In two Yemeni governorates, 4 in 10 infants under six months showed anthropometric failure

Nutrition programmes in emergencies tend to concentrate on children aged 6 to 59 months, the window in which most screening tools are validated. A study published in Nutrients on 13 July looks at the age group below that line — infants under six months — and finds that undernutrition is already widespread by the time it is usually first measured [s1].

What was measured

The facility-based survey ran for thirteen months across five health facilities in two Yemeni governorates, and analysed 5,053 infants under six months attending routine health visits [s1]. Researchers measured weight, length and mid-upper arm circumference, and computed the standard anthropometric indicators: weight-for-age, length-for-age and weight-for-length z-scores [s1].

The prevalence figures

Among the infants, 19.3% were wasted (95% CI 15.1–24.3), 17.7% were stunted (95% CI 13.8–22.5), and 30.5% were underweight (95% CI 21.8–40.7) [s1]. Combining the forms, 40.9% met the composite index of anthropometric failure — a summary measure that counts a child as failing if they show any of wasting, stunting or being underweight (95% CI 34.7–47.5) [s1].

Within that group, the failures were often stacked: 17.7% had a single form of anthropometric failure and 23.3% had concurrent failures (95% CIs 12.5–24.4 and 17.9–29.7) [s1]. The most severe composite category, the composite index of severe anthropometric failure, applied to 12% of all infants (95% CI 10.8–13.4) [s1]. Concurrent failure matters clinically because a child who is simultaneously wasted and stunted, or wasted and underweight, carries a higher risk of death than one showing any single deficit.

The composite indices exist precisely to catch this. Reporting wasting, stunting and underweight separately can understate how many children are in trouble, because the same child may appear in more than one column or, conversely, be missed if a programme tracks only one indicator. The composite index of anthropometric failure counts a child once if they show any form of failure, giving a single figure for the share of infants who are failing to grow normally by at least one measure — here, more than four in ten [s1].

The measurement finding

The study's second contribution is methodological, and it is the kind of detail that changes what a screening programme catches. The authors compared how well weight-for-age below −2 z-scores and mid-upper arm circumference identified infants with anthropometric failure [s1]. Weight-for-age overlapped with a far higher share of affected infants than the arm-circumference tape did: 74.4% versus 28.8% for the composite failure index, and 92.1% for the most severe category [s1].

Mid-upper arm circumference is the backbone of community malnutrition screening because it is fast, cheap and needs no scale. But it was designed and validated for older children, and this study suggests that in the under-six-month group it misses most of the undernutrition that weight-for-age detects [s1]. A programme relying on the tape alone in this age band would, on these figures, screen past the majority of malnourished infants.

The authors examined this deliberately, assessing the overlap of different arm-circumference thresholds and weight-for-age below −2 z-scores against the various forms of anthropometric failure [s1]. The consistency of the pattern across those comparisons — weight-for-age catching more in every case — is what turns a single statistic into a screening argument. It does not mean the tape is useless; it means that for the youngest infants, a scale-based measure identifies children the tape would let through, and that a programme with the capacity to weigh infants would find more of the ones who need feeding support.

Limits

This is a cross-sectional study at health facilities, so it describes infants brought to routine visits in two governorates, not a random sample of all infants in Yemen — and children who never reach a facility may differ from those who do. The prevalence figures carry wide confidence intervals, especially for underweight (21.8 to 40.7%), reflecting the sample size and clustering [s1]. The measurement comparison is an association within this dataset, not a trial of screening strategies.

What the study does support is a specific caution: that in conflict-affected Yemen, undernutrition is present in infants below the age most tools are built for, and that the standard field screen may not be the right instrument to find it [s1]. The authors' implication is that weight-for-age deserves a larger role in screening the youngest infants.