Twins, birth weight, cognition, and handedness

A comprehensive study by National Institute of Mental Health (NIMH) researchers has tied prenatal growth to postnatal brain maturation in 85 monozygotic (MZ, identical) and 54 dizygotic (DZ, fraternal) twin pairs, followed from ages 5 to 25 y (1). MRI scans and intelligence quotient (IQ) data were obtained for the twins and matched nontwins, although age at intellectual assessment was not given. Among the many important findings were that (i) subtle differences in prenatal growth (based on normal birth weight) were linked to brain surface area alterations, (ii) these brain effects seemed to impact late-maturing cortices associated with higher cognitive functioning, and (iii) greater birth weight within MZ twin pairs was associated with superior postnatal cognition.

However, a key finding that was neither noted nor addressed was that both twin groups scored significantly below the nontwins in IQ, with MZ twins scoring lowest: (MZ: 109, SD = 12.4; DZ: 112, SD = 12.2; nontwins: 116, SD = 12.7). This finding is contrary to results from recent twin studies showing that adolescent and adult twins do not score below nontwins on general intelligence tests (2, 3).

What might account, at least in part, for this discrepancy? A study of 8-y-old MZ twins compared the IQ scores of pairs in which one twin was right-handed and one twin was left-handed (4). (Approximately 25% of MZ twin pairs show opposite-handedness.) These pairs were further divided according to whether the left-handed twin had a lower or higher birth weight than the right-handed twin partner (cotwin). It was reasoned that because prenatal pathology is associated with left-handedness, then IQ scores should be reduced in MZ pairs in which left-handed twins weigh less at birth than their right-handed cotwins. Conversely, IQ scores should be higher in MZ twins when left-handed twins weigh more at birth than their right-handed cotwins, possibly reflecting delayed splitting of the fertilized egg and a relative absence of prenatal pathology. These expectations were confirmed, although the twins were not selected for minimal birth weight differences as in the NIMH study. Nevertheless, organizing the NIMH MZ twins by handedness might clarify the IQ score ordering across the MZ, DZ, and nontwin groups, extending the reach of the many important findings. Handedness data were presented in an earlier NIMH article on age- and sex-related differences in brain development (5). However, in that study the MZ and DZ twins were combined into a single sample, with handedness data provided for individual twins, not for twin pairs, preventing detailed analyses of prenatal factors affecting intelligence.

The authors correctly noted that birth weight does not fully reflect or predict pre- and postnatal brain growth complexities. Events beyond those tied to birth weight may, therefore, contribute to the twins’ lower IQ scores in that sample. Experiential factors, for example some twins’ suboptimal language learning environments (reflecting divided parental attention), might also explain the twins’ lower IQ scores, especially if testing had occurred in childhood rather than in adulthood. In summary, findings from the present article underline the importance of examining associations among prenatal influences, experiential factors, and cognitive functioning.