Association of Medical Stimulants With Mortality in the US From 2010 to 2017

The role of nonopioids in the drug overdose epidemic in the US is frequently overshadowed. From 2015 to 2016, mortality rates involving cocaine and psychostimulants were smaller than opioids, but were rising faster than opioids.¹ We examined individual substances listed on death certificates to better understand stimulant-associated mortality and assess rates over time.

We used the Drug Involved Mortality database from the National Center for Health Statistics,² which comprised drug-related terms mentioned on death certificates for every death in the US from 2010 to 2017. All involved drugs from the death certificate part I, part II, and box 43 were used. Demographic characteristics as listed on the death certificate, polysubstance deaths, and mortality rates were analyzed for 3 subgroups: (1) all stimulants, (2) illicit stimulants (eg, cocaine), and (3) medical stimulants (eg, methylphenidate) (subgroups are defined in the eTable in the Supplement). If a death involved multiple drugs, it was counted in each of the drug groups; this resulted in double counting between rate estimates for individual drugs. Annual rate ratios (ARRs) were modeled using a Poisson mixed model regression. The Colorado Multiple institutional review board confirmed this analysis of deidentified decedent data to be not human participants research, and it was considered exempt from informed consent. Statistical analysis was conducted by the Researched Abuse, Diversion and Addiction-Related Surveillance System using SAS, version 9.4 (SAS Institute), and statistical analysis was set at P < .05.

Of the 1 220 143 deaths with involved drugs listed on the death certificate, 130 560 (10.7% of all decedents) were found to have stimulants listed. Among stimulant-involved deaths, 93 689 decedents (71.8%) were men, the median (interquartile range) age was 45 (34-54) years, and 98 635 (75.5%) were White (Table). Of these, 120 803 certificates (92.5%) listed only illicit stimulants, 5544 (4.2%) listed only medical stimulants, and 3524 listed both types (2.7%). Among illicit stimulants, there were 77 013 deaths (61.9%) involving cocaine, 49 602 deaths (39.9%) involving methamphetamine, and 817 deaths (0.7%) involving 3,4 methylenedioxymethamphetamine. Among medical stimulants, there were 8240 deaths (90.9%) involving amphetamine, 295 deaths (0.3%) involving methylphenidate, and 615 deaths (0.7%) involving pseudoephedrine. Among all 3 stimulant groups, the proportion of deaths that also involved opioids was substantial; concomittant use of benzodiazepines and antidepressants was also not uncommon.

Stimulant mortality has risen rapidly since 2010 (Figure). The mortality rate involving all stimulants rose from 2.913 deaths per 100 000 population in 2010 to 9.690 in 2017. Mortality rates increased among all medical stimulants (ARR, 1.226; 95% CI, 1.202-1.250), amphetamine (ARR, 1.118; 95% CI, 1.082-1.155), cocaine (ARR, 1.234; 95% CI, 1.222-1.245), and methamphetamine (ARR, 1.278; 95% CI, 1.261-1.295).

Our study showed an increase in mortality involving stimulants and identifies the substances that drive this upward trend. For many stimulants, there was a doubling in the mortality rate approximately every 4 years, which corresponds to an ARR of 1.20. Stimulants are often used in combination with opioids and other drugs, as polysubstance use complicates treatment for use disorder. There are only 2 International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes for stimulants, and drugs listed by the medical certifier contain more information. Therefore, this study is more specific than studies that use ICD codes alone.^3,4 The limitations of the study include that amphetamine is in the metabolic pathway of other stimulants, possibly producing misclassification, and the change in accuracy over time of identifying involved substances is not known.

The rapid rise in stimulant-involved mortality is concerning given the lack of tools to combat increasing mortality. To our knowledge, there are currently no medications approved to treat stimulant use disorder nor a reversal agent for stimulant-induced overdose. Opioid-involved mortality in 2000 had a similar rate (approximately 3 deaths per 100 000 population) and rose more slowly than our results have shown for stimulant-involved mortality.⁵ A notable difference is that stimulant-involved mortality predominantly stems from illicit drugs. These results should be a call to address stimulant-associated use before mortality reaches epidemic levels.

Accepted for Publication: November 3, 2020.

Published Online: February 1, 2021. doi:10.1001/jamainternmed.2020.7850

Corresponding Author: Joshua C. Black, PhD, Rocky Mountain Poison & Drug Safety, A Division of Denver Health, 1391 N Speer Blvd, Ste 600, M/C 0180, Denver, CO 80204 (joshua.black@rmpds.org).

Author Contributions: Dr Black and Ms Bau had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Black, Iwanicki, Dart.

Acquisition, analysis, or interpretation of data: Black, Bau, Iwanicki.

Drafting of the manuscript: Black, Bau.

Critical revision of the manuscript for important intellectual content: Black, Iwanicki, Dart.

Statistical analysis: Black, Bau.

Obtained funding: Dart.

Administrative, technical, or material support: Black, Iwanicki, Dart.

Supervision: Iwanicki, Dart.

Conflict of Interest Disclosures: Dr Black reported that the Researched Abuse, Diversion and Addiction-Related Surveillance (RADARS) System received funding from Shire Plc for general surveillance of stimulant prescription drug products outside of the submitted work and the receipt of grants from the US Food and Drug Administration (FDA) (75F40120C00151) outside the submitted work.

Funding/Support: The RADARS System is supported by subscriptions from pharmaceutical manufacturers, government and nongovernment agencies for surveillance, and research and reporting services. The RADARS System is the property of The Denver Health and Hospital Authority (DHHA), a political subdivision of the State of Colorado. No subscriber participated in the conception, analysis, drafting, or review of this manuscript.

Role of the Funder/Sponsor: The DHHA funded the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The findings and conclusions in this paper are those of the authors and do not necessarily represent the views of the Research Data Center, National Center for Health Statistics (NCHS), or the US Centers for Disease Control and Prevention.

Additional Contributions: We thank Philip Pendergast, PhD, US Census Bureau, and Jaylan Richardson, MS, NCHS, for their assistance in creating the data set. No compensation was received.