Three separate reports in recent months describe the same shape. In northern Yemen, admissions of severely malnourished children with complications rose 62 per cent in a year. In Baidoa, Somalia, acute malnutrition among displaced children screened by MSF ran near 50 per cent, with readmissions above 60 per cent. In southern Afghanistan, admissions to therapeutic feeding centres were more than 30 per cent above the three-year average for the same months.

Three countries, three different conflicts, three different climates. One common factor named in all three reports: money.

What actually gets cut

A funding reduction does not arrive as a decision to treat fewer children. It arrives as a decision about which parts of a system to keep running, and the parts that go first are consistently the same.

Nutrition programmes have a layered structure. At the base are community health workers who screen children — measuring mid-upper arm circumference, checking for oedema — and refer them early. Above that sit outpatient therapeutic feeding centres, where uncomplicated cases are treated at home with ready-to-use therapeutic food. At the top are inpatient units for children with medical complications, which need beds, therapeutic milk, antibiotics and staff around the clock.

The base of that structure is the cheapest and the easiest to stop. The top is the most expensive and the hardest to defend cutting, because closing an inpatient ward is visibly a decision to let children die. So screening and outpatient coverage shrink first, and the inpatient wards stay open — and then fill up.

The mechanism, stated plainly

MSF’s field teams describe it as children reaching them too late, in critical condition, with preventable complications. Their medical coordinator in southern Afghanistan put it as a breakdown in the systems designed to detect and treat malnutrition early — arguing that an effective response needs every layer working, from outpatient identification through to inpatient care.

The WHO recorded the scale of the withdrawal in Afghanistan specifically: significant reductions in international funding since early 2025 led to the suspension or closure of 445 health facilities, including 203 mobile health and nutrition teams. Mobile teams are precisely the layer that finds children in places without a clinic.

Distance does the rest. Where the nearest service closes, families travel further and pay more to reach care. A mother quoted by MSF in Yemen described a two-hour journey costing about £47, from a family of twelve with one earner. That is a filter, and what it filters out is early treatment.

The argument on the other side

It would be too easy to present this as an obvious error by donor governments, and it is worth setting out their case properly.

Donors argue that budgets are finite and shrinking, that domestic political support for aid spending has weakened across several major contributors, and that when money must be cut, triage toward the most immediately life-saving interventions is the defensible principle. They also argue — with some evidence — that long-running humanitarian nutrition programmes can substitute for national health systems rather than building them, and that indefinite external funding of a country’s basic child health provision creates a dependency that is itself a policy failure.

Humanitarian organisations do not generally dispute the dependency point. Their response is about sequencing: that a national system cannot be handed responsibilities during an active conflict or a drought displacement crisis, and that withdrawing the outer layers of a nutrition programme before anything replaces them produces the readmission cycle visible in Baidoa — the same children treated repeatedly, at the most expensive point of care.

Both arguments are about what happens over years. The reports above are about what is happening this year, and on that narrower question the evidence points one way.

What to look for in the next set of figures

Three indicators are more informative than a caseload total. The proportion of admissions arriving with complications, which measures how late children are reaching care. The readmission rate, which measures whether anything is changing at home. And the number of functioning outpatient sites, which is the thing that actually got cut and is rarely reported.

A rising caseload alone can mean better detection. Those three, moving together, cannot.

Related: One Hospital in Northern Yemen and The Children of Baidoa Keep Coming Back.

Sources

Country figures from Médecins Sans Frontières reporting on Afghanistan, Yemen and Somalia; health facility closure figures from the World Health Organization as cited by MSF USA.