Nancy L. Segal Ph.D.


Twin Injuries -- Genes or Environments?

Are our chances of injury partly in our genes?

Posted Mar 09, 2011

Twin children may or may not be more likely than non-twins to be affected by unintentional injuries. However, the possibility of genetic influence on accident-proneness is reasonable, given the demonstrated genetic components underlying physical characteristics and personality traits. 

An early twin study by Matheny (1986) reported a relationship between children’s temperamental patterns, such as activity level and attentiveness, and injury liability. Subsequent work by the same author supported the finding that children’s psychological characteristics are associated with accident proneness, but also showed that this association was affected by the sex of the child and features of the rearing parents and home environment (Matheny, 1987). For example, greater childhood adaptability, eating and sleeping regularity, attentiveness and positivity were associated with lower injury liability. Boys were also more likely than girls to receive medical attention. Mothers who described themselves as more emotionally stable, energetic, socially engaged or reflective had children who showed a lower risk for childhood injuries.  The homes of these children were less noisy, showed less confusion and included more ample child development resources. The contribution of child characteristics to injury liability was higher in a cohort of children followed from age six to nine years, relative to a cohort of children followed from one to three years. This result is a likely reflection of the somewhat greater freedom and mobility of older children.

A later study by Phillips and Matheny (1995) similarly examined the occurrence of accidents and injuries in a cohort of twins between birth and three years of age.  Twin correlations in injury liability showed genetic dominance effects in the absence of additive variance. The investigators noted, however, that such results could only be explained with reference to epistatic or other interaction effects. As before, injuries were more frequent among males than females, although this result applied only to twins whose parents responded to detailed questioning about their children. Activity and impulsivity were the strongest childhood predictors of accidental injury.

In a more recent study, Ordonana, Caspi and Moffitt (2008) argued that genetic effects on accidental injury in childhood suggest the presence of “intermediate phenotypes between genes and risk of injury (e.g., temperamental or behavioral characteristics under genetic influence).” They noted that a great deal about the interaction of these factors and how they may eventuate in accidents and other mishaps is unknown. As such, they conducted an extensive study of the problem, using 1,027 same-sex twin pairs assessed within two months of their fifth birthday. The twins were enrolled in the Environmental Risk (E-risk) Longitudinal Twin Study and comprised two cohorts of twins born in England and Wales in 1994 and 1995.  Zygosity was assigned on the basis of a physical resemblance questionnaire, with unclear cases classified by DNA analysis.

The twins’ mothers were interviewed twice such that data were gathered for each co-twin on separate occasions; thus, the information was obtained retrospectively.  Specifically, parents were asked to indicate the occurrence of any accident or injury on the part of each twin child that required medical attention.  The socioeconomic status of each family, assigned with reference to income, place of residence and other factors, was examined with reference to childhood injuries.

The final sample of MZ and DZ twins showed an equal distribution of injuries: None: 64.3% of MZ twins and 65.4% of DZ twins; One: 26.7% of MZ twins and 24.3% of DZ twins; and Two or More: 9.1% of MZ twins and 10.3% of DZ twins. The MZ twin correlation of .14 for one or more injuries was twice as large as the corresponding DZ twin correlation (.06).  However, effects specific to the child played a greater role as show by the difference between 1.0 (all sources of influence) and the MZ correlation of .14 (shared genes and shared environments) or .86.  In other words, having an injury within the first five years of life was only modestly associated with genetic factors, whereas specific environmental effects played the biggest role.  Surprisingly, a genetic effect on having two or more injuries did not show genetic influence. This is surprising because children who are hardest to control might be expected to show the greatest risk for injury. However, attentive parents might mitigate this effect by introducing more constant supervision or other safeguards.

As in the earlier twin studies, being male posed the greatest risk for accidental childhood injuries, regardless of frequency, and child-specific environmental risk factors made greater contributions to injury risk than genetic factors. Other important contributions came from the younger age of the mother and the presence of childhood externalizing difficulties. The authors asserted that children’s behavioral styles should be considered in this area, in particular with respect to how they interact with features of the environment.

Ordonana and colleagues noted some important limitations of their work, namely that the data were gathered retrospectively and that twins from opposite-sex twin pairs were not included. Several ideas come to mind in this regard. First, following twins longitudinally might capture the nature and source of their accidents more effectively than parental recall. Second, it may be that twins are more likely to suffer accidental injury than nontwins. This is because rearing two children is often more difficult than raising one child because caregiver attention becomes divided. Furthermore, parents and others may leave twins on their own to a greater degree, assuming they are safe and happy in each other’s company.  This may allow twin children (especially MZ twins) to collaborate in unsafe ways. Alternatively, in the event that one twin is harmed, the other child may be available to alert their parents or to seek other assistance. Research shows that twin children are more likely to be abused than nontwin children (see Segal, 2000), but no one has, to my knowledge, conducted a comparable analysis of accidental injuries. 

An interesting and important finding provided by all the studies reviewed is that being male poses a greater injury risk than being female. It would be of interest to know if females from opposite-sex twin pairs have a mitigating influence on their twin brothers in this regard, or if twin males are more likely to succeed in engaging their sisters in risk-taking behaviors. 

This article initially appeared in the Journal TWIN RESEARCH AND HUMAN GENETICS.

About the Author

Nancy L. Segal, Ph.D., is a professor of psychology and the Director of the Twin Studies Center, at California State University, Fullerton.

More Posts