Baby-led weaning vs spoon-feeding: what the trials actually found

Babies allowed to feed themselves from the start of solids are, on the best evidence, no slimmer, no more likely to be iron-deficient and no more likely to choke than spoon-fed babies — but only when parents are actively coached to offer safe, iron-rich finger foods [s1][s2][s3]. The popular case for baby-led weaning, that it prevents overweight, was not borne out in the one large randomised trial built to test it [s1]. The practical verdict is that either approach can be done well, and that the safety and nutrition details matter more than the method's label.

What the two approaches are

In traditional complementary feeding, an adult spoons pureed food to a baby who gradually moves on to lumps and finger foods. In baby-led weaning, the infant self-feeds whole soft foods from the very beginning of solids, skipping purees, on the theory that self-regulation of intake protects against overeating. Guidance recommends starting complementary foods at around six months while continuing milk feeds [s1].

Most of the reliable evidence comes from a single, well-run programme in New Zealand: the Baby-Led Introduction to SolidS (BLISS) trial, which randomised 206 women in late pregnancy to a control group given usual care or to a modified baby-led approach with eight support contacts from before birth to nine months [s1]. "Modified" is the key word: the intervention was designed specifically to address the two biggest worries about self-feeding — choking and iron — by teaching parents which foods to offer [s2][s3].

Does it prevent overweight? No measurable effect

The trial's primary question was whether self-feeding produces a healthier body weight, and the answer was no. Body mass index z-score did not differ significantly between groups at 12 months (control 0.20, BLISS 0.44; adjusted difference 0.21, 95% confidence interval 0.07 lower to 0.48 higher) or at 24 months (adjusted difference 0.16, 95% CI 0.13 lower to 0.45 higher) [s1]. At 24 months, 5 of 78 children (6.4%) were overweight in the control group compared with 9 of 87 (10.3%) in the baby-led group, a difference that was not statistically significant (relative risk 1.8, 95% CI 0.6 to 5.7) [s1].

There were differences in eating behaviour: baby-led infants were reported to have lower satiety responsiveness at 24 months and, at 12 months, less food fussiness and greater enjoyment of food [s1]. But the headline claim — that self-feeding yields a leaner child — did not hold up in the data [s1].

Does it cause iron deficiency? Not with the right coaching

Iron is the nutrient of greatest concern once milk alone stops being enough, and self-feeding was suspected of delivering too little. In the trial's iron analysis, differences in median dietary iron intake between groups were not significant at 7 months (0.6 mg/day, 95% CI 1.0 lower to 2.3 higher) or 12 months (0.1 mg/day lower, 95% CI 1.6 lower to 1.4 higher) [s2]. Nor were there significant differences at 12 months in plasma ferritin (difference 2.6 µg/L, 95% CI 10.9 lower to 5.8 higher) or in the prevalence of depleted iron stores, early functional iron deficiency or iron deficiency anaemia (all P ≥ 0.65) [s2]. The authors' conclusion is precise: a baby-led approach does not appear to raise the risk of iron deficiency when parents are advised to offer high-iron foods at each meal [s2].

Does it cause choking? No more than spoon-feeding

The most feared risk is choking, and here the evidence is reassuring but comes with a warning. In the choking analysis, 35% of all infants choked at least once between 6 and 8 months of age, with no significant difference in the number of choking events between groups at any time point (all P > 0.20) [s3]. Baby-led infants gagged more often at 6 months (relative risk 1.56, 95% CI 1.13 to 2.17) but less often at 8 months (RR 0.60, 95% CI 0.42 to 0.87) — consistent with earlier practice at managing textures [s3].

The warning is that both groups were doing something risky: at 7 and 12 months, 52% and 94% of infants respectively were offered foods that pose a choking risk during a 3-day food record, again with no difference between groups [s3]. In other words, self-feeding done with safety coaching was not more dangerous than traditional feeding — but unsafe foods were common in both, which is where the real hazard lies [s3].

Whichever route a family takes, the timing and manner of introducing common allergens is a separate, important question; we cover it in our explainer on early allergen introduction with solids, and the evidence on milk feeding in our review of breastfeeding and the PROBIT trial.

How to read this without overreaching

The evidence supports a permissive, safety-first conclusion. Baby-led weaning is a reasonable choice that did not, in the trial, deliver its promised weight benefit, but also did not increase iron deficiency or choking when parents were taught to offer safe, iron-rich foods [s1][s2][s3]. The method matters less than the execution.

The limits are real. BLISS tested a supported, modified version of baby-led weaning, not the unstructured do-it-yourself version many families follow, and the authors caution that their reassuring findings may not transfer to it [s1]. The trial was a single centre in one country, retention was 80.5% at 24 months, and it was not powered to detect rare choking events [s1][s3].

This article is informational and is not medical advice. It does not tell any family how or when to introduce solids to a particular baby; those decisions should be discussed with a qualified clinician or health visitor.