Cognition in Small-for-Gestational-Age Children

About 10% of newborns are classified as small for gestational age (SGA), defined as birth weight below the 10th percentile for gestational week. Most SGA newborns recover normal growth in the first two years. Their cognitive trajectory, by contrast, does not converge with their term-born peers: across longitudinal cohorts, SGA individuals show a roughly 8-point IQ gap from infancy through adulthood, with a specific profile of weakness in perceptual reasoning and processing speed and a different pathophysiology from prematurity-related deficits. The most informative recent evidence comes from the Bavarian Longitudinal Study, where Eves and colleagues (2020) tracked SGA versus appropriate-for-gestational-age (AGA) cognitive performance from infancy through age 26.

What the Eves 2020 study actually found

The Bavarian Longitudinal Study followed 481 individuals — 121 born SGA and 360 born AGA — at six time points: 5 months, 20 months, 4 years 8 months, 6 years 3 months, 8 years 5 months, and 26 years. Cognitive ability was assessed with age-appropriate instruments (Bayley, Kaufman, K-ABC, WISC, WAIS) and rescaled to a common IQ metric.

The headline finding was a stable, durable cognitive gap:

• SGA individuals scored on average 8 points lower on IQ than AGA individuals across the six assessments.
• The gap was present at the earliest assessment (5 months) and persisted through age 26.
• The gap narrowed only slightly into adulthood — from roughly 9 points in childhood to roughly 7 points at 26.
• The effect was independent of very preterm/very low birthweight birth status, low socioeconomic status, and parent–infant relationship quality, each of which contributed additional variance.

One reframing matters for interpretation. Earlier literature debated whether SGA cognitive deficits were essentially absorbed by socioeconomic factors — i.e., that SGA was a marker of disadvantage rather than a cause of cognitive differences. The Bavarian analysis adjusted for SES alongside SGA, found that SES contributed comparable variance to SGA on its own, and showed that the SGA effect remained independently significant. SES and SGA both matter; neither explains the other away.

What is the cognitive profile?

Term-born SGA children do not show a uniform “low IQ” picture. Vitale and colleagues (2026), assessing term-SGA children with the WISC-IV, found a specific pattern:

• Perceptual Reasoning Index (PRI): significantly lower than AGA controls.
• Processing Speed Index (PSI): significantly lower than AGA controls.
• Verbal Comprehension Index (VCI): preserved, no significant difference.
• Working Memory Index (WMI): preserved, no significant difference.

The same study found elevated internalizing behavioral symptoms (anxiety, withdrawal, somatic complaints) on the Child Behavior Checklist, and birth head circumference predicted perceptual-reasoning performance — a structural-developmental signal consistent with the brain-imaging evidence reviewed below. Notably, growth hormone therapy — sometimes used in SGA children with persistent short stature — showed no association with cognitive outcomes in this analysis, consistent with broader evidence that hormone therapy addresses growth without resolving the underlying cognitive trajectory.

The specific PRI/PSI profile aligns with a brain-development pattern where visuospatial processing and processing-speed networks are more vulnerable than language and verbal-memory networks. This profile has practical educational implications: SGA-born children may compensate well in language-dependent settings (verbal classroom instruction, reading, vocabulary) and look less impaired than they are, while struggling with visuospatial reasoning, math operations involving spatial mapping, and tasks under time pressure.

What does brain imaging show?

The structural evidence comes from two complementary sources: in utero fetal MRI and childhood follow-up imaging.

Meijerink et al. (2025), in a systematic review of fetal MRI studies of growth-restricted fetuses, summarized convergent findings across studies: total brain volume reduced by approximately 10% relative to normally-growing fetuses at comparable gestational ages, altered cortical folding with deeper fissures and asymmetric gyrification, and reduced N-acetylaspartate-to-choline ratios on MR spectroscopy indicating altered neuronal metabolism. The pattern was visible from the second trimester onward and was more pronounced in early-onset growth restriction (before 32 weeks) than in late-onset cases.

Korkalainen et al. (2022) extended the picture into childhood. In 8- to 10-year-olds born with fetal growth restriction (n = 32) compared with age-matched controls (n = 27), the children showed smaller total intracranial volumes but no significant differences in gray or white matter volumes individually. Diffusion tensor imaging, however, revealed elevated mean diffusivity and radial diffusivity in large white matter tracts, indicating microstructural alterations not detectable on volumetric analysis. The dissociation matters: by school age, the gross volumetric impact of growth restriction on the brain has largely closed, but the microstructural signature — which carries variance for processing speed and integrative cognitive functions — persists.

What is the underlying mechanism?

The dominant cause of intrauterine growth restriction is placental insufficiency: the placenta delivers inadequate oxygen and nutrients to the fetus, typically from a combination of maternal vascular factors (preeclampsia, hypertension, chronic disease), placental structural abnormalities (small placental size, infarction), or both. The consequences for fetal brain development unfold along several pathways.

