As theories continue to emerge for the dramatic turnaround in overdose deaths, a new study indicates that it’s largely unrelated to the decreased proportion of fentanyl in law enforcement drug seizures—even though the previously increasing prevalence of overdose mortality and of fentanyl in seized drugs appeared to be more closely related.
The abrupt change in trajectory of overdose mortality was first identified in September 2024, and shows that after rising nearly every year since at least 1999, deaths peaked around August 2023. Since then they’ve dropped continuously, though not quite as sharply as they were initially.
In a study published January 16 in JAMA Health Forum, researchers analyzed data from all 50 states and the District of Columbia between mid-2023 and late 2024, the period when overdose deaths peaked and then started to fall. Overdose mortality data were sourced from the Centers for Disease Control and Prevention Wide-Ranging Online Data for Epidemiologic Research (better known as CDC WONDER) database, while monthly reports of law enforcement seizures were sourced from the National Forensic Laboratory Information System.
Out of the total law enforcement seizures of fentanyl, carfentanil, cocaine, heroin, methamphetamine or xylazine during that time period, the proportion of fentanyl seizures decreased from 28.8 percent to 23.2 percent. The study suggests that less than 10 percent of the decline in fatal overdose is associated with the decreasing proportion of fentanyl reports.
“The association between the two trends was significant but modest compared with the overall decline in deaths and appeared to decline in magnitude with time,” the authors wrote. “These findings suggest that other factors are likely contributing to the decline in mortality.”
For several years leading up to the sudden turnaround in overdose mortality rates, the rising number of deaths was accompanied by an increasing proportion of fentanyl in law enforcement seizures. The researchers noted that “both indicators rose in parallel” and followed a “well-documented geographic pattern.” The same does not appear to be true when deaths are trending downward, at least not to the same degree.
The 5.6-percentage point decline in proportion of fentanyl reports coincided with a 5.4-percentage point increase for cocaine reports.
Overdose deaths increased sharply between 2018 and 2022, from approximately 68,000 people to approximately 110,000. The researchers found that during that time the proportion of fentanyl reports in overall drug seizures rose 17 percentage points. But by September 2024 the annual death rate had dropped to approximately 88,000 people, while the drop in fentanyl seizures was not nearly as drastic—only 5.6 percentage points.
The authors suggest a number of factors that have likely contributed more substantially to the decrease in overdose deaths including other shifts in the drug supply, shifts in patterns of drug use and wider uptake of overdose prevention efforts such as naloxone distribution and access to medications for opioid use disorder.
Though some have linked the reduction in overdose deaths to a reduction in the fentanyl supply, this has often been mischaracterized or conflated with a decrease in fentanyl purity—the amount of fentanyl contained in drug samples sold as fentanyl. Fentanyl is too potent to sell on its own, because the amount of drug material would simply be too tiny, so it’s bulked out with a variety of nonpsychoactive substances. Other higher-risk substances are also sometimes mixed in, either intentionally or unintentionally.
The space left by fentanyl seems to have been filled by seizures of cocaine, with the authors noting that the 5.6-percentage point decline in proportion of fentanyl reports coincided with a 5.4-percentage point increase for cocaine reports. However, the authors also caution against viewing law enforcement seizure data as representative of the actual drug supply.
“While widely used, this measure assumes that seizures represented a sample of the illicit drug market; in reality, enforcement patterns and seizure targeting vary across jurisdictions and over time. The proxy indicator also does not capture weight, purity, potency, form or composition (carfentanil vs fentanyl with xylazine), and it excludes prescribed opioids. If any of these factors change over time within states, it could have introduced bias in our results.”
Image via United States Customs and Border Protection