The Fourth Wave Breaks — and Tells Us Exactly Who Was Left Behind

Maria Feather, a peer recovery specialist at the Native American Connections crisis center in Phoenix, heard about the new overdose data on the morning of June 4 — the way most people in her field hear about research: someone shared it in a group text. She pulled up the study on her phone, read the headline figure, and then scrolled until she found the number she already knew was there. American Indian and Alaska Native people: 50.8 overdose deaths per 100,000. A decline of only 20.1 percent, the lowest rate of improvement among any racial group tracked.

“That’s us,” she told a colleague over the phone. “The wave broke everywhere else, and it’s still coming for us.”

The study she was reading — “Charting the Decline of the Fourth Wave,” published June 2 in the journal Addiction by researchers at UC San Diego — is the most granular accounting yet of what happened to U.S. drug overdose deaths between 2023 and 2024. The headline is genuinely historic: the national overdose death rate dropped 24.4 percent in a single year, falling to 23.7 per 100,000 people. That’s the first time every wave of the overdose crisis — from prescription opioids to heroin to illicit fentanyl to fentanyl-stimulant combinations — declined simultaneously.

It is a meaningful number. Anyone who has spent time in this field knows that headline numbers almost never move this fast. This one did. The people responsible — harm reduction workers who distributed naloxone, addiction medicine physicians who prescribed buprenorphine without requiring counseling attachments, syringe exchange programs that kept people alive long enough to want treatment — deserve to hear that the work moved the needle.

But Maria Feather’s instinct to scroll past the headline and find the granular data is the correct instinct for anyone trying to understand what the decline actually means. The numbers beneath the 24.4 percent tell a story the headline can’t carry.

The Decline Is Real, and Uneven in Ways That Matter

Lead author Joseph R. Friedman and colleagues at UC San Diego analyzed CDC mortality data across race, ethnicity, substance class, and geography — producing what may be the most detailed breakdown of the 2023-2024 shift in print.

The disparity that stopped Maria Feather cold is the one that should stop policymakers cold as well. American Indian and Alaska Native individuals had the highest absolute overdose death rate of any group — 50.8 per 100,000 — and experienced only a 20.1 percent decline, the weakest improvement in the dataset. That rate is 2.15 times the national average. The gap between AIAN communities and the national trend is not an artifact of small sample size. It reflects decades of systematic underinvestment in tribal behavioral health infrastructure, geographic barriers to treatment access, and federal trust responsibility obligations that have gone chronically underfunded.

Black and African American individuals, meanwhile, experienced the largest percentage decline of any group: 29.3 percent, bringing their rate to 35.99 per 100,000. That is the fastest single-year improvement in the dataset. And it still leaves Black Americans dying at 1.51 times the national average.

Non-Hispanic white Americans saw a 23.9 percent decline — roughly the national average. The overdose crisis began, in its modern form, in predominantly white communities in Appalachia and New England. Those communities built political will and got early investment in naloxone distribution, treatment expansion, and harm reduction infrastructure. The data suggests those investments paid off — and that they did not pay off equally for communities that received less of them.

The right read here is not that the 2024 decline is bad news dressed up as good news. It isn’t. A 24.4 percent single-year drop in overdose death rates is extraordinary. The right read is that the crisis response was never uniformly deployed, and the data is now showing us exactly where the gaps were.

The right read is that the crisis response was never uniformly deployed, and the data is now showing us exactly where the gaps were.

Friedman and colleagues did not speculate in the paper about the precise mechanisms. But the pattern in the data — largest declines where naloxone access, syringe services, and buprenorphine prescribing expanded earliest; slowest declines where those programs remained limited or were actively opposed — is not subtle.

What’s Falling, and What Isn’t

The 2024 decline was driven almost entirely by falling illicit fentanyl deaths. Deaths involving fentanyl without stimulants dropped from 31,193 in 2023 to 19,673 in 2024 — a 36.9 percent decline. Deaths involving fentanyl combined with stimulants (typically methamphetamine or cocaine) fell from 41,583 to 28,062, a 32.5 percent decline. The fourth wave — the fentanyl-stimulant combination that drove overdose rates to their peak in 2022 and 2023 — broke in 2024. That’s what “first recorded drop in all four waves simultaneously” means.

What did not fall are the trends that have been quietly building throughout the fentanyl era and now represent the emerging shape of the crisis.

