Seven years after Niteroi released mosquitoes carrying a dengue-blocking bacterium, reported cases are 88.8 percent below the city's own decade-long baseline, and the bacterium is still there.
In 2024 Brazil recorded its worst dengue year on the books: 3,157 cases for every 100,000 people. In Niteroi, a city of about half a million across the bay from Rio de Janeiro, the figure was 374. Across Rio de Janeiro state as a whole it was 1,884. The difference in Niteroi is Wolbachia, a bacterium that lives naturally inside insects and, once it is inside a mosquito, sharply reduces that mosquito's ability to pass dengue on. Between 2017 and 2023 Niteroi released its own mosquitoes carrying it, zone by zone, until the bacterium had spread through the wild population. Seven years after the first release, it is still there.
How we know
Niteroi did this in stages. Between 2017 and 2019 it released mosquitoes carrying the wMel strain of the bacterium across four zones covering 371,000 residents, and between November 2022 and July 2023 it added a fifth zone of 111,000, becoming the first Brazilian city with citywide coverage. Researchers at the World Mosquito Program then compared what followed with the decade before, drawing on Brazil's national notification system, which logs every dengue case a doctor reports, from 2007 to September 2024. Their tool was an interrupted time series model, a method that establishes the trend running before an intervention and measures how far the later data departs from it.
In the ten years before the releases the city averaged 4,349 dengue cases a year, 913 for every 100,000 residents. In the five years after they finished it averaged 439, or 91 per 100,000. The model puts the fall at 88.8 percent, with a confidence interval, the range the true value is likely to sit in, of 77.9 to 94.4 percent. For the release years themselves the estimate was 90.2 percent.
Trapping the mosquitoes explains why the effect held. Four years after release, Wolbachia was present in at least 95 percent of the Aedes aegypti sampled in every monitored neighbourhood, and it was still high in the earliest release sites seven years on. The last zone took it up quickly too: three months after its releases the bacterium was in 51 percent of the mosquitoes sampled, and after five months, 92 percent. Infected females pass the bacterium to their offspring through the egg, so once a wild population carries it, it keeps it without further releases. A second Brazilian city offers an independent check. In Campo Grande, where more than 100 million mosquitoes were released between December 2020 and December 2023 across a population of about 882,850, dengue fell 63.2 percent in the neighbourhoods where the bacterium settled in at least 60 percent of mosquitoes, according to a study published in The Lancet Regional Health Americas in December 2025. By 2024 the bacterium was in 86.4 percent of mosquitoes across that city, and 89 percent of the treated areas were above the 60 percent mark.
Why it matters
Two cities are not the world, but the numbers are large and they come from the same routine records that a health ministry uses to count any other outbreak. During Brazil's 2024 epidemic the authors estimate Niteroi avoided between 5,242 and 11,660 dengue cases, 74 to 87 percent of the burden it would otherwise have expected. Before the releases the city sat among the ten municipalities with the most dengue per head in Rio de Janeiro state. It has ranked among the state's lowest every year since 2022. Over the five years to September 2024 it also notified only 131 chikungunya cases and 34 Zika cases, both diseases the same mosquito carries.
What is not settled is whether this becomes ordinary public health practice. The World Health Organization's Vector Control Advisory Group, the panel that assesses ways of fighting disease-carrying insects, concluded in 2020 that field trials had demonstrated an effect on disease and advised the agency to start developing guidelines. Six years on, that guidance is still being written: the group appraising the evidence is expected to meet in the first quarter of 2027, and recommendations are anticipated in late 2027. Several countries are already releasing or evaluating the mosquitoes without waiting for it. Niteroi's records are a measure of what the waiting costs, and the next few years will show how many other cities can repeat them.
Neither study is a randomised trial, and both rest on Brazil's routine notifications, which are mostly probable cases judged on symptoms rather than confirmed in a laboratory, so an unknown share are other fevers. The Niteroi authors say they cannot fully exclude underlying trends, or shifts in how people sought care and how cases were reported. In Campo Grande the protection showed up only where the bacterium was well established: areas still part-way through releases recorded a 9.2 percent fall that was not statistically significant. Aedes albopictus, another mosquito that can carry dengue, is still present in Niteroi, laboratory work shows the block on the virus is not complete, and the World Health Organization has issued no formal recommendation.
How does a bacterium stop dengue?
The bacterium lives inside the mosquito and sharply reduces its ability to pass dengue and other viruses on. Infected females pass it to their offspring through the egg, so once a wild population carries it, it keeps it without further releases. The block is not perfect: laboratory work shows some mosquitoes can still become infectious at very high viral loads.
Could something else explain the fall in Niteroi?
The authors say they cannot fully exclude underlying trends, or shifts in how people sought care and how cases were reported. Two things make coincidence harder to accept. Niteroi went from among the ten municipalities with the most dengue per head in Rio de Janeiro state to among the state's lowest every year since 2022, and a second city, Campo Grande, recorded a 63.2 percent fall after the same intervention.
Does it work everywhere the mosquitoes are released?
Not evenly. In Campo Grande the 63.2 percent reduction appeared only in neighbourhoods where the bacterium settled in at least 60 percent of mosquitoes, while areas still part-way through releases showed a 9.2 percent fall that was not statistically significant. In neighbouring Rio de Janeiro the levels dropped sharply in mid-2022, which the Niteroi authors say needs further investigation.
Has the World Health Organization approved this?
Not yet. Its Vector Control Advisory Group concluded in 2020 that field trials had demonstrated an effect on disease and advised the agency to begin developing guidelines. The group appraising the evidence is expected to meet in the first quarter of 2027, with recommendations anticipated in late 2027, and several countries and partners are already implementing or evaluating the approach.
- Tropical Medicine and Infectious Disease, Long-Term Durability and Public Health Impact of City-Wide wMel Wolbachia Mosquito Releases in Niteroi, Brazil, During a Dengue Epidemic Surge Primary
- The Lancet Regional Health Americas, The impact of large-scale release of Wolbachia mosquitoes on dengue incidence in Campo Grande, Brazil: an ecological study
- World Health Organization, WHO update on the assessment of Wolbachia population replacement for the control of Aedes-borne diseases
One verified piece of progress, every other day.
Good news, checked against primary sources. Free.
Powered by Buttondown