Revisiting sources of left-handedness in multiple-birth individuals

Multiple births have been celebrated by many couples, yet they pose moderate to severe physical hazards to both mothers and infants, one being infants’ lower birthweight relative to nontwins. Low birthweight, often linked to premature birth, is a recognized risk factor for a range of developmental delays known to be more prevalent among twins and other multiples than singletons (1). The extent to which birthweight is associated with left-handedness and other developmental sequelae in triplets from Japan and The Netherlands is the focus of a recent study by Heikkilä et al. (2). Given the long-standing debates surrounding factors affecting handedness in multiple-birth individuals, as well as the expanding multiple-birth population, a fresh look at the origins and implications of left-handedness in triplets is both welcome and timely.

The availability of extensive twin and triplet registries in Japan and The Netherlands enabled the researchers to complete the largest study to date addressing associations between birthweight and left-handedness in multiple-birth individuals. Triplets in Japan were additionally recruited from mothers of twins organizations and referrals from public health professionals. In both samples mothers completed mailed questionnaires requesting information about their children’s handedness (right, left, or ambidextrous), although some triplets provided this information at age 14 y. Note that ambidextrous individuals were excluded from the analyses, although in some previous studies they have been grouped together with left-handers (3). The writing hand is a reliable index of hand preference, although the reliance on self-reporting and the inability to assess the extent to which degree of handedness is a meaningful factor in early and later development are possible caveats recognized by the authors. For example, a recent book presented the life histories and research findings on doubly switched-at-birth monozygotic (MZ) male twins from Colombia, South America, among them results from a standard handedness questionnaire. It was found that while all four twins were right-handed, two of the twins (from different pairs) were close to the range of ambidexterity, especially one with an extremely low birthweight (4). Parents of the Japanese and Dutch triplets also indicated the ages at which key motor milestone measures, such as sitting and standing, were reached. Infant head circumference was available for the Japanese triplets only.

Factors explaining hand preference in multiple-birth individuals have long been the focus of scientific inquiry. It was believed for many years that delayed division of the fertilized egg, either at or after 8 d postconception, created “mirror-imaging effects” in some traits, such as hand preference, dermatogylphic features, and hair whorl. Delayed division was used to explain the well-documented finding that ∼25% of MZ (or identical) twins are opposite-handed (5). That belief is still held by some, although it has been seriously questioned in recent years, given that the presumed index of the time of zygotic splitting (i.e., the number of placenta and fetal membranes) has not been conclusively confirmed (Fig. 1). Furthermore, this event would not account for the fact that left-handedness occurs with equal frequency in MZ and DZ (dizygotic or fraternal) twin pairs; DZ twins originate from the separate fertilization of two ova by two separate sperm, so their biological origins are quite distinct from those of MZ twins. Indeed, Heikkilä et al. (2) do not find differences in the frequency of right- and left-handedness between zygosity groups. It is reasonable to suppose that left-handedness in triplets and other multiples is variously linked to adverse birth effects, genetic factors, and/or unusual laterality determination.

Fig. 1.

The traditional models of twinning. All four figures show MZ twins, but with different arrangements of the placentae and fetal membranes. The first two figures also apply to DZ twins, explaining why zygosity can be misdiagnosed. The cords of monochorionic–monoamniotic twins are typically entangled. Illustrations © Levent Efe (artist).

A number of important and interesting findings emerged from this analysis. The prevalence of left-handedness was lower in the Japanese sample than in the Dutch sample. This difference could reflect differences in cultural views, specifically unfavorable attitudes toward left-handedness in mostly rural areas of Japan; pressure toward right-handedness has recently relaxed in urban areas (6, 7). It is also possible that differences in methodology are responsible for this finding. In addition, the frequency of left-handedness among the Japanese males and females did not differ and was actually slightly higher among the females. In contrast, most studies find a significant excess of left-handedness among males, as did the Dutch study. This sex difference may reflect the greater vulnerability of male infants to adverse birth events, although the precise mechanisms by which handedness may “shift” remain uncertain, as the authors note.

