The Dose That Rewired the Village
A trial in 110 Honduran villages found a maternal health programme rewired friendship and money ties. At a low dose, people left out lost connections.
Teach a few households in Honduras, and the people left out can lose their ties.
Between November 2016 and August 2018, community health workers taught maternal, newborn, and child health in 110 remote villages in Honduras. The trial covered 8,331 people. It did not teach every house the same way. Some villages got no classes. In others, 5 to 30 percent of households were taught. In others, half to all of them were.
Nicholas Christakis, Marios Papamichalis and Laura Forastiere at Yale, with Edoardo Airoldi of Temple, mapped who asked whom for health advice, who was friends, and who lent money, before the classes and after. The paper is in Nature, 14 October 2026. Yale wrote it up on 22 October.
In the low-dose villages, people who took the classes tended to drop untreated neighbours they used to ask about health. The village got less connected. In the high-dose villages, treated people added ties. Friendship and financial links moved with the health links. The pattern was still visible three years after the classes stopped, in a subset of 56 villages.
The lessons were about pregnancy and infants. The side effect was the map of the village. Christakis's point, in the Yale account, is that a programme can change who is connected to whom even when that was not the aim.
Key Facts
- 8,331 people, 110 villages, randomised doses of a 22-month maternal and child health programme
- Classes ran November 2016 to August 2018
- Low dose, 5–30 percent of households: treated people severed health-advice ties to untreated neighbours
- High dose, 50–100 percent: treated people increased ties
- Friendship and financial networks moved with the health network
- The pattern lasted about three years after the programme ended
- Nature, 14 October 2026; Yale News, 22 October 2026
Why This Matters
Health programmes are scored on clinic visits and vaccines. This one moved who a person can borrow from. The practical lesson is uncomfortable and useful: a partial rollout can isolate the households you did not reach. Saturation is not a luxury. In this trial it was the difference between a village that pulled together and a village that pulled apart.
What We Don't Know Yet
Whether a planner can hit the high-dose pattern on purpose in another country. The classes ended in 2018. This is the network paper, not a new programme, and not a measure of infant survival. We are not quoting secondary percentages for how many ties broke. The direction is in the Nature paper. The exact percentages in some write-ups are not a figure we are willing to put in a reader's mouth.