Deaths involving stimulants without fentanyl — primarily methamphetamine and cocaine used on their own — continued to rise. As fentanyl-involved deaths fall and stimulant-involved deaths climb, stimulants are becoming an ever-larger share of the overdose picture. Friedman’s team noted in the paper that “stimulants may soon surpass opioids” as the primary substance class driving addiction-related mortality in the United States. That is not a warning about the distant future. That is a description of a near-term trend visible in 2024 data.

The other number still moving in the wrong direction: xylazine. The veterinary sedative found adulterating the illicit drug supply — concentrated heavily in the Philadelphia corridor but spreading nationally — continued claiming more lives. Unlike opioids, xylazine does not respond to naloxone, which is why the wound-care crisis it creates is so acute. There are no approved medications for xylazine dependence. The people encountering it have almost no pharmacological safety net.

Together, these two trajectories — stimulants rising, xylazine rising — describe the shape of the next crisis. The infrastructure being built to address fentanyl will not automatically transfer. Naloxone doesn’t reverse a methamphetamine overdose. Buprenorphine wasn’t designed for stimulant use disorder.

The Funding Paradox

The decline in overdose deaths happened against a backdrop of serious federal funding uncertainty. The year 2026 began with SAMHSA sending termination notices to roughly 2,700 grantees — close to $2 billion in behavioral health grants — before reversing course within 24 hours under intense pressure from providers, advocates, and members of Congress. The whiplash was damaging regardless of the reversal. Programs laid off staff. Trainings were cancelled. Services shut down. The check eventually came through, but the 24 hours in between cost real people real things.

The underlying budget conversation did not resolve. The administration’s proposed “Behavioral Health Innovation Block Grant” — which would consolidate three existing SAMHSA grant streams into a single fund — would reduce total behavioral health block grant funding by $465 million compared to current spending. And it would eliminate the categorical mandates that currently push money toward the programs states don’t always fund voluntarily: workforce development, syringe service programs, harm reduction outreach, rural treatment expansion.

The timing is almost precisely wrong. A 24.4 percent decline in overdose deaths is strong evidence that the infrastructure built by those grants does something. The data says: keep funding it. The proposed budget says: cut it by $465 million and let states decide.

States have historically not decided in favor of harm reduction when given the choice. Arizona, to its credit, is an exception in some respects — AHCCCS coverage is broad, and the state’s $1.215 billion in opioid settlement funds creates a 15-year runway for treatment investment. But the settlement money isn’t infinite, and the categorical protections in the block grant are what force spending toward the services that research consistently shows reduce deaths: naloxone access, medication treatment, and harm reduction outreach.

States have historically not decided in favor of harm reduction when given the choice.

Remove those mandates, and the settlement dollars will drift toward politically safer investments — education campaigns, drug courts, abstinence-only residential programs — rather than the interventions the 2024 data shows were working.

What the Data Is Asking For

The Friedman et al. paper was published in an academic journal two weeks ago. It has been cited in a handful of addiction medicine and public health newsletters, shared by researchers on social media, and largely ignored by mainstream news. That is the typical lifecycle of a study like this, and it is a structural problem in how the country processes evidence about a crisis it has been living with for thirty years.

The finding that the overdose death rate fell 24.4 percent deserves a policy response proportional to its significance: sustained investment in what worked, targeted expansion into the communities where the decline was slowest, and an honest accounting of the next phase of the crisis before stimulant and xylazine deaths climb to where fentanyl deaths were two years ago.

What it should not produce is complacency. The people celebrating the 2024 numbers and the people worrying about 2026 are both right. Maria Feather in Phoenix can hold both of those things in her head at once — because she has to.

“I’ve been doing this work for eleven years,” she said. “Every time the numbers get better for someone, they get worse for someone else who doesn’t have a name yet in the data. I want the celebration and the reckoning to be the same sentence.”

The data, now published in Addiction, gives us the sentence. The question is whether the people who write budgets and set funding priorities will read it.

---

The Friedman et al. paper — “Charting the Decline of the Fourth Wave: US Overdose Deaths by Race, Ethnicity, and Substance Involvement” — was published June 2, 2026 in the journal Addiction. Full text available via Wiley and PMC. The CDC NCHS data brief underlying the analysis is at cdc.gov. Find treatment options in Arizona at /find-help.