Among the most provocative findings was that birthweight was significantly lower among the left-handed triplets than the right-handed triplets. This difference was observed in both populations and remained stable when the analyses were restricted to individuals older than 5 y of age. Also of particular interest was the association between birthweight and handedness even after controlling for birth order, sex, maternal age, and infertility treatment. Birth order was not related to handedness, a somewhat surprising finding, given that naturally delivered second-born twins typically experience a more difficult birth than first-born twins. This perinatal effect is due largely to the decreased size of the uterus following delivery of the first-born infant. Consequently, second-born twins are more likely to experience respiratory distress, neonatal trauma, and infection and are at greater risk for newborn death than their first-born cotwins. Greater intervals between births also predict less favorable outcomes for second-born twins (8). Birth order does, however, have very different implications for multiple-birth infants delivered by cesarean section and it is unclear if that distinction was made in the Heikkilä et al. (2) study.

Gestational age was not associated with left-handedness in the Japanese sample, an unexpected finding because of the link between prematurity and low birthweight. Birthweight was, however, a significant predictor of handedness. In the Dutch sample gestational age and birthweight were associated with handedness, but only when considered separately. Interestingly, controlling for gestational age demonstrated that left-handedness is associated with lower birthweight, a result that may be one of the most important and far-reaching effects to emerge from this analysis. First, birthweight emerged as the most salient factor implicated in preference for the left hand. Second, the cooccurrence of low birthweight and other birth difficulties in multiples is well-established, but the relationships among left-handedness, birthweight, and developmental delays are not clearly defined.

Uncertainties surrounding the implications of birthweight for other developmental traits is evident in the comparative analysis of motor and head development in the left- and right-handed triplets. Delays in motor development were observed in both the Japanese and Dutch samples, although to somewhat different degrees. Earlier motor development also occurred in conjunction with higher birthweight; however, these differences were not maintained when controlling for birthweight. A smaller infant head circumference was also detected among left-handers than among right-handers in the Japanese sample. A positive association between birthweight and head circumference was noted, but handedness was unrelated to head circumference when birthweight was controlled.

While it is beyond the scope of this paper, in the future it might be of interest to examine familial handedness. Some studies find that parents of left-handed twins are more often left-handed than their own siblings who are not parents of twins (9). Increased left-handedness among the parents of these triplets might be consistent with a familial component, such that developmental delays might occur less frequently. Several questions also come to mind. A previous study looked at twins as a function of whether or not the left- or right-handed MZ twin had the lower birthweight. It was found that lower-birthweight left-handers had lower intelligence quotient (IQ) scores than their right-handed cotwins, suggesting trauma. However, when left-handers had the higher birthweight they had higher IQs than their right-handed cotwin, suggesting unusual asymmetry formation (10). Heikkilä et al. (2) attempted to organize their triplets in this manner with respect to motor development, but they lacked a sufficient number of opposite-handed MZ pairs. Future studies by these investigators and others might benefit from such analyses.

The science behind the fascination of twins, triplets, and other higher-order multiples has progressed rapidly in recent years, due largely to increases in twinning rates and advances in molecular genetic techniques, as suggested above. Among Western nations ∼1/33 individuals are twins, compared with 1/60 individuals in 1980 (11). This dramatic rise is explained by the proliferation of assisted reproductive technologies (ART), such as in vitro fertilization, and delays in maternal age at conception as women establish themselves professionally (12). Both of these changes are responsible for increases in DZ twinning, whereas ART has been additionally linked to increases in MZ twinning, albeit to a lesser degree (5, 13). The Heikkilä et al. (2) study clearly underlines the need for the better management of multiple-birth pregnancies, with a view toward extending gestational age and/or identifying more effective postnatal caretaking regimens for these newborns. The fluid-filled biobag, a device designed to offset prematurity in newborn lambs, has produced promising results and is under consideration for use with humans in several years (14). Until then it is incumbent upon medical researchers and health professionals to consider ways to assist the development of low-birthweight newborns. The Heikkilä et al. (2) study, by highlighting the developmental significance of low birthweight, is a key step in the right direction.