The Autopsy / Case File · Overdose & Substance
The Gap That Reopened
For a decade, dying of a drug overdose in America was, county for county, more likely if you lived somewhere rural. Then it wasn't. Then, more recently, the specific drugs on each side of that line started swapping places again — and the federal data tracking all of it runs a year or two behind the supply chain it's trying to measure.
CDC's National Center for Health Statistics has tracked age-adjusted overdose death rates by urban-rural county classification since 1999. The rates ran roughly even from 2004 through 2006, then rural counties pulled ahead and stayed ahead for nine straight years, from 2007 through 2015, driven overwhelmingly by prescription opioids. By 2016 the lines crossed back: urban rates hit 20.0 per 100,000 against rural's 18.7, and the following year the gap widened further, 22.0 versus 20.0, as heroin, cocaine, and fentanyl — drugs that have stayed persistently more urban — took over as the crisis's primary engines. The most recent county-level data, from 2020, shows the urban lead holding at 28.6 versus 26.2 nationally, even as psychostimulants and diverted prescription pills continue outpacing fentanyl specifically in rural counties, and even as newer qualitative research from 2024 finds fentanyl itself now turning up in rural drug supplies that didn't carry it before.
Join the discussion →
Overdose mortality gets described nationally, in one curve, more often than it gets described geographically — a single line climbing from 14.7 deaths per 100,000 in 2014 to a peak of 32.6 in 2022 before declining through 2024. That single line hides a fact CDC's own county-level surveillance has documented for a quarter century: the gap between rural and urban overdose death rates hasn't just narrowed or widened once. It has flipped, more than once, and the specific drug behind each flip has been different every time. Understanding the crisis's geography means tracking three separate, overlapping stories — who was dying where, of what, and how long it took the data to notice the pattern had already changed again.
Three Periods, One Data Brief
NCHS Data Brief No. 345, covering 1999 through 2017, lays out the sequence in age-adjusted rates per 100,000. In 1999, urban counties had the higher rate, 6.4 against rural's 4.0, and that urban lead held through 2003 — with rural rates actually climbing faster in percentage terms even while trailing in absolute numbers, growing roughly 20 percent a year against urban's 9 percent. From 2004 through 2006 the two lines converged and ran together. Then, starting in 2007, rural counties took the lead outright and didn't give it back for nine years, growing 13 percent annually from 2003 to 2006 and continuing to outpace urban counties' more modest climb through 2014. By 2016, the pattern reversed again: urban counties posted 20.0 deaths per 100,000 against rural's 18.7. In 2017 the gap widened further — 22.0 versus 20.0 — with urban rates accelerating at roughly 17 percent a year against rural's 9 percent.
What Actually Drove Rural's Decade in the Lead
The nine-year rural lead wasn't a single drug crisis; it was a specific one. Through the mid-2000s and into the 2010s, deaths involving natural and semisynthetic opioids — the pharmaceutical category that covers oxycodone, hydrocodone, and morphine, the pills a doctor actually prescribes rather than a street-manufactured substitute — ran higher in rural counties than urban ones for most of that period, only converging by 2018 and 2019 as rural rates for that specific drug class declined. That pattern lines up with a well-documented feature of the early prescription-opioid era: per-capita opioid prescribing ran higher in many rural and small-town counties, especially ones marked by economic decline, manual-labor industries with high rates of occupational injury, and a disability-heavy population — the same demographic and economic profile researchers have repeatedly tied to the geography of prescription-opioid mortality specifically, as distinct from the geography of the broader overdose crisis that followed it.
Heroin and cocaine told a nearly opposite geographic story across the same years. NCHS's data brief is explicit on this point: rates of drug overdose deaths involving heroin or cocaine ran consistently higher in urban than in rural counties for the entire period studied, without the reversal seen in the topline numbers. By 2017, heroin's urban rate was 5.2 per 100,000 against rural's 2.9; cocaine's urban rate was 4.6 against rural's 2.4. Synthetic opioids other than methadone — the category dominated by illicitly manufactured fentanyl — were already running higher in urban counties too by 2017, at 9.3 versus 7.0. The topline rural lead through 2015 wasn't proof that rural America had a bigger overdose problem in some general sense; it was proof that rural America had a much bigger prescription-opioid problem specifically, one large enough to outweigh urban counties' larger heroin, cocaine, and early fentanyl problem in the national tally — until the prescription-pill numbers came down and the fentanyl numbers kept climbing.
Rural counties didn't have the bigger overdose crisis. They had the bigger prescription-opioid crisis — and for nine years, that was enough to win the topline number.
2020: The Gap Widens, and a New Split Appears
The most recent urban-rural comparison NCHS has published, covering 2020, shows the reversal holding and, in some respects, deepening. The national age-adjusted rate was 28.6 per 100,000 in urban counties against 26.2 in rural ones. Broken out by drug class, the urban lead was largest for synthetic opioids — 18.3 versus 14.3, a 28 percent gap — and for cocaine, where the urban rate of 6.4 was more than double rural's 3.0. Heroin remained a smaller but consistent urban-favoring category, 4.2 versus 3.2. But two drug categories ran the other direction entirely: natural and semisynthetic opioids were still slightly higher in rural counties, 4.5 versus 4.0, and psychostimulants — the category covering methamphetamine — were substantially higher in rural counties, 9.4 versus 7.2, a 31 percent gap in rural's favor. The geographic split by sex ran differently too: male overdose rates were markedly higher in urban counties (40.4 versus 34.1), while female rates were actually slightly higher in rural ones (17.9 versus 17.0). By race, the widest urban-rural gap of all appeared among non-Hispanic Black Americans, whose urban overdose rate of 37.4 per 100,000 was nearly double the rural rate of 18.9 — while non-Hispanic American Indian and Alaska Native Americans had the highest overdose rate of any group measured in either setting, 44.3 urban and 39.8 rural.
