John DeFries is the Director of the Center, and each of the co- investigators are responsible for the direction of their specific component projects. John DeFries and David Fulker are co-investigators on a component that involves the administration of a battery of standardized psychometric tests and the development and application of advanced behavioral-genetic analyses. Richard Olson directs the development and administration of experimental measures for component reading, language, and perceptual processes that may play important roles in different reading disabilities. Bruce Pennington is exploring the relations between ADHD, executive functions, and reading disabilities. Shelley Smith is directing research on linkage analysis and physical mapping of associated genes. Richard Olson and Barbara Wise conduct studies on the computer-based remediation of reading and related deficits in phonological processing. Bruce Pennington and Pauline Filipek are conducting magnetic resonance imaging (MRI) studies of twins’ brain morphometry and relations to reading disability.

Many important results have been published in numerous articles from the Colorado Reading Project. Some of these are cited in the two review papers mentioned above. I will not attempt to summarize all this work here. Instead, I will focus primarily on some of the major results from our experimental measures of reading, spelling, and language that show significant genetic influence. I will conclude with a discussion of the implications for environmental intervention and briefly review work with Barbara Wise on the computer-based remediation of reading and related phonological disabilities.

The use of a common subject pool of identical and fraternal twins provides an important collaborative link between our component research projects. The twins are 3rd to 12th graders from 27 Colorado school districts. If one or both members of a twin pair have school records suggesting problems in reading, math, and/or ADHD, they are invited for testing in laboratories at the University of Colorado and the University of Denver. A smaller normal comparison twin sample with no school history of problems in these areas is also tested on the same measures.

Twins are studied because their behavioral data are informative about the relative balance of genetic and environmental influences. Identical or monozygotic (MZ) twins share all their genes and their home environment. Fraternal or dizygotic (DZ) twins also share their home environment, but only half of their segregating genes, on average. Thus, a greater similarity between MZ twins compared to DZ twins provides evidence for a genetic influence on the behavior studied. Behavior-genetic analysis of twin data can also separate the environmental influence that is shared by the twins from that which is not shared: Differences between MZ twins indicate non-shared environment influences and test error; the influence of shared-environment is indicated when the DZ twins’ average similarity is greater than expected from their average 50% genetic similarity.

DeFries and Fulker (1985) developed a powerful method, now referred to as DF analysis, for assessing genetic influence on a group deficit when twin pairs are selected for one or both member’s deviant position on a normally distributed dimension such as reading. DF analysis compares the average regression toward the normal population mean for MZ and DZ cotwins who do not meet the affected severity criterion. From this information, it is possible to derive estimates of the average genetic, shared environment, and non-shared environment influences on deviant group membership. DF analyses have been used to assess average genetic influence on the group deficit (below the local 10th percentile) for a composite measure of word recognition, reading comprehension, and spelling from the Peabody Individual Achievement Test (Dunn & Markwardt, 1970). The most recent analysis by DeFries and Alarcon (1996) estimated that the heritability of the group deficit in this composite measure was h2g = .56. This means that approximately half of the group deficit is due to genetic factors.

DF analyses have also been conducted for group deficits in phoneme awareness, phonological decoding, and orthographic coding (Olson, Wise, Conners, Rack, & Fulker, 1989; Olson, Forsberg, & Wise, 1994). Phoneme awareness is measured by language tasks that require the isolation and manipulation of phonemes within spoken words or nonwords. Performance in these tasks is highly correlated with reading skill, particularly with the component reading skill of phonological decoding. We measure phonological decoding through both the oral and silent reading of nonwords. Our orthographic coding measures assess subjects’ sensitivity to the precise spelling patterns for words in the comparison of a word with a homophonic nonword (e.g., rain rane) and in the choice between homophones (bear bare) to fit the meaning of a spoken sentence. All of these tasks are significantly correlated with each other and with measures of printed word recognition. Estimates of the average heritability for group deficits in these measures range from about h2g = .4 for phonological decoding accuracy and word recognition without time constraints, to about h2g = .6 for phonological decoding and word recognition that is both fluent and accurate (Gayan, Datta, Castles, & Olson, 1997). Thus, it appears that the measurement of both speed and accuracy in phonological decoding and word recognition may tap additional genetic variance for group deficits. This result is consistent with observations by Bruck (1990) that “remediated” adults, who have attained normal levels of word recognition accuracy through extensive reading practice, often have continuing deficits in reading fluency.

