Pediatrics · Part One · Before Birth and Birth
Lesson 04 / 57
Nourishing Two Nervous Systems: What the Maternal Nutrition Trials Actually Found
In 2,399 women given 800 milligrams of DHA a day through pregnancy, the difference in their children’s cognitive score at 18 months was 0.01 points.
Maternal nutrition shapes fetal brain development mainly through deficiency rather than through supplements. Children of women with low first-trimester iodine were more likely to score in the lowest quartile for verbal IQ, odds ratio 1.58. Yet 200 micrograms of iodine daily left verbal IQ 0.7 points below placebo at age five to six, and 800 milligrams of DHA daily moved cognitive scores by 0.01 points. The Unified Model of Tone reads nutrition as one input to a developing state.
Iodine supplement trial, verbal IQ at age 5 to 6
0.7 points below placebo, p=0.77
DHA supplement trial, cognitive score at 18 months
difference of 0.01 points, p=0.99
Low first-trimester iodine and lowest-quartile verbal IQ
odds ratio 1.58
Anemia diagnosed before 30 weeks and intellectual disability
odds ratio 2.20
What is being measured
Nutrient status is what a laboratory reads in a pregnant woman: urinary iodine concentration, hemoglobin and iron stores, plasma fatty acid pattern, serum vitamin D. Intake is what a capsule changes. Status and intake are different numbers, and studies of the two answer different questions.
Why the two halves of the research disagree
A nutrient behaves like a floor rather than a dial. Below the floor, missing material limits what tissue can be assembled, and outcomes track the shortfall. Above it, more of the same material meets a system already stocked. A trial recruiting well fed women tests the flat top of that curve.
01The shape of the evidence
Deficiency moves child outcomes, and supplementing a well fed woman usually does not
Maternal nutrition research splits in two, and the halves point in different directions. Studies that measure how much of a nutrient a woman carries find real differences in her child’s later scores. Trials that add that nutrient to women who already have enough mostly find nothing.
The two results are not in conflict. They are what a threshold looks like when it is tested from both sides.
What the trials settled
Randomized trials are where causal claims get settled, and several large ones came back empty. Iodine at 200 micrograms daily produced a verbal IQ difference of 0.7 points favoring placebo (Gowachirapant 2017). DHA at 800 milligrams daily produced a cognitive difference of 0.01 points (Makrides 2010).
Folic acid around conception is the one nutrient intervention with a decisive trial result behind it, and the preconception window carries that evidence. It is the exception that proves how much a genuine deficiency matters.
02Findings
What the research shows
The figures below come from randomized trials, Cochrane reviews, a birth cohort and a national register study.
03Iodine
Iodine status tracks child scores, and the iodine supplement trial found nothing
Iodine is the clearest case, because the same nutrient has been studied both ways within a few years. A British birth cohort measured urinary iodine in stored first-trimester samples from 1,040 women. The median was 91.1 micrograms per liter (Bath 2013).
The analysis adjusted for 21 socioeconomic, parental and child factors. Children of women below the sufficiency cutoff more often scored in the lowest quartile for verbal IQ at age eight (odds ratio 1.58), reading accuracy (1.69) and reading comprehension (1.54). Scores worsened stepwise as maternal iodine fell further.
A meta-analysis organized the wider literature by design. Average effect sizes ran from 0.46 to 0.68, translated to 6.9 to 10.2 IQ points lower in iodine deficient children (Bougma 2013). The authors list their own limits: weak designs, omitted confounders and small samples.
Then the trial arrived
832 women in Bangalore and Bangkok were randomized to 200 micrograms of iodine daily or placebo, with 87 percent compliance (Gowachirapant 2017). At age five to six, verbal IQ was 89.5 against 90.2. Performance IQ was 97.5 against 99.1, and the global executive composite was 90.6 against 91.5.
Every confidence interval crossed zero. Those women entered with a median urinary iodine of 131 micrograms per liter, well above the 91.1 in the British cohort. The trial topped up women near the threshold rather than correcting women far below it.