State-level variation sat underneath all of it: urban rates exceeded rural rates in 23 states in 2020, rural rates exceeded urban in 8, and the two were statistically similar in the remaining 16 — meaning the national topline number, in either direction, has always been an average masking sharply different local realities depending on which specific drug supply had reached which specific county first.
The Supply Chain Doesn't Wait for the Data Brief
Every number above describes a supply chain as it existed one to six years before publication — NCHS's most recent urban-rural-specific comparison covers 2020 data, released in 2022. Newer qualitative research suggests the geographic drug-type split that data captured may already be eroding. A 2024 multi-site study published in Harm Reduction Journal, part of the NIH-funded Rural Opioid Initiative, interviewed people who use drugs across rural regions of the country and found a consistent, cross-site theme: participants, regardless of their location or which drug they primarily used, described a growing awareness that their local drug supply — including supplies that historically hadn't carried it — was now contaminated with fentanyl. That finding doesn't show up in any topline mortality rate yet, because mortality data lags behind the drug supply by design; a death has to happen, get certified, get coded, and get aggregated before it appears in a data brief. But it is exactly the kind of early signal that, historically, has preceded the next reversal in this dataset — the same way rising rural prescription-opioid mortality in the early 2000s preceded the rural lead that showed up in the topline numbers years later, and the same way urban fentanyl's early growth preceded the urban lead that reasserted itself by 2016.
None of this means the rural-urban overdose gap is a wash, or that geography stopped mattering once the topline lead changed hands. It means the geography itself has never been static, and treating a single year's national urban-versus-rural comparison as a fixed fact about where the crisis lives is a way of being surprised, again, the next time the underlying drug supply moves somewhere the data hasn't caught up to yet. Rural counties spent nine years in the lead on the strength of one drug class. Urban counties have now held the lead for most of a decade on the strength of several others. The four-year-old lag in the country's best available county-level data is the space where the next version of this story is very likely already happening, uncounted.
And Then You're Dead
Two people can die of the same drug in counties three hundred miles apart, and the dataset that's supposed to explain the difference between them is already describing a supply chain that no longer exists in quite that shape. The gap between rural and urban overdose deaths has closed, reopened, and swapped which drug sits on which side of it three separate times since 1999 — and there is no reason, on this evidence, to expect it has finished moving.
That's it. That's the whole thing.
Sources
- Age-adjusted drug overdose death rates by urban-rural county classification, 1999–2017, showing rates similar from 2004 through 2006, higher in rural counties from 2007 through 2015, and higher in urban counties again by 2016 (20.0 urban vs. 18.7 rural) and 2017 (22.0 vs. 20.0), with drug-type breakdowns for heroin, cocaine, synthetic opioids, and natural/semisynthetic opioids — Spencer MR, Miniño AM, Warner M, "Drug Overdose Deaths in the United States, 1999–2017," NCHS Data Brief No. 345, CDC National Center for Health Statistics, December 2019
- Age-adjusted drug overdose death rates by urban-rural county classification, 2020, showing an overall rate of 28.6 per 100,000 in urban counties vs. 26.2 in rural counties, breakdowns by sex, race/ethnicity, drug type (synthetic opioids 18.3 vs. 14.3; cocaine 6.4 vs. 3.0; heroin 4.2 vs. 3.2; natural/semisynthetic opioids 4.0 vs. 4.5; psychostimulants 7.2 vs. 9.4), and state-level distribution (23 states with higher urban rates, 8 with higher rural rates, 16 similar) — Garnett MF, Miniño AM, "Urban-Rural Differences in Drug Overdose Death Rates, 2020," NCHS Data Brief No. 440, CDC National Center for Health Statistics, July 2022
- National overdose death rate trend, 14.7 per 100,000 in 2014 rising to a peak of 32.6 in 2022 before declining through 2024, and illicitly manufactured fentanyl's role in an estimated 70 percent of 2023 overdose deaths — CDC National Center for Health Statistics, National Vital Statistics System, Drug Overdose Deaths data page, 2025
- Multi-site qualitative findings from the NIH-funded Rural Opioid Initiative documenting a consistent theme among rural people who use drugs: growing awareness and concern of fentanyl contamination spreading into drug supplies that historically had not carried it, regardless of participants' location or primary drug of use — "Strategies used to reduce harms associated with fentanyl exposure among rural people who use drugs: multi-site qualitative findings from the rural opioid initiative," Harm Reduction Journal, 2024
- Photo: "Crawfordville Pharmacy," Crawfordville, Florida, The Bushranger, CC BY-SA 4.0, via Wikimedia Commons
Comments
Loading…