cardiovascular system functions in both acute and chronic ways.^16–20 For example, cocaine may trigger myocardial infarction by directly affecting myocardial tissue and indirectly via vascular thrombosis, while chronic methamphetamine use can contribute to aortic dissection via hypertensive effects.^16,19 Notably, psychostimulant-implicated circulatory deaths increased considerably during the past decade, aligning with escalations in overdose mortality.³³ Clinicians should take particular note of methamphetamine and other stimulant use among patients, particularly as consumption of and related overdoses increased in recent years, a trend overshadowed by opioid-related mortality. Although fentanyl is the dominant opioid in overdose deaths in recent years, it emerged as a predominant opioid for PDI circulatory deaths only in 2020. Yet notably, sensitivity analyses on multiple drug classes versus single drug classes indicate that cocaine and methamphetamine use without other drugs heavily affected PDI circulatory deaths. Finally, despite increasing normalization and consumption across the population, particularly midlife adults, cannabis contributes minimally towards cardiovascular mortality.

Increases in PDI circulatory mortality were sharper among males than females, cohering with higher rates of generalized cardiovascular mortality among males as well as patterns of drug use and dependence, although gender gaps in substance use appear to be narrowing.^25,34 Should these patterns continue, they may contribute towards an increasing gap between men and women in age-adjusted cardiovascular mortality. Additionally, distinctions in PDI circulatory mortality follows familiar patterns whereby Black Americans experience greater adverse outcomes with cardiovascular health,^26,28 despite contradicting the drug mortality trend of greater deaths among White Americans until the most recent years. PDI deaths may be a point of intervention to consider within broader efforts to reduce racial disparities in cardiovascular mortality. Also, midlife is a key period for PDI circulatory mortality. Although drug problems are popularly represented as occurring during adolescence and young adulthood, a greater need to account for substance use within midlife exists given escalations in PDI circulatory deaths among these groups. Lastly, patterns by state indicate that trends in either overall circulatory deaths or overdose deaths alone should not guide clinical practice and health resource allocations. Some states that perform relatively well on either cardiovascular or overdose mortality nonetheless have higher rates of PDI circulatory mortality (and vice versa). Addressing PDI mortality may require intervening on a complex interaction of cardiovascular health, substance use patterns, and other factors.

This trend poses challenges for cardiovascular specialists and other clinicians who have contributed to improvements in cardiovascular health. Several considerations may help disrupt these trends and contribute to restoring the trajectory of decline in cardiovascular mortality. First, clinicians can emphasize to patients that drug use has ramifications for mortality beyond overdose; conversations about substance use with patients may provide avenues for intervention on cardiovascular mortality. Efforts to reduce stigma within clinical encounters may contribute towards such dialogue. Furthermore, reducing stigma within these conversations may especially benefit midlife adults and Black adults, who disproportionately experience PDI cardiovascular mortality. Additionally, efforts to ensure that people who use drugs receive care for cardiovascular conditions may ameliorate some excess mortality. Working with harm reduction agencies and organizations engaged in public health efforts with people who use drugs may facilitate such linkages to cardiovascular care. Reaching people with underlying hypertensive conditions is especially important as the proportion of PDI deaths for this specific condition is proportionally higher than for overall cardiovascular mortality. While considerable attention is devoted to overdose risks, further highlighting the role of acute and chronic substance use on cardiovascular health may enable prevention efforts to limit premature mortality.

Limitations

Despite many strengths, limitations must be considered. The analyses focus on the U.S., where drug mortality has escalated greatly, but more research is needed to identify global patterns. Also, scholars have identified that mortality data quality for drug-related causes of death vary across states, which may affect some estimates, including how the underlying cause of death gets coded between sites.³⁵ Further, particularly for specific drugs implicated, variation over time exists in the percentage of deaths identifying specific substances; e.g. for overdoses, this ranged from 75%–79% from 1999 through 2013 but increased to 94% by 2020.¹ Although unconfirmable, the results presented here may be underestimates, since drugs as a contributing but non-underlying cause of death may be overlooked by medical examiners when the underlying cause of death is not drug poisoning or otherwise suspicious. As such, these results indicate that medical examiners should heighten attention to the presence of drugs with cardiovascular deaths.

Conclusions

The results highlight 21^st century trends in PDI cardiovascular mortality within the U.S. Increases in PDI circulatory mortality are set against a backdrop of overall declines in cardiovascular deaths. Shifts in PDI circulatory mortality have not been distributed evenly across the population. Efforts to reverse these trends could help reinvigorate decades long trends of declining cardiovascular mortality in the U.S., which has stagnated in recent years.

Abbreviations:

PDI

psychotropic-drug-implicated

DOT

diseases of the

ICD-10

International Classification of Diseases, Tenth Revision

CDC

Centers for Disease Control and Prevention

WONDER

Wide-ranging ONline Data for Epidemiologic Research

IRB

Institutional Review Board