04Omega-3 and DHA
The DHA trial record is mixed, and the largest trial found no cognitive difference
DHA is the supplement most often sold to pregnant women on a brain-development argument, and it has the largest trial record of any nutrient here. That record is mixed, and the biggest study in it is null.
The DOMInO trial randomized 2,399 women at under 21 weeks across five Australian maternity hospitals. They took 800 milligrams of DHA daily or vegetable oil, and 96.7 percent completed (Makrides 2010). Among 726 children assessed at 18 months, the adjusted cognitive difference was 0.01 points and the language difference was 1.42 points lower.
A meta-analysis pooled 11 trials and 5,272 participants and found no differences in cognitive, language or motor scores (Gould 2013). One exception appeared, a mean difference of 3.92 points at ages 2 to 5, drawn from 2 trials at high risk of bias.
A Norwegian trial gave cod liver oil from week 18 of pregnancy to three months after delivery, and reported higher intelligence scores at four years. Retested at seven, the 82 supplemented children and the 61 controls showed no significant differences (Helland 2008).
What omega-3 does change
A Cochrane review of 70 trials and 19,927 women found a real effect in a different outcome. Preterm birth before 37 weeks fell from 13.4 to 11.9 percent, and birth before 34 weeks fell from 4.6 to 2.7 percent (Middleton 2018). Gestation beyond 42 weeks rose from 1.6 to 2.6 percent. On cognition, IQ, vision and language the review reports very few differences.
05Iron and vitamin D
Iron carries the strongest human signal, and vitamin D carries none for the brain
Iron produces the largest association in this literature, and it comes from a register rather than a trial. A Swedish cohort followed 532,232 children born between 1987 and 2010 (Wiegersma 2019). Against mothers with no anemia diagnosis, anemia diagnosed in the first 30 weeks carried adjusted odds ratios of 1.44 for autism, 1.37 for attention-deficit/hyperactivity disorder and 2.20 for intellectual disability.
A matched sibling comparison gave 2.25 and 2.59 for autism and intellectual disability. Anemia diagnosed after 30 weeks carried greatly diminished associations. That finding is about timing in a population, read from diagnostic registers. It describes groups, not a choice any woman made. Anemia in pregnancy is common. It is screened for in routine antenatal care, and it is treated.
What iron supplementation reliably does is fix iron. A Cochrane review of 57 trials and 48,971 women found maternal anemia at 4.0 against 7.4 percent, and iron-deficiency anemia at term at 5.0 against 18.4 percent (Finkelstein 2024). Low birthweight fell from 6.1 to 5.2 percent, while preterm birth and neonatal death were unchanged.
Vitamin D
Vitamin D has no established neurodevelopmental effect in pregnancy, because the trials have not measured one. The 2024 Cochrane update moved 20 of the previous 30 studies to awaiting classification after assessing trustworthiness, leaving 10 (Palacios 2024). What remains is very uncertain evidence on pre-eclampsia and preterm birth. No child neurodevelopmental outcome is reported at all.
06What the nutrients build
Each of these nutrients has a structural job, and the job runs on a schedule
The trials above ask whether adding a nutrient to a well fed woman changes a later score. They do not ask whether the nutrient is needed. It plainly is, and each one has a known place in the assembly of a nervous system.
Folate supplies methyl groups for the rapid cell division that closes the neural tube in the first month, before most pregnancies are confirmed. That is the one place a trial settled the question, and the preconception window carries the result.
DHA is construction material rather than fuel. It is built into the membranes of neurons and of the retina, and it accumulates fastest across the third trimester and the first two years. A membrane rich in it behaves differently from one that is not, which is where the interest in supply began.
Thyroid, myelin and dopamine
Iodine has no direct route to the brain. It becomes thyroid hormone, and thyroid hormone drives myelination and the timing of neuronal migration. That is why real deficiency is expensive, at 6.9 to 10.2 IQ points across designs (Bougma 2013). Adding 200 micrograms to a stocked woman moved verbal IQ by 0.7 points (Gowachirapant 2017).
