Jacquelyn J. Gillis

Attention deficit disorder in reading-disabled twins: Evidence for a genetic etiology

In order to assess the genetic etiology of attention deficit hyperactivity disorder (ADHD), the basic regression model for the analysis of selected twin data (DeFries & Fulker, 1985, 1988) was fitted to questionnaire data (DICA: Diagnostic Interview for Children and Adolescents; Herjanic, Campbell, & Reich, 1982) for 37 identical and 37 fraternal twin pairs tested in the Colorado Reading Project. Results of this analysis suggest that ADHD is highly heritable. Moreover, adjusting DICA scores for either IQ or reading performance differences did not substantially change parameter estimates. In future analyses of larger data sets, distinguishing between possible subtypes of attentional problems (e.g., ADD with or without hyperactivity) may facilitate tests of more searching etiological questions.

Confirmatory Factor Analysis of Word Recognition and Process Measures in the Colorado Reading Project

Measures of word recognition (REC) and two component skills, phonological coding (PHON) and orthographic coding (ORTH), were subjected to multivariate behavioral genetic analysis. Data were obtained from a sample of identical and fraternal twin pairs wherein at least one member of each pair was reading disabled (RD), and from a sample of twins wherein both members of each pair read in the normal range. Confirmatory factor analysis was used to fit the genetic, common environmental, and specific environmental covariance components for REC, PHON, and ORTH within the RD and normal simples. The resulting heritability estimates for REC, PHON, and ORTH were 0.59, 0.41, and 0.05 in the RD sample, and 0.35, 0.52, and 0.20 in the normal sample. After dropping the nonsignificant common environment parameters from the models, the genetic correlations between REC and PHON and between REC and ORTH were respectively 0.81 and 0.45 in the RD sample, and 0.68 and 0.45 in the normal sample. Differences between the genetic correlations were significant in the RD sample (p<0.005), marginally significant in the normal sample (p<0.10), and highly significant in the combined sample (p<0.001), indicating that genetic influences on individual differences in REC are more strongly related to genetic variance in PHON than in ORTH. These results are consistent with previous demonstrations of substantial genetic covariance between the disabled groups deficits in REC and PHON, but not between REC and ORTH (Olson et al., 1989; Olson and Rack, 1990).

Genetic Etiology of Spelling Deficits in the Colorado and London Twin Studies of Reading Disability

The basic multiple regression model for the analysis of selected twin data (DeFries and Fulker 1985, 1988) was fitted to spelling data from 100 pairs of MZ twins and 71 pairs of same-sex DZ twins tested in the Colorado Reading Project (DeFries, Olson, Pennington and Smith 1991), and to data from 12 pairs of MZ twins and 15 pairs of same-sex DZ twins tested in the London twin study of reading disability (Stevenson, Graham, Fredman and McLoughlin 1984, 1987). Estimates of hg2 obtained from analyses of these data suggest that about 60% of the deficit of probands is due to heritable influences in both samples. When a regression model was fitted separately to data from males and females in the combined Colorado and London samples, resulting estimates of hg2 were 0.66±0.18 and 0.56±0.19, respectively, a nonsignificant difference. Collaborative analyses of data from additional twin studies of reading disability would facilitate more rigorous tests of hypotheses of differential genetic etiology as a function of group membership.

Differential Genetic Aetiology of Reading Disability as a Function of Age

In order to test the hypothesis that the genetic aetiology of reading disability differs as a function of age (Stevenson et al., 1987). composite reading performance data from 99 pairs of identical twins and 73 pairs of same-sex fraternal twins in which at least one member of each pair was reading disabled were subjected to multiple regression analysis. A basic regression model that predicted a cotwins score from the probands score and the coefficient of relationship was fitted separately to data from younger (8.3 – 11.5 years) and older (11.5 – 20.2 years) twin pairs. Resulting estimates of h2g (an index of the extent to which the deficit of pro bands is due to heritable influences) were .60 ± .20 and .38 ± .19, respectively. Although these results are consistent with the hypOthesis that genetic factors may be less important as a cause of reading problems in older children. the difference between the two estimates was not significant. When the basic model was fitted to the data from twins in three diff...

