In rural China, HIV keeps spreading among older men along transport routes

Treatment-as-prevention was supposed to close the door on HIV transmission: put enough people on antiretroviral therapy, suppress their virus, and the epidemic starves. A study published on 29 May shows why that logic can still leave a stubborn tail of transmission — and uses molecular data to name exactly who and where it is [s1]. Analysing HIV in rural southwest China, where outbreaks persist among the elderly despite "Treat All" policies, researchers triangulated genetic networks, phylodynamic and spatial analysis, and behavioural surveys across multi-source data from 1999 to 2021 (N = 5,094) [s1].

Treatment worked, but not enough

Antiretroviral therapy did what it is meant to do, up to a point: it reduced community-level transmission risk by 21.0% [s1]. But a 21% reduction is not the near-elimination that full, sustained viral suppression can achieve — and the study set out to explain the residual.

The answer, in the authors' framing, is a small set of "molecular super-spreaders." Across four epidemic phases they identified 107 such individuals, predominantly males aged 50 to 69 [s1]. These were not a random slice of people living with HIV. They showed poor treatment retention, with dropout rates reaching 28.2%, and by the fourth phase 38.5% had unsuppressed or missing viral loads [s1]. In other words, the people at the centre of continued transmission were disproportionately those who had fallen out of, or never been reliably held by, the treatment system.

Geography as a driver

The most striking finding is spatial. Fully 94.4% of the super-spreaders lived within 2 km of transportation arteries, against a population baseline of 77.9% — a concentration too large to be an artefact of where people happen to live [s1]. Townships within that roadside buffer recorded significantly higher cumulative HIV cases independently of population size, with an adjusted incidence rate ratio of 1.43 (95% CI 1.14-1.78) [s1].

The authors describe these roadside settlements as infrastructure-anchored "risk hubs" that sustain viral dispersal through what they call a "double-proximity" mechanism [s1]. The behavioural surveys fill in the picture: older men frequent nearby periodic markets along transit corridors for commercial sex, while female sex workers move in a linear pattern across townships, timing their movements to the rhythm of those markets [s1]. Transport routes, periodic markets, and an ageing male clientele intersect to keep transmission alive in places a population-wide treatment programme was not built to reach.

Why it matters beyond China

The specifics are local, but the lesson is not. Programmes worldwide have leaned on the promise that scaling up treatment will drive incidence toward zero. This study is a concrete demonstration of the limit of that promise when a definable minority — here, older men with poor treatment retention, clustered along transport corridors — falls outside the system's effective reach. A 21.0% reduction in community transmission is real progress, but the residual epidemic it leaves behind is structured, not diffuse, and structure is something interventions can target [s1].

That is the authors' central argument. Because the drivers are identifiable by who (older men with treatment gaps), where (within 2 km of transport arteries), and how (commercial sex tied to periodic markets), the epidemic is amenable to what they call node-level testing and intervention — bridging clinical treatment and public-health prevention rather than treating them as separate tracks [s1]. The methodology, they suggest, is a way to mitigate aging-driven epidemics in resource-limited regions more broadly [s1].

The ageing dimension

The age profile deserves emphasis. HIV programmes are often designed around younger key populations, but here the persistent transmission sits among men aged 50 to 69 [s1]. Older men living with HIV are easy to overlook in prevention messaging and in services oriented toward the young, yet in this setting they are central both as a group with worse treatment retention and as participants in the market-based commercial sex networks that sustain spread [s1]. An epidemic that has aged with its population needs services that have aged with it too.

The limits

This is an observational, retrospective reconstruction built on genetic networks and behavioural surveys, not a trial, and the "super-spreader" designation is inferred from molecular clustering rather than directly witnessed transmission. The 21.0% figure describes a modelled reduction in community-level risk, and the spatial associations, while adjusted, describe correlation in a specific rural region of one country. What the study offers is not a universal number but a template: a way to locate the residual engine of an epidemic precisely enough to act on it.

For countries chasing the last stretch toward HIV control, that precision is the point. The uncomfortable implication is that "Treat All" alone does not finish the job where a hard-to-reach minority keeps transmission going — and that finding them may depend as much on a map of roads and markets as on a clinic register [s1].

Sources

• Delineating the "who-where-how" of persistent HIV epidemics: a 23-year longitudinal genetic network, phylodynamic, and spatial analysis for precise intervention in rural China — Emerging Microbes & Infections, 2026-05-29