Early-life health turns less on feeding duration than on conditions and access

Duration alone tells a thin story

In a preprint, researchers following a Canadian cohort report that breastfeeding duration trajectories, constant versus decreasing, are not associated with the risk of infant challenging behaviours, and not associated with the risk of infant emergent medical visits. That is a null result, and null results are easy to misread in both directions. It does not say feeding does not matter; it says that, in this cohort, the pattern of duration over time did not track these two outcomes. This suggests advice framed purely around lasting longer may be aiming at the wrong lever for these particular outcomes, and that studies should separate duration from exclusivity, timing and the conditions around feeding before drawing conclusions. m:…0865

A second preprint from the same research group sharpens the point. Its authors report that maternal ADHD is associated with a reduced probability of six-month exclusive breastfeeding, with an average treatment effect on the treated of minus 0.105, and that mothers with ADHD in the cohort show 90% intention to breastfeed but only 28% actual six-month exclusive breastfeeding. The gap between intention and outcome is the finding: this is not a group that needs persuading, but a group whose plans do not survive contact with the practical demands of sustained exclusive feeding. The same preprint reports higher rates of prenatal depression at 38%, cigarette smoking at 13% and alcohol use at 19% among mothers with ADHD in the cohort, which suggests support needs to address co-occurring conditions, not feeding advice alone. m:…7126

Support, not intention, is the gap

This suggests the policy question changes shape. If intention is already at 90% and sustained exclusive breastfeeding at 28%, then more messaging about benefits is unlikely to close a 62-point gap, while practical support, screening for depression, and help with the daily work of feeding might. The preprint does not test interventions, so it cannot say which support would work; it can only locate the gap. But locating it matters, because a programme designed around persuasion would spend its effort where the evidence says the problem is not. The honest reading is narrower and more useful: for mothers with ADHD, the barrier appears to be follow-through under real-life conditions, not willingness. m:…7126

A third preprint shifts the focus earlier, to pregnancy itself. Its authors report that higher maternal glycated albumin in early and mid-late pregnancy is associated with an increased risk of large-for-gestational-age infants, while HbA1c is not associated with large birth size at either timepoint. Glycated albumin-total is lower in women who develop gestational diabetes during early pregnancy compared with non-diabetic controls, the authors report. This suggests the standard marker may miss a signal that a shorter-term marker captures, at least for this outcome. For clinicians and researchers, the implication is a testable one: whether glycated albumin adds information beyond HbA1c for identifying pregnancies at risk of large birth size, not a replacement of existing screening on preprint evidence alone. m:…2429

Conditions before birth, access after

The fourth preprint brings access into the picture. Its authors report that combined oral rehydration salts and zinc coverage for childhood diarrhoea in Bangladesh increased from 35.0% in 2019 to 42.9% in 2025, real progress that still leaves most children without the combined treatment. Wealth inequality in combined coverage persists in 2025 with a slope index of 13.0 percentage points, the authors report, and pharmacy-only care is associated with lower combined coverage compared with public care, with an adjusted prevalence ratio of 0.59. This suggests the constraint is not only supply or knowledge, but where families seek care and what they can afford when they get there. m:…3299

Read together, the four preprints tell a consistent story about early-life health: duration of feeding, on its own, explains less than the conditions around mother and child. A feeding trajectory shows no link to two infant outcomes; a maternal condition predicts whether feeding plans survive; a maternal blood marker tracks birth size better than the standard test; and a cheap, proven diarrhoea treatment still reaches richer children first. This suggests the highest-value work is less about telling families to do more of one thing for longer, and more about identifying which mothers need support, which pregnancies need closer monitoring, and which care routes actually deliver treatment. All four findings are preprints, so each needs confirmation, but the direction they point is the same. m:…0865 m:…7126 m:…2429 m:…3299