Confirmatory Factor Analysis of Reading and Mathematics Performance: A Twin Study

Reading and mathematics performance data from a sample of 264 reading-disabled twin pairs and 182 matched control twin pairs were subjected to multivariate behavior genetic analysis. The factor structure of reading and math performance measures was found to be highly similar for both groups. Consistent with previous findings obtained using alternative methods, a significant heritable component to individual differences in reading performance was found both within the reading-disabled (h2 = 0.78) and control (h2 = 0.74) twin samples. In addition, a substantial genetic influence on mathematics performance was found (h2 = 0.51 and 0.60 in the reading-disabled and control samples, respectively), although shared environmental influences common to both members of a twin pair also contribute significantly to the variance in math scores of both groups (c2 = 0.44 and 0.37). Moreover, genetic influences accounted for 98% of the observed correlation between reading and math performance within the sample of reading-disabled twin pairs, and for 55% of the observed correlation in the control sample. Thus, individual differences in both reading and mathematics performance are highly heritable and appear to be caused by many of the same genetic influences.

Validity of school history as a diagnostic criterion for reading disability

In order to assess the validity of school history information as a diagnostic criterion for reading disability, reading performance data from 304 children with a positive history for reading problems were compared to those from 319 children with a negative school history. These children are members of twin pairs who are participants in the Colorado Reading Project (DeFries 1985). A multivariate analysis of variance of Reading Recognition, Reading Comprehension, and Spelling subtest scores from the Peabody Individual Achievement Test (Dunn and Markwardt 1970) yielded a highly significant (p <0.001) difference between groups for both the multivariate and univariate comparisons. The difference between the unweighted mean scores of the school-history positive and negative groups for Reading Recognition was 2.1 standard deviations. Moreover, when discriminant weights estimated from an independent sample of 140 reading-disabled and 140 control nontwin children were used to reclassify the sample of twins, 75% of those with a positive history for reading problems were classified as being reading disabled and 95% of those with a negative history were classified as unaffected. Thus, results of this study establish the validity of school history information and suggest that such information may have considerable utility for ascertaining samples of reading-disabled children.

Gender differences in cognitive abilities of reading-disabled twins.

Gender differences were assessed in three research-identified samples of children who were members of twin pairs: (1) 120 male and 124 female probands from same-sex identical and fraternal twin pairs in which at least one member of each pair is reading disabled; (2) a comparison sample of 148 males and 161 females from same-sex twin pairs with no history of reading problems; and (3) 34 pairs of opposite-sex fraternal twin pairs in which at least one member of each pair is reading disabled. Results of multivariate analyses of variance of psychometric test data from the two samples of same-sex twin pairs, in which the male and female subjects were reared in different homes, suggest that profiles of gender differences are similar in reading-disabled and control children. Moreover, this pattern of gender differences also tended to occur in opposite-sex twin pairs, who shared prenatal, as well as early postnatal, environmental influences. In general, reading-disabled males obtained higher average scores than affected females on Wechsler (1974, 1981) Verbal and Performance IQ, but lower scores on Reading Recognition and Spelling subtests of the Peabody Individual Achievement Test (Dunn and Markwardt 1970). With regard to the Wechsler subtests, reading-disabled males achieved higher average scores on Information, Picture Completion, Block Design, and Object Assembly, but lower scores on Coding. Although significant and relatively consistent gender differences in cognitive measures were found in three samples included in this study, these differences account for only a small portion of the observed variance.

Familial resemblance for measures of reading performance in families of reading-disabled and control twins

In order to assess familial resemblance for measures of reading performance, data from 314 pairs of twins in which at least one member of each pair is reading-disabled [142 monozygotic (MZ) and 172 dizygotic (DZ) twin pairs], 273 matched control pairs (131 MZ and 142 DZ pairs), and their parents were subjected to both correlation and regression analyses. Results indicate that parent-offspring resemblance in families of reading-disabled probands does not differ substantially from that in families of controls. In general, the correlations and regressions for MZ twin pairs are greater than those for DZ twins; thus, individual differences in reading performance are due at least in part to heritable influences. As expected, regression coefficients are consistently larger than correlation coefficients for both parent-offspring and proband-cotwin comparisons in the reading-disabled sample, illustrating that regression analyses are more appropriate than correlations for assessing familial resemblance in selected samples.