Iron does two jobs at once. It is required to build myelin, and it runs the dopamine systems of the basal ganglia that carry movement and motivation. A fetus banks its own stores late in pregnancy, which is one reason anemia before 30 weeks tracked outcomes that anemia after 30 weeks did not (Wiegersma 2019).
The shared state
Mother and fetus do not share a brain. They share a chemistry. Glucose, fatty acids and thyroid hormone all reach the fetus through one placenta, so maternal chemistry is the medium this development happens in.
That is why a thyroid stands between iodine and a myelinating brain. It is also why a fetus banking iron across the last weeks draws from a supply its mother is running at the same time. Nutrition in pregnancy is one circulation under sustained load rather than two separate accounts.
What care during pregnancy involves
Care in pregnancy is examination and palpation of the pelvis, the low back and the upper neck, with sustained contact graded to the person on the table rather than a twist or a thrust. Anything obstetric is referred promptly.
The model expects a body carrying less mechanical and autonomic load to meet a demanding season with more in reserve. That is a claim about the state a pregnancy is carried in. Recording maternal autonomic variability and recovery time across the trimesters, in the same women, would show it. Nutrition decisions belong with the obstetric or midwifery team.
07The model’s claim
What the Unified Model of Tone predicts about maternal nutrition
Everything above is established science, including null trials that run against the intervention being sold. What follows is this model’s reading of that record, stated as ours rather than drawn from the papers cited.
The model rests on a law that lists a meal among its plainest cases. No input creates an outcome on its own; an input interacting with a tone creates an outcome. A nutrient is therefore a constraint on the receiving organization rather than a lever attached to intelligence, and a constraint already satisfied cannot be relieved further. The same capsule is a different event in a stocked body than in a depleted one. That is why 800 milligrams of DHA given to well fed women produced a difference of 0.01 points (Makrides 2010).
The prediction
From that follows a claim the nutrition literature does not make. No capsule arrives at an empty body, so measured maternal status, rather than assigned supplement group, is the variable carrying information. That information should show up across several infant readouts at once: autonomic regulation, feeding coordination, sleep and postural control. Those are this page’s versions of the variability, coupling, responsiveness and recovery the model treats as one factor.
It also predicts where the null trials came from. A trial that hands every woman the same dose tests the input alone and averages away the tone receiving it. Population recruitment enrolls mostly sufficient women, so the treatable range is thin. Trials recruiting on measured deficiency should separate from placebo where population-recruited trials did not.
That is a different claim from anything about treatment. This is a claim about how development is organized rather than about what treatment does. If maternal nutrient status, infant autonomic regulation, feeding coordination and postural control are shown to move together, the unification claim is confirmed.
08The tone reading
How maternal nutrition expresses tone
Every topic in this library expresses all of tone. In maternal nutrition three aspects carry the signature, because the nutrient that tracked lower verbal IQ in one cohort moved nothing at all in a trial.
Input quality
Nutrient status is the raw material a nervous system is assembled from. A median urinary iodine of 91.1 and one of 131 are different starting stock.
Constraint
A nutrient sets a floor rather than a dial. Below it the shortfall bounds what tissue can form, and above it more material adds nothing.
Time course
Anemia diagnosed before 30 weeks tracked raised odds of later diagnoses. The identical diagnosis after 30 weeks tracked almost nothing. Timing decided the reading.
The remaining foundations run through this topic as well. Load: pregnancy is a sustained metabolic demand met from one supply. Coupling: iodine reaches the brain through thyroid hormone rather than directly. Set point: iron stores and thyroid output are regulated targets, not stockpiles. Gain: the same intake produces different status in different women. Prediction: the placenta allocates against a forecast. Oscillation: appetite, sleep and nausea cycle across trimesters. These are readings of one organization rather than separate systems, which is the core claim of the Unified Model of Tone.
09Across the library
How this page relates to the rest of the library
Maternal nutrition sits between what a pregnancy supplies and what a newborn nervous system runs on.