Olson et al. (1989) suggested that the group deficit in orthographic coding might have a lower heritability than the other component skills. We argued that accurate recognition of the precise spelling for a word would depend more on the environmental effects of exposure to that word in print. However, a trend toward lower heritability in the small sample of twins analyzed by Olson et al. (1989) was not replicated in larger samples (Olson et al., 1994; Gayan et al., 1997). In these later analyses, group accuracy deficits in the orthographic tasks have heritability levels that are about h2g = .6.

Bivariate DF analyses were employed by Olson et al. (1994) to assess the degree of common genetic influence on the shared variance among the above measures: The question was, to what degree are the group deficits in the measures due to the same or different genetic influences? The results indicated that there were significant common genetic effects on group deficits in all the measures, but there was also evidence of some independent genetic effects on each measure. Shared and independent genetic effects on individual differences in each of the measures have recently been confirmed in both normal and reading disabled groups using appropriate factor models (Gayan & Olson, in press). The partial genetic independence of phonological and orthographic skills is of particular interest to researchers who study individual differences or subtypes within the reading disabled population and who wonder about the role of genes and environment in these differences (c.f., Castles and Coltheart, 1993; Olson et al., 1985).

The DF analyses for group deficits described above do not provide evidence on the genetic influence for any individual disabled reader within the group. We could begin to make more accurate predictions about the degree of genetic influence on individual deficits if we found evidence that the level of heritability varies, depending on individual variables such as age, gender, or IQ. Fortunately, DF analyses can be extended to assess the size and statistical significance of such effects on the average level of heritability. Gender does not appear to be related to the degree of genetic influence on reading deficits (DeFries, Gillis, & Wadsworth, 1993). Nonsignificant trends in relation to age reported by Wadsworth, Gillis, DeFries, & Fulker (1989) have been further explored by DeFries, Alarcon, and Olson (1997). The latter study found trends toward decreasing heritability with age for word recognition and increasing heritabiltiy with age for spelling. The opposing direction of these trends resulted in a statistically significant interaction. In addition, there is now significant evidence for the importance of IQ in the genetic etiology of deficits in word recognition: Disabled readers with relatively high IQ scores tend to have a stronger genetic etiology than those with relatively low IQ (Olson, Forsberg, Gayan, & DeFries, in press). The concurrence of low IQ and low reading may be more likely due to some shared family environment that constrains both reading and general cognitive development. With high IQ, the reading environment may tend to be better, and reading failure more likely due to genetic constraints. We are currently conducting additional analyses to better understand this apparent relation between IQ and genetic influence on reading deficits.

Behavioral comparisons of MZ and DZ twins has yielded valuable information about the balance of genetic and environmental influence on group deficits in reading and related language skills. Further analyses of genetic influence on reading disabilities in relation to IQ and other individual characteristics will provide more specific information about this balance across individuals. However, the further specification and understanding of genetic mechanisms at the individual level will ultimately depend on the identification of specific genes that are associated with reading disability. Recent linkage analyses suggest that a gene or genes on the short arm of chromosome 6, close to the HLA region, may account for a significant proportion of reading disabilities. Preliminary evidence for this linkage was first obtained by Smith, Kimberling, & Pennington (1991) using data from 19 extended families with a history of reading problems. Cardon, Smith, Fulker, Kimberling, Pennington, & DeFries (1994) applied more powerful linkage analyses to these family data and added a sample of 46 DZ twin families from the Colorado Reading Project. Taken together, the extended family and twin data provided highly significant evidence for a genetic linkage to reading disability near the HLA region of chromosome 6.

