Long-term breastfeeding linked to lower risk of developmental delays

In the largest study to date on infant nutrition and child development, Israeli scientists analyzed data from 570,532 children and came to a simple but important conclusion: the longer a child is breastfed, the lower the risk of delays in speech, socialization, and motor skills — even after accounting for dozens of possible “confounding” factors. The study was published in the journal JAMA Network Open.

Background

Early childhood is a "window of opportunity" for laying the foundation for cognitive, speech and social development. Nutrition in the first months of life is one of the few modifiable factors. The World Health Organization recommends exclusive breastfeeding (BF) for the first 6 months, then continuing BF with complementary feeding for up to 2 years and longer.

What is already known. Many studies have found a link between breastfeeding and better cognitive performance in children. Proposed mechanisms include breast milk composition (long-chain PUFA, hormones, oligosaccharides), effects on the microbiome and immune system, and indirect effects of close skin-to-skin contact.

What is the problem with previous studies:

1. Confounding: Families with higher education and income are more likely to choose and maintain breastfeeding for longer—and these same factors themselves improve child development. Many studies have under- or completely omitted these variables.
2. Selectivity of samples and small sizes: results are difficult to generalize to the population.
3. Recall bias: feeding data were often collected retrospectively.
4. A rough classification of feeding: “breast versus formula” without taking into account duration and exclusivity, although in real life mixed feeding and different trajectories of breastfeeding cessation are common.
5. Medical factors of birth: Prematurity and low birth weight both reduce the chances of successful breastfeeding and increase the risk of developmental delays; without careful adjustment, it is easy to obtain distorted conclusions.

What was studied

The team linked two national bases:

• Routine developmental monitoring in children's clinics (Tipat Halav): During visits from birth to 6 years, nurses record growth, nutrition and achievement of age milestones.
• Disability Assignments from the National Insurance Institute: The registry reflects cases of severe neurodevelopmental disabilities (e.g. autism, severe ADHD) that actually require support and entitle to payments.

The analysis included children born after the 35th week without severe neonatal pathology and who came for at least one visit at 2–3 years. The key “exposure” was the duration and exclusivity of breastfeeding (BF).

How it was calculated

The researchers conducted three independent tests to avoid confusing the effect of breastfeeding with the influence of other factors:

1. Regressions adjusted for gestational age, birth weight, birth order, birth characteristics, maternal age and education, marital status, postpartum depression (EPDS), socioeconomic status, etc.
2. Matching pairs of "similar" children on all variables where the groups differed significantly.
3. Within-family analysis of 37,704 sibling pairs with different breastfeeding durations: this design partially “zeroes out” unobservable family characteristics (values, parenting style, genetics).

We separately checked whether prematurity (35–36 weeks) changes the picture.

Key Results (Percentages and Odds)

• Overall, 52% of children were breastfed for at least 6 months; of these, approximately 42% were exclusively breastfed during this period.
• Compared to breastfeeding <6 months:
  • Exclusive breastfeeding ≥6 months was associated with lower odds of delays (language/socialization/motor): AOR 0.73 (95% CI 0.71–0.76).
  • Non-exclusive breastfeeding ≥6 months - AOR 0.86 (0.83–0.88).
• Dose-response curve: risk reduction is most pronounced in the first months and reaches a plateau at approximately 10–12 months.
• Sibling analysis ("children in the same family"):
  • A child with ≥6 months of breastfeeding has a lower likelihood of delayed milestones than his/her brother/sister with <6 months of breastfeeding: OR 0.91 (0.86–0.97).
  • The odds of a serious neurodevelopmental diagnosis are lower: OR 0.73 (0.66–0.82).
• Absolute differences in the matched sample:
  • Any delay in milestones in late preterm infants: 8.7% with GA <6 months vs 6.8% with GA ≥6 months (–1.9 percentage points).
  • Any NDS (neurodevelopmental condition) in late preterm infants: 3.7% vs 2.5% (–1.27 percentage points).
  • For full-term babies, the absolute gains are smaller, but noticeable: -1.18 percentage points for delays and -0.73 percentage points for NRS.
• The signal was particularly robust for speech and socialization; statistics were insufficient for rare severe motor diagnoses.

Importantly, prematurity did not change the direction of the association (interaction not significant), but the higher baseline risk in preterms made the absolute benefit appear larger.

Why is this compelling?

• Very large sample size and national coverage.
• Many confounders are taken into account (including postpartum depression and social risk).
• Three independent analytical approaches showed the same thing.
• Sibling design reduces the influence of "invisible" family factors.

What this does not prove

This is an observational study. It does not prove causality and does not separate the "milk effect" from the "close contact effect" (skin-to-skin, frequent interaction). Also, it did not assess subtleties like expressed milk vs. latching, support at work, etc. And although the clinics cover >70% of children, the context is Israel, which is important for transfer to other health systems.

How do the findings relate to the recommendations?

The results support the WHO recommendations: aim for exclusive breastfeeding for the first 6 months, and then continue with complementary feeding. At the same time, the authors emphasize: the goal is to minimize functional gaps, and not to “eradicate neurodiversity.”

What this means for parents and doctors

• If breastfeeding is possible and comfortable, each additional month is associated with a slightly lower risk of delays, especially speech and social ones.
• If breastfeeding does not work or is contraindicated, you have not "spoilt" anything. Development is multi-component: sleep patterns, communication, reading, games, hearing and vision, timely screenings - all this also has a strong influence.
• For premature babies, the absolute “benefit gain” from prolonged breastfeeding may be greater – it is worth discussing lactation support in the ward and after discharge.

What is needed at the level of policy and health services

• Breastfeeding-friendly maternity hospitals (BFHI initiative), infant rooms with counseling.
• Flexible employment and parental leave, conditions for pumping.
• Correct information and limiting aggressive marketing of breast milk substitutes.

Summary

Children who were breastfed for at least six months were less likely to have delays in key milestones and were less likely to have serious neurodevelopmental diagnoses—an effect that was noticeable even when strictly controlling for sociomedical differences and when comparing siblings within the same family. This is another compelling argument to support families who are willing and able to continue breastfeeding.