The folate evidence, and the weeks of development that precede a positive test.
The maternal autonomic state as an input, and what the research establishes.
What reaches a fetus through the uterine wall, and on what schedule.
The sensory channels that come online before birth, each with a dated window.
The system this nutrition supplies, running on about a quarter of its adult brain volume.
The fidelity of what a system receives, treated as a measurable state in adults.
Sustained metabolic and physiologic demand, and how it is read.
10Frequently asked
Questions families ask about nutrition in pregnancy
Does what I eat in pregnancy change my baby’s brain?
Deficiency does, and topping up an adequate diet mostly does not. Children of women with low first-trimester iodine more often scored in the lowest quartile for verbal IQ, odds ratio 1.58. Maternal anemia diagnosed before 30 weeks carried an odds ratio of 2.20 for intellectual disability. Yet randomized trials of iodine and DHA in reasonably fed women returned differences of 0.7 and 0.01 points. Eating adequately matters. Stacking supplements on top of adequate has not shown a cognitive return.
Should I take a DHA or fish oil supplement?
That belongs with your maternity team, and the evidence is worth knowing first. The largest trial gave 800 milligrams of DHA daily to 2,399 women and found a cognitive difference of 0.01 points in their children at 18 months, with language scores 1.42 points lower. A meta-analysis of 11 trials and 5,272 participants found no differences. Omega-3 does reduce preterm birth, from 13.4 to 11.9 percent, so the case for it is about birth timing rather than intelligence.
Do I need extra iodine?
Your status matters more than your supplement. In a British cohort with a median urinary iodine of 91.1 micrograms per liter, low status tracked lower verbal IQ, reading accuracy and reading comprehension. Scores worsened stepwise as iodine fell further. But a randomized trial of 200 micrograms daily in 832 women found verbal IQ of 89.5 against 90.2 at age five to six, and those women started near the threshold. Ask your maternity team about your own status rather than guessing at it.
What about iron?
Iron carries the strongest signal here, and the finding is about timing. Among 532,232 Swedish children, maternal anemia diagnosed in the first 30 weeks was associated with autism, attention-deficit/hyperactivity disorder and intellectual disability, the largest odds ratio being 2.20. The same diagnosis after 30 weeks carried greatly diminished associations. Supplementation reliably corrects the iron itself, cutting maternal anemia from 7.4 to 4.0 percent across 57 trials. Anemia is screened for routinely in antenatal care, and it is treated.
Does vitamin D help my baby’s brain?
No trial has measured that. The 2024 Cochrane review moved 20 of the previous 30 studies to awaiting classification after assessing their trustworthiness, leaving 10 studies, with the main comparison resting on 8 studies and 2,313 women. What remains is very uncertain evidence on pre-eclampsia and preterm birth, and a possible reduction in low birthweight from 3 studies of 371 infants. No child neurodevelopmental outcome is reported anywhere in it. Vitamin D dosing still belongs with your clinician.
What do these nutrients actually do in a developing brain?
Each has a structural job. Folate supplies methyl groups for the cell division that closes the neural tube in the first month. DHA is built into the membranes of neurons and the retina. Iodine becomes thyroid hormone, which drives myelination and the timing of neuronal migration. Iron is needed for myelin and for the dopamine circuits of the basal ganglia, and a fetus banks stores late in pregnancy. Deficiency limits the work. Extra material for a stocked body does not extend it.
What does the Unified Model of Tone say about maternal nutrition?
That a nutrient acts as a constraint on one organized state rather than as a lever attached to intelligence. A constraint already satisfied cannot be relieved further, which is the model’s reading of why 800 milligrams of DHA moved cognitive scores by 0.01 points. From that the model predicts that measured maternal status, not assigned supplement group, should carry information across infant autonomic regulation, feeding coordination, sleep and postural control, with compensation deciding which of those readings expresses it. That prediction is testable and separate from treatment.
11The sources
References
10 primary sources, each linked to its PubMed record. Figures quoted on this page were checked against the published abstract.
Related evidence