Brain-sparing redistribution. When fetal oxygen supply falls below threshold, the fetal circulation redistributes toward vital organs — brain, heart, adrenals — at the expense of trunk and limbs. The “brain-sparing” mechanism, detectable on Doppler as increased middle cerebral artery flow, is protective in the short term but is not fully sparing in the structural sense. Sustained brain-sparing is associated with altered cortical development, suggesting that the redistribution maintains tissue mass at the cost of organizational quality.

Chronic mild hypoxia. Even when overt asphyxia is avoided, chronic mild hypoxemia disrupts the energy-intensive processes of late-fetal brain development: neuronal proliferation, migration, dendritic arborization, and synaptogenesis. Selective vulnerability of oligodendrocyte precursors during the third trimester explains why white-matter microstructure is a sensitive indicator of growth-restriction history.

Neuroinflammation and oxidative stress. Placental insufficiency activates fetal inflammatory pathways, and oxidative stress is a documented contributor to neuronal injury in growth-restricted fetuses. These pathways converge on the same vulnerable cell populations as hypoxia and likely interact with them.

Postnatal nutritional and environmental factors. Postnatal catch-up growth and feeding patterns modulate downstream development. Aggressive postnatal nutrition can produce somatic catch-up but does not consistently translate into cognitive catch-up — suggesting that the prenatal injury sets a developmental trajectory that is partially fixed by birth.

The meta-analytic picture

Sacchi and colleagues (2020) conducted the largest systematic review and meta-analysis on this question, pooling 60 studies with 52,822 children across diverse cohorts. The pooled analysis found a small-to-moderate cognitive disadvantage in SGA/IUGR children versus AGA peers (SMD around -0.5 to -0.6 SD on most cognitive composites), with the effect:

• Present in both preterm-SGA and term-SGA subgroups.
• Larger in early-onset growth restriction than late-onset.
• Largely independent of perinatal complications when these were measured separately.
• Persistent through school-age follow-up in studies that extended that long.

The meta-analytic effect is roughly half the magnitude of the prematurity effect (compare with the SMD around -0.86 from extremely preterm cohorts) but, unlike prematurity, applies to a substantially larger population — most SGA newborns are not preterm.

What does this mean for parents and clinicians?

The actionable points fall into four areas.

SGA status warrants developmental surveillance, not parental alarm. The 8-point IQ gap is real but small at the individual level — most SGA children have cognitive ability well within normal range, and the population-level effect does not predict individual trajectory. Routine developmental surveillance through pediatric well-child visits will identify the subset that needs evaluation.

The cognitive profile is specific. Educators and parents should be alert to relative weakness in visuospatial reasoning and processing speed, which can be missed when language-based screening is the only assessment used. A child who reads well but struggles with timed math or visual-pattern tasks may have an SGA-typical profile that benefits from targeted support.

Socioeconomic factors compound, not substitute. The Eves 2020 finding that SGA and SES each contribute independently means that families and communities can meaningfully change trajectory through environmental support — quality early childhood education, language-rich home environments, responsive caregiving — even though SGA-related deficits cannot be fully reversed.

Internalizing symptoms deserve attention. The Vitale 2026 finding of elevated internalizing scores aligns with broader literature suggesting that SGA-born children may be at higher risk for anxiety and depression in childhood and adolescence. Mental-health screening should be part of long-term follow-up.

What remains uncertain

Several questions are not yet resolved. Whether postnatal interventions — early intervention programs, specific nutritional protocols, or targeted cognitive training — can measurably narrow the IQ gap in SGA children is not established. Most reported intervention trials have been small, short, or methodologically limited. The role of growth hormone therapy in cognitive outcomes is similarly unsettled, with the recent Vitale 2026 finding of no GH-cognition association adding to a literature that has produced mixed signals across decades.

The boundary between constitutional small size (genetically programmed, no underlying pathology) and pathological growth restriction (placental insufficiency, infection, genetic syndromes) is also not always clean in the clinical record. Some cohorts that label all SGA infants together may dilute or mask effects that are specific to the pathological-growth-restriction subgroup. More refined classification — using fetal Doppler abnormalities, placental pathology, and growth velocity rather than birthweight alone — produces stronger and more specific cognitive associations than birthweight-cutoff studies.

The takeaway

Children born small for gestational age show a stable 8-point IQ gap that persists from infancy into adulthood, with a specific profile of weakness in perceptual reasoning and processing speed alongside preserved verbal and working-memory abilities. The deficit is most pronounced in early-onset, placental-insufficiency-driven growth restriction, has structural correlates in fetal and childhood brain MRI, and is independent of the obvious confounders of prematurity and socioeconomic status. The cognitive trajectory is partially fixed by birth, but the practical levers — developmental surveillance, environmental enrichment, targeted educational support, and attention to internalizing mental-health symptoms — meaningfully shape downstream outcomes.