Reading disability in the above linkage studies was ascertained through a composite measure of word recognition, reading comprehension, and spelling from the Peabody Individual Achievement Test (Dunn & Markwardt, 1970). Gayan et al. (1995) used the same analytic methods and DZ twin sample as Cardon et al. (1994) to look for linkage to deficits in the specific skills of phoneme awareness, phonological decoding, orthographic coding, and fluent word recognition. These results and recent unpublished analyses of a new DZ twin and sibling sample suggest that deficits in phoneme awareness, phonological decoding, and orthographic coding show strong evidence for linkage to genetic markers in the HLA region of chromosome 6. The evidence was less strong for two measures of word recognition. A similar pattern of results for phoneme awareness and word recognition was recently reported by Grigorenko et al. (1997). They used data from extended families with a history of reading disabilities and found strong linkage for deficits in phoneme awareness to the same HLA region of chromosome 6. Linkage appeared to be weaker in this region for deficits in word recognition, which were more strongly linked to a region on chromosome 15. The possibility of differential genetic linkages for different component skills in reading and language is intriguing, but a much larger subject sample is needed to test the statistical significance of these differences. This additional data is now being collected by our Center, with additional genetic markers including other regions of the genome.

A strong genetic linkage for deficits in phoneme awareness has now been confirmed in two independent laboratories for the same region of chromosome 6. The next step is to more precisely specify the gene or genes’ location(s) and ultimately clone the gene(s), identify the coded protein(s), and determine the influence of the protein(s) on the developing nervous system and related behavior. Much more research will be needed to reach the latter goal, but recent advances in methods for locating genes may soon allow us to identify individuals who have a gene or genes that place them at risk for reading disability. Early information about a genetic risk could be used to provide additional support in a child’s early language and reading environment. Providing support prior to school entry could help avoid the stigmatizing effect of school failure in reading.

It should be emphasized that having a genetic risk does not imply that a reading disability is inevitable. All genetic effects are mediated by the environment. For example, phenylketonuria is a genetic disorder that can lead to severe mental retardation, but restricting the child’s diet to reduce the ingestion of an amino acid called phenylalanine can substantially reduce or eliminate the deleterious effects of this genetic disorder. There is no evidence that such a simple dietary control could reduce a specific genetic influence on reading disability, but the readers of Perspectives are well aware that other forms of environmental intervention, such as the Orton-Gillingham program, can have a significant impact.

Since 1986, we have been exploring the use of talking computers in the schools to support 2nd to 5th grade children with reading disabilities in their word decoding while reading stories. The programs allow children with reading disabilities to read interesting stories on the computer that are more appropriate for their age level, and independently obtain spoken decoding support by targeting difficult words with a mouse. More recently, we have incorporated additional programs designed to improve disabled readers’ phoneme awareness and phonological decoding (Wise and Olson, 1995). This research will be reviewed in greater detail by Barbara Wise in a subsequent issue of Perspectives. The main point to be made here is that these and other programs can substantially improve disabled readers’ phoneme awareness and phonological decoding, skills that are critical for reading development and that have a very strong genetic influence on their group deficits. It is clear that the improvement of these and other reading-related skills in children with reading disabilities often requires extraordinary environmental support. Computer programs incorporating both synthetic and digitized speech can efficiently provide much of this support.

The varied and convergent research perspectives on reading disability in the Colorado Learning Disabilities Research Center and the other NICHD- sponsored centers have resulted in some major advances over the past decade. As this work continues, we look forward to learning much more about the causes and optimal treatment of different reading disabilities. We are most grateful for the continuing support of the NICHD and the Orton Dyslexia Society in this endeavor.

References

Bruck, M. (1990). Word-recognition skills of adults with childhood diagnoses of dyslexia. Developmental Psychology, 26, 439-454.

Cardon, L. R., Smith, S., Fulker, D., Kimberling, W., Pennington, B. & DeFries, J. (1994). Quantitative trait locus for reading disability on chromosome 6. Science, 266, 276-279.

Castles, A., & Coltheart, M. (1993). Varieties of developmental dyslexia. Cognition, 47, 149-180.

DeFries, J. C., & Alarcon, M. (1996). Genetics of specific reading disability. Mental Retardation and Developmental Disabilities Research Reviews, 2, 39-47.

