Breastfeeding: what is established, and what is not
Some of the benefits are proven by randomized trial. Others shrink when you compare siblings.
What actually happens in the United States, for babies born in 2019:
| Milestone | Share of infants |
|---|---|
| Ever received any breast milk | 83.2% |
| Any breast milk at 1 month | 78.6% |
| Any breast milk at 6 months | 55.8% |
| Exclusive breast milk at 6 months | 24.9% |
Eighty-three percent start. Twenty-five percent are still exclusively breastfeeding at six months, which is what is recommended. That gap is not a failure of individual determination. It is what happens in a country without universal paid leave.
The evidence, separated by how strong it is. There is one large randomized trial of breastfeeding promotion, involving 17,046 mother and infant pairs, which is the only design that can separate the effect of breastfeeding from the effect of being the kind of family that breastfeeds.
| Finding | Strength |
|---|---|
| Fewer gut infections in the first year: 9.1% against 13.2% | Randomized. Holds. |
| Less eczema: 3.3% against 6.3% | Randomized. Holds. |
| Verbal IQ at 6.5 years, 7.5 points higher | Randomized. Holds, but full-scale IQ did not reach significance. |
| Height, weight, body mass index, blood pressure at 6.5 years | Randomized. No difference. |
| Most long-term child health and wellbeing measures | Comparing siblings within families, differences largely disappear. |
The trial's own conclusion about growth and blood pressure was that Previously reported beneficial effects on these outcomes may be the result of unc[ontrolled confounding].
A sibling analysis of 25 years of US data found that once children were compared with their own brothers and sisters, all but one of eleven outcomes lost statistical significance.
The point is that the pressure applied to women is out of proportion to what the evidence supports, and that a woman who cannot or does not breastfeed has not harmed her child's future in the ways she is often led to fear. Both halves of that are true and both belong in the same conversation.
Mastitis affected 20% of women in the six months after birth in a 1998 Australian prospective cohort of 1,075, with 75% of cases in the first seven weeks. It is old and not from the United States, but it remains the most commonly cited incidence figure. Management guidance changed substantially in 2022, when the Academy of Breastfeeding Medicine reframed mastitis as a spectrum running from ductal narrowing through inflammation to infection. Several older pieces of advice were reversed. If you are given mastitis advice, it is fair to ask whether it is based on the current protocol.
On supply, the most common reason for stopping early is a perception of insufficient milk, which is usually not true insufficiency. Genuine low supply exists and is worth investigating, but the first question is almost always about how often and how effectively milk is being removed, not about supplements.