Azithromycin campaigns to cut child deaths raised gut drug resistance in Niger

Giving young children a broad-spectrum antibiotic twice a year has been shown to keep more of them alive in the highest-mortality parts of sub-Saharan Africa. The obvious worry has always been what that does to antibiotic resistance. A large trial in Niger has now measured it, and published the answer in Nature Medicine on 17 March 2026 [s1].

The trial, called AVENIR, found that mass azithromycin distribution to under-fives did raise resistance to the drug's class in children's guts — modestly but measurably — while a parallel signal in the nose did not reach statistical significance [s1].

The design

AVENIR was a double-blind, cluster-randomised, placebo-controlled trial [s1]. It randomised 3,000 communities in a 1:1:1 ratio to three arms: two years of twice-yearly distribution of azithromycin to children aged 1-59 months (the child-azithromycin arm); azithromycin to infants aged 1-11 months with placebo to the older children (the infant-azithromycin arm); or placebo to all children aged 1-59 months [s1]. The comparison of infant-only against child-wide dosing matters because it tests whether narrowing the target age group could preserve the survival benefit while spreading less antibiotic.

Mortality and safety were the trial's earlier co-primary results and have been published separately [s1]. The Nature Medicine paper reports the other co-primary endpoint: resistance [s1]. For that analysis, 150 communities — 50 per arm — were selected, yielding 4,382 rectal and 4,402 nasopharyngeal samples to measure changes in macrolide resistance in the gut and the nasopharynx [s1].

What it found

The trial met its primary resistance endpoint for the gut but not for the nasopharynx [s1].

In the gut, the macrolide resistance burden, expressed as a fold change between arms, was highest when child-wide dosing was compared with placebo: 1.16 (95% confidence interval 1.06-1.28; p<0.01) [s1]. Comparing child-wide dosing with infant-only dosing gave a fold change of 1.13 (95% CI 1.02-1.23; p=0.01) [s1]. The infant-only arm against placebo produced a fold change of 1.04 (95% CI 0.94-1.15; p=0.66), which was not statistically significant [s1].

That pattern is the study's most policy-relevant finding. The resistance signal tracked how many children were dosed: giving the antibiotic to all under-fives drove the largest increase, while restricting it to infants produced no detectable rise over placebo in the gut [s1].

In the nasopharynx, none of the between-arm differences reached significance [s1]. The fold changes were 2.14 (95% CI 0.93-4.99) for child-wide dosing versus placebo, 2.08 (95% CI 0.93-4.69) for infant-only versus placebo, and 1.03 (95% CI 0.46-2.30) for child-wide versus infant-only dosing — wide intervals that cross one [s1].

Why it matters

Mass azithromycin distribution sits on a genuine trade-off. On one side is a proven reduction in childhood deaths in high-mortality settings; on the other is the risk of accelerating resistance to a drug class that treats many common infections. WHO estimates that bacterial antimicrobial resistance was associated with more than 4.7 million deaths globally in 2021, which is the backdrop against which any programme that deliberately spreads an antibiotic has to be judged [s2]. That toll is also why the agency's global plan is tied to the 2024 UN General Assembly target of a 10% reduction in bacterial AMR-associated deaths in humans by 2030 — a goal that pulls in the opposite direction from expanding antibiotic distribution [s2].

AVENIR does not resolve that trade-off, but it quantifies part of it and points to a lever. The resistance rise was concentrated in the gut and scaled with the number of children treated, and the infant-only strategy showed no significant gut increase over placebo [s1]. If a narrower age target can retain enough of the survival benefit — a question the mortality analysis addresses — that would be a way to buy the lives while limiting the resistance cost.

The limits

The resistance analysis covered a subset of communities and two years of dosing; it measures the resistome shift over that window, not the downstream clinical consequences of it, and not what happens if programmes run longer [s1]. Fold changes summarise a complex microbial picture, and the non-significant nasopharyngeal result is a finding of uncertainty rather than of safety — the intervals are too wide to rule out a real effect [s1].

What to watch

The authors' own conclusion is a governance one: close monitoring of resistance should be an essential component of any azithromycin distribution programme aimed at child survival [s1]. As countries weigh scaling these campaigns, AVENIR turns an abstract worry into a measurable quantity and hands policymakers a concrete dial — who gets dosed — to turn. The next question is whether the infant-only approach that limited resistance here also keeps enough children alive to justify it.

Sources

• Mass azithromycin distribution and antibiotic resistance in the gut and nasopharynx: a cluster-randomized trial — Nature Medicine, 17 March 2026
• Antimicrobial resistance — Fact sheet — World Health Organization