Releasing sterile male mosquitoes cut dengue risk by about 70% in a Singapore trial

A cluster-randomised trial in Singapore reports that repeated release of male Aedes aegypti mosquitoes carrying Wolbachia bacteria suppressed wild mosquito populations and reduced the risk of symptomatic dengue by roughly 70% [s1]. The results were published in the New England Journal of Medicine on 11 February.

Dengue vector control has been short of interventions with trial-quality evidence behind them. This one has a randomised design, a large population, and a 24-month follow-up.

The biology being exploited

Wild-type female A. aegypti that mate with males infected with the wAlbB strain of Wolbachia pipientis produce nonviable offspring, through a phenomenon called cytoplasmic incompatibility [s1]. Repeated releases of infected males can therefore suppress a wild-type population without insecticide [s1].

The distinction from other Wolbachia programmes matters. Some approaches release both sexes to establish Wolbachia in the local mosquito population permanently, which reduces the mosquitoes' ability to transmit virus. This trial used a suppression strategy: males only, functioning as a biological sterile-insect technique, requiring continued releases to hold the population down.

The trial

Fifteen geographic population clusters were divided into two groups: eight received deployments of male Wolbachia-infected mosquitoes and seven received none [s1]. A total of 393,236 residents lived in the intervention clusters and 331,192 in the control clusters [s1].

The design used test-negative controls. The primary endpoint was diagnosis of symptomatic dengue virus infection of any severity caused by any serotype, measured as the odds ratio for the distribution of Wolbachia exposure among laboratory-confirmed reported dengue cases compared with test-negative controls [s1].

A test-negative design compares people who sought testing and were positive against people who sought testing and were negative. It is a common approach for vaccine effectiveness because it partly controls for differences in health-seeking behaviour — the comparison group has already demonstrated willingness to be tested.

What happened to the mosquitoes

Adult wild-type A. aegypti populations were suppressed across the intervention clusters [s1].

Baseline average abundance — the number of adult female mosquitoes trapped divided by the number of traps — was 0.18 in the intervention clusters and 0.19 in the controls [s1]. From three months after the intervention began until the end of the 24-month trial period, average abundance was 0.041 in intervention clusters and 0.277 in controls [s1].

That is a fall of roughly three-quarters from baseline in the intervention arm, while the control arm rose. The divergence in the control clusters is a reminder that the counterfactual is not static.

What happened to dengue

In the intention-to-treat analysis at six months or more, 354 of 5,722 tests in intervention clusters were dengue-positive (6%), compared with 1,519 of 7,080 tests in control clusters (21%) [s1].

Protective efficacy, calculated as one minus the odds ratio times 100, ranged from 71% to 72% with three to twelve months or more of Wolbachia mosquito exposure, corresponding to odds ratios of 0.28 to 0.29 [s1].

The authors conclude that release of sterile Wolbachia-infected male A. aegypti reduced vector populations and the risk of dengue infection in Singapore [s1]. The trial was funded by the Singapore Ministry of Finance and others and is registered as NCT05505682 [s1].

What limits the reading

Three things are worth holding.

Setting. Singapore is a tropical city-state with dense, largely high-rise housing, strong existing vector surveillance, and a well-resourced national programme. Those conditions affect both how efficiently releases can be delivered and how completely a suppression effect can be maintained against reinvasion. Whether comparable efficacy transfers to sprawling or peri-urban settings with weaker infrastructure is not answered here.

Suppression is not permanent. Because this strategy removes rather than replaces the mosquito population, the effect depends on continued production and release of infected males. The trial ran 24 months; it does not tell us what the abundance curve looks like when releases stop.

Cluster-randomised, not blinded. Residents and health services knew, in practice, which areas received deployments. The test-negative design mitigates some of the resulting bias in health-seeking, but not all of it.

The percentage figures — 6% versus 21% of tests positive — should also not be read as population incidence. They are proportions among people who presented for dengue testing, which is the correct comparison for the study design but a different quantity from the risk faced by an average resident.

Why this matters beyond Singapore

Dengue's global range has been expanding, and the intervention toolbox is thin: source reduction, insecticides against which resistance is widespread, and vaccines with complex indication restrictions. A vector-control method with cluster-randomised evidence of roughly 70% protective efficacy against symptomatic disease is a substantive addition to that list, whatever the transfer questions.

It is also a method with real operational demands. Producing and sex-sorting large volumes of male mosquitoes, then releasing them repeatedly across a landscape for years, is an industrial programme, not a campaign.

What to watch

Whether equivalent trials in different settings — lower-density housing, weaker surveillance, higher baseline transmission — reproduce the efficacy estimate.

What happens to abundance and to dengue incidence in the intervention clusters over a longer horizon, particularly if release intensity varies.

And whether cost per person protected is published. For an intervention that must be sustained indefinitely, that number determines whether the efficacy result is usable anywhere it has not already been tried.

Sources

• [s1] Lim JT, Chong CS, Chang CC, et al. (Project Wolbachia–Singapore Consortium), "Dengue Suppression by Male Wolbachia-Infected Mosquitoes," The New England Journal of Medicine, published 11 February 2026. https://doi.org/10.1056/NEJMoa2503304