David W. Fulker - US grants

Although it is widely believed that individual differences among infants are not predictive of cognitive ability later in life, this conclusion is based on studies that used infant tests heavily weighted with sensorimotor items. In the proposed 3-year study, a novel design will be employed to assess the extent to which nonmotoric measures at 5 and 7 months of age can predict adult IQ: Correlating midchild scores (average of twin children) with midparent IQ (parental average) provides an "instant" facsimile of a longitudinal study from infancy to adulthood. The proposed research will include 450 infants (200 pairs of twins and 50 singletons) and their 500 parents. The infant measures to be used include some measures that have been suggested as precursors of later IQ: orienting responses; Fagan's measure of preference for visual novelty; strength of hand preference; and verbal and test-taking behaviors. In addition, we shall test the efficacy of infant visual reaction time and a delayed memory version of the Fagan measure in predicting adult IQ. Parents will be tested on a battery of measures that includes the WAIS-R, measures of specific cognitive abilities, and a measure of reaction time as a function of increasing bits of information. In addition to developing a brief, reliable, and easily administered battery of early infancy measures that predicts adult IQ, the proposed research will test two general hypotheses: (1) infant precursors of adult cognitive abilities predict adult general cognitive ability (g, or IQ) rather than specific cognitive abilities; and (2) unlike extant infant tests, infant measures that predict adult IQ show substantial genetic influence, are genetically intercorrelated (as shown by twin cross-correlations), and show genetic continuity from infancy to adulthood (as estimated from our model of midparent/midtwin relationships). The practical significance of this research lies in the diagnostic and therapeutic benefit that will ensue from the ability to predict adult cognitive ability from measures obtained in infancy. Its theoretical significance lies in its re-evaluation of the nature of infant intelligence.

Little is known about the etiology of individual differences in psychological development during middle childhood, an era marked by changes in cognitive and social behavior nearly as dramatic as those seen during the transition from infancy to early childhood or during adolescence. The proposed research will apply the most powerful quantitative (biometrical) genetic methodology, the "full" adoption design in which data are collected from both birth and adoptive parents, to the study of psychological development in 7-year-olds. The sample is unique: over 400 adopted and matched nonadopted children who previously have been studied in their homes at 1, 2, 3, and 4 years of age using an extensive set of psychological and environmental assessments. The birth parents and adoptive parents of the adoptees and the parents of the nonadopted children have been administered a 3-hour battery of psychological measures. The proposed 5-year project will provide for testing the 7-year-old adopted and nonadopted children on a multidimensional battery of psychological measures including general and specific cognitive abilities, communicative skills, school achievement, temperament, behavioral problems, motoric development, and social interaction. The project will capitalize on the potent design and the extensive data set previously collected as part of the Colorado Adoption Project in applying multivariate, longitudinal, and quatitative genetic analyses to data obtained during the important developmental period marked by the beginning of formal schooling. The results of these analyses, which will substantially advance our understanding of the etiology of individual differences in functioning during this critical epoch, are likely to have important implications for childrearing, education, and mental health.

The goal of the proposed research is to continue the rapid screening of infant cognitive measures, obtained at 9 months of age and earlier, as predictors of adult IQ. Infant twins will be assessed to establish the genetic and environmental bases of the infant measures, and their parents' IQ will be measured using the WAIS-R and a specific cognitive abilities test battery. Mean parental IQ will be taken as the best available predictor of infants' later IQ, with the strength of the relationship being estimated by midtwin/midparent correlations and regressions. In addition, however, all infant twins will now become subjects in the independently funded MacArthur Longitudinal Twin Study, providing a direct longitudinal perspective up to 3 years of age that will enhance the value of and strengthen the proposed research. During the continuation period, we will select and strengthen those infant measures showing most promise as predictors of parental IQ. These involve non-motoric, attentional, and information processing tasks. Specifically, they are Fagan's infant intelligence test based on visual novelty preference and Haith's visual reaction time anticipation task. In view of the success of the novelty and anticipation measures, an auditory novelty preference task and a manipulatory anticipation (reaching) measure will be added. Haith's visual anticipation task and the two additional tasks will be administered by him in his laboratory. Measures of complexity of vocalization and other less directly related measures such as tester ratings (IBR), hand preference, orientation tasks, and selected Bayley items also will be included. We will extend the age of testing to include 5 months so that the study will be even more relevant to early prediction of IQ, and will double the sample size to include 400 twin pairs and 100 singletons in order to put our exciting preliminary findings on a firm empirical basis. The practical significance of this research lies in the diagnostic and therapeutic benefit that will ensue from the ability to predict adult intelligence from a brief and reliable battery of measures obtained in infancy. Its theoretical significance lies in the re- evaluation of the nature of infant intelligence.