DeFries, J. C., Alarcon, M., & Olson, R. K. (1997). Genetics and dyslexia: Developmental differences in the etiologies of reading and spelling deficits. in C. Hulme & M. Snowling (Eds.), Dyslexia: Biology, cognition, and intervention (pp. 20-37. London: Whurr Publishers Ltd.

Defries, J. C., Filipek, P. A., Fulker, D. W., Olson, R. K., Pennington, B. F., Smith, S. D., & Wise, B. W. (1997). Colorado Learning Disabilities Research Center. Learning Disability Quarterly, 8, 7- 19.

DeFries, J. C., & Fulker, D. W. (1985). Multiple regression analysis of twin data. Behavior Genetics, 15, 467-473.

DeFries, J.C., Gillis, J.J., & Wadsworth, S.J. (1993). Genes and genders: A twin study of reading disability. In A.M. Galaburda (Ed.), Dyslexia and development: Neurobiological aspects of extra- ordinary brains (pp. 187-204).

DeFries, J.C., Olson, R.K., Pennington, B.F., & Smith, S.D. (1991). Colorado reading project: An update. In D.D. Duane & D.B. Gray (Eds.), The reading brain: The biological basis of dyslexia (pp. 53-87). Parkton, MD: York Press.

Dunn, L. M., & Markwardt, F. C. (1970). Examiner's manual: Peabody Individual Achievement Test, Circle Pines, MN: American Guidance Service.

Gayan, J., Datta, H. E., Castles, A., & Olson, R. K. (1997). The etiology of group deficits in word decoding across levels of phonological decoding and orthographic coding. Paper presented at the meeting of the Society for the Scientific Study of Reading, Chicago, 3/23/97.

Gayan, J., & Olson, R. K. (in press). Common and specific genetic effects on reading measures. Behavioral Genetics.

Gayan, J., Olson, R.K., Cardon, L.R., Smith, S.D., Fulker, D.W., Kimberling, W.J., Pennington, B.F., & DeFries, J.C. (1995). Quantitative trait locus for different measure of reading disability. Behavior Genetics, 25, 266.

Grigorenko, E. L., Wood, F. B., Meyer, M. S., Hart, L. A., Speed, W. C., Shuster, B. S., & Pauls, D. L. (1997). Susceptibility loci for distinct components of developmental dyslexia on chromosomes 6 and 15. American Journal of Human Genetics, 60, 27-39.

Olson, R. K., Forsberg, H., Gayan, J., & DeFries, J. C. (in press). A behavioral-genetic analysis of reading disabilities and component processes. In R. M. Klein & P. A. McMullen (Eds.), Converging methods for understanding reading and dyslexia. Cambridge, MA: MIT Press.

Olson, R. K., Forsberg, H., & Wise, B. (1994). Genes, environment, and the development of orthographic skills. In V.W. Berninger (Ed.), The varieties of orthographic knowledge I: Theoretical and developmental issues (pp. 27-71). Dordrecht, The Netherlands: Kluwer Academic Publishers.

Olson, R.K., Kliegl, R., Davidson, B.J., & Foltz, G. (1985). Individual and developmental differences in reading disability. In G.E. MacKinnon and T.G. Waller (Eds.), Reading research: Advances in theory and practice, Vol. 4. New York: Academic Press. pp. 1-64

Olson, R. K., Wise, B., Conners, F., Rack, J., and Fulker, D. (1989). Specific deficits in component reading and language skills: Genetic and environmental influences. Journal of Learning Disabilities, 22, 6, 339- 348.

Smith, S. D., Pennington, B. F., & Kimberling, W. J. (1991). Screening for multiple genes influencing dyslexia. Reading and Writing: An Interdisciplinary Journal, 3, 285-298.

Wadsworth, S.J., Gillis, J.J., DeFries, J.C., & Fulker, D.W. (1989). Differential genetic aetiology of reading disability as a function of age. Irish Journal of Psychology, 10, 509-520.

Wise, B. W., & Olson, R. K. (1995). Computer-based phonological awareness and reading instruction. Annals of Dyslexia, 45, 99- 122.