Little is known about the etiology of individual differences in adjustment to the typical stresses of early adolescence. The purpose of the proposed research is to extend the longitudinal Colorado Adoption Project (CAP), initiated in 1975, into early adolescence. With the support of a grant from the Spencer Foundation, the adjustment of 185, 84, and 34 adopted and nonadopted children will have been assessed via telephone and questionnaire at 9, 10, and 11 years, respectively. In the proposed study, the total number of adopted and nonadopted children tested at 9, 10, and 11 will be increased to 576, 495, and 441, respectively. In addition, to test battery will be administered in our laboratory when the children are 12 years of age in order to validate the telephone test results, to administer the WISC-R IQ test, and to complete the test series for early adolescence. Importantly, these adopted and nonadopted children have been studied previously in their homes at 1, 2, 3, and 4 and in the laboratory at 7 years of age using an extensive set of psychological and environmental assessments; in addition, the biological and adoptive parents of the adoptees and parents of the matched nonadopted children have taken the WAIS-R IQ test and have completed a 3-hour battery of psychological measures. Because of the unique prospective, longitudinal design of this adoption study, the size and representativeness of its sample, its demonstrated absence of selective placement, and its low rate of attrition, the CAP has become a landmark study of the genetic and environmental origins of individual differences. For these reasons, and because such a study could not be conducted again given the drastic decline in the numbers of newborns available for adoption, it is imperative to capitalize on this opportunity to conduct the first prospective adoption study of adjustment during early adolescence.

Little is known about the etiology of individual differences during early adolescence. The purpose of the present proposal is to complete the evaluation of adopted and non-adopted children in the Colorado Adoption Project (CAP) at ages 9, 10, 11 and 12 years of age, on a variety of measures including adjustment, personality, general and specific cognitive abilities, social attitudes, physical growth and other variables. These children have been evaluated previously at ages 1, 2, 3, 4 and 7 in other components of the CAP and are, or will be tested at 13 through 16 years of age in another component. The number of CAP test sessions conducted with CAP subjects at ages 9 through 12 during the proposed continuation will be 327, bringing the total to 2,528 sessions. A novel feature of this continuation is the inclusion of the first 400 pairs of twins who have been tested previously on CAP measures at ages 1, 2, 3, 4, 5 and 7 as part of a combined adoption, sibling and twin study of development. We refer to this twin sample as the Colorado Twin Study (CTS) and will conduct 2,388 test sessions with these subjects during the proposed continuation, bring total twin tests to 2,456. The combination of these two comparable samples of subjects greatly strengthens the scope and power of inference regarding developmental processes in this study. The strengths of each approach to genetic and environmental etiology are combined and the limitations of each separate approach are overcome in this integrated genetically informative design. Because of the unique prospective, longitudinal design of both the adoption and twin study, the size and representativeness of their samples, the low rates of attrition, and, in the case of the adoption study, the demonstrated absence of selective placement, together with the use of many of the same measures in the two studies, and their full integrations at ages 4 and 7 years, this continuation provides the opportunity of establishing a truly landmark developmental study of the genetic and environmental etiology of individual differences during early adolescence. Given the investment in the CAP during the previous two decades, the investment in the CTS during the past decade, and the decline in the numbers of newborns available for adoption, a study of this scope and magnitude is unlikely ever to be undertaken again. It is, therefore, crucial to our research endeavor to take advantage of this unique opportunity.

A genetic contribution to anxiety-related disorders is not in doubt, but the nature of that contribution is still poorly understood. We propose to conduct a quantitative trait loci (QTL) study of fear- related/emotional behavior in the mouse as a first step to understanding possible genetic mechanisms in humans. An extensive test battery will be used to characterize fear-related behaviors in the mice and to identify potential confounds with general activity, photophobia, and thigmotaxis. We will use both simple univariate QTL mapping techniques and more recent multivariate mapping techniques to locate QTLs in F2 mice. These methods have been remarkably successful in mapping QTLs in plants and more recently in animal studies. The studies outlined in this proposal will use an inbred strain of mice that was derived from lines that were selectively bred for what may be fear- related behaviors. The mouse stains (developed by DeFries) were selectively bred for differences in open-field activity; two high (H1 and H2) and two low (L1 and L2) strains were developed. The H-L strains also differ in open-field defecation (L more H). F2 hybrids will be developed from L1 - H1 crosses and L2-H2 crosses. The F2 mice will be evaluated for activity in home cage and several measures of fear- related and depressive behaviors. Approximately 800 F2 animals will be scored in each cross (1600 animals total) and DNA from these same animals will be screened for polymorphisms using 84 microsatellites. The strategy of testing the replicability of the QLT method using two crosses derived from the mouse should serve as a rigorous test of the QTL methodology and could also facilitate the identification of fear- related genes in humans.