The ACE Model and LISREL

Karl G Jöreskog¹, Stefan Mattson¹

Different parameterisations of the ACE model are considered and their advantages and dissadvantages are discussed. It is shown that the commonly used latent variable (LV) parameterisation often give rise to computational problems in LISREL because of its non-unique solution. By contrast, the variance component (VC) parameterisation avoids these problems, is easier to apply, and gives correct standard errors of the estimated variance components due to genes, common and unique environment. The ACE model is often estimated by the maximum likelihood (ML) method assuming a normally distributed phenotype. However, this method can lead to incorrect standard errors and chi-square goodness-of-fit measures when the phenotype is non-normal. We report results of a simulation study that demonstrate the superiority of the asymptotically distribution free (ADF) method when the phenotype is non-normal. The case when the phenotype is measured only on an ordinal scale is also considered. This leads to complicated identification problems. It is shown how these can be resolved and that the relative variance components of genes, common and unique environment can be estimated by maximising simultaneous multinomial likelihoods of two or more observed contingency tables. The methods discussed are illustrated on real data.

Address:   Department of Statistics, Uppsala University, P O Box 513, S-75120 Uppsala, Sweden, Phone +46 18 4711165, Fax +46 18 554422, E.mail Karl.Joreskog@Statisik.uu.se

¹Department of Statistics, Uppsala University, Sweden

The Wisconsin Twin Panel: Temperament in Context²

H. Hill Goldsmith¹ & Kathryn S. Lemery¹

We report the establishment of a statewide infancy and early childhood twin panel for behavioral and health-related research. The base population comprised approximately 700 twin pairs per year since 1989. The ethnic breakdown of twin mothers was 87% White, 12% Black, and 1% American Indian, Asian and other. We interviewed the twins' primary caregiver over the telephone for a subset of over 500 families with 3 to 7 year old twins. Additionally, we sent a packet of questionnaires for both mothers and fathers to complete. We asked parents about their twins' temperaments, the sibling relationship, their own parenting style and personality, and levels of family emotional expressiveness. Considering the relationship between difficult temperament in the twins and sibling conflict (average phenotypic r = .40), we found that similarity between the twins was explained largely by shared genes for difficult temperament (MZ ICR = .76; DZ ICR = .25), and shared environment for sibling conflict (MZ ICR = .68; DZ ICR = .57). We fit a direction of causation model to this data. From detailed temperament questionnaires, we identified subgroups of twins with extreme temperaments for further study. We fit DF extreme group regressions to this data. We are currently planning a laboratory followup with structured psychopathology assessment of extreme temperamental groups.

Address:   Psychology Department, 1202 West Johnson Street, University of Wisconsin-Madison, Madison, WI 53706., (608) 262-1040 phone, (608) 265-3649 fax, hhgoldsm@facstaff.wisc.edu

¹Department of Psychology, University of Wisconsin-Madison, 1202 West Johnson Street, Madison, Wisconsin 53706 ²Supported by NIMH Grant MH50560

Preliminary Results from the Twins' Early Development Study: Language Delay and Mild Mental Retardation in Infancy.

Thalia C. Eley¹, & The Twins' Early Development Study²

This paper presents initial findings from the first birth-cohort in the Twins' Early Development Study. This is the first twin study of vocabulary development and mild mental retardation in infancy. Vocabulary scores (L. Fenson, P. S. Dale, J. S. Reznick, E. Bates, D. Thal, & S. J. Pethick, 1994, Monog. Soc. Res. Ch. Dev.,59: Serial No. 242, 1-173) and a parent-administered test of non-verbal cognitive ability (PARCA: K. J. Saudino, P. Dale, B. R. Oliver, S. A. Petrill, V. Richardson, M. Rutter, E. Simonoff, J. Stevenson, R. Plomin, in press, Br. J. Dev. Psych.) were collected from 3033 pairs of two-year-old twins. Model-fitting analyses of the variance-covariance matrices for the full distribution, and extreme group analyses using DF regression (J. C. DeFries, & D. W. Fulker, 1985, Beh. Genet.,15,467-473) for the lowest 5% of the sample were conducted for the vocabulary and non-verbal cognitive measures. Extreme group heritability for vocabulary scores (h²_g = .59, 95% CI .24-.94) was substantially greater than individual differences heritability (h² = ??, 95% CI ??-??). The reverse pattern was found for shared environment (c²_g = .31, 95% CI .00-.92, c² = ??, 95% CI ??-??). In contrast, the results from the PARCA indicate similar levels of genetic and shared environment influence on individual differences in the normal range and extreme group membership (h²_g = .25, 95% CI .05-.46, h² = ??, 95% CI ??-??, c²_g = .54, 95% CI .22-.86, c² = ??, 95% CI). These results suggest that while the genes influencing language disability may differ from those influencing language ability, the genetic influences on mild mental retardation are similar to those on the full range of cognitive ability in infancy.

Address:   Social, Genetic and Developmental Psychiatry Research Centre, Institute of Psychiatry, De'Crespigny Park, London SE5 8AF

¹Social, Genetic and Developmental Psychiatry Research Centre, Institute of Psychiatry, University of London, UK. ²The Twins' Early Development Study is funded by the Medical Research Council, UK. Principal Investigator: Robert Plomin; Co-Investigators: Dorothy Bishop, Philip Dale, Emily Simonoff, Jim Stevenson, Eric Taylor; Project Co-ordinator: Bonamy Oliver

The Netherlands Twin Register

Dorret I. Boomsma¹

The Netherlands Twin Register (NTR) was established in 1987 and is used as a research tool for developmental genetic studies in newborn twins and for longitudinal genetic studies in adolescent and young-adult twins and their families. The NTR registers around 50% of all newborn twins in The Netherlands each year. In 1987, 936 twin pairs and in 1995 1527 twin pairs were registered. All parents of young twins receive questionnaires about growth and development every one to two years. In the entire sample of newborn twins data on birth weight, height and growth, motor development, health- and behavior problems are collected. At ages 3, 7, 10 and 12 years childhood psychopathology is assessed by the Child Behavior Check List (CBCL). Results in 4000 3-year old twin pairs show high heritabilities for nearly all dimensions of behavioral problems. Subsamples of twins participate in specific projects. Over 200 twin pairs participated in a study of brain maturation (at ages 5 and 7 years), and cognition (at ages 5, 7, 10 and 12 years). This sample is also used to study the association of cognition with childhood behavioral problems. We have good experiences with collecting DNA samples via mouth swabs. Currently, these samples are used to determine zygosity.

Address:   De Boelelaan 1111, 1081 HV Amsterdam, The Netherlands, Phone 31-20-4448787, Fax 31-20-4448832, Email: dorret@psy.vu.nl

¹Vrije Universiteit, Dept of Biological Psychology

The Georgia Twin Registry: Sample Description and Research Issues

Irwin D. Waldman¹, Soo Hyun Rhee¹, Sarah A. Feigon¹, & Jessica L. Bar¹

In this paper we describe a new registry of 1600 twin pairs, ages 6 - 18, that were recruited over the past 4 years. Through a search of computerized birth records, we identified all multiple births in the state of Georgia during 1973-1991 and were given access to their complete birth records. Using parents' social security numbers, we then located the twin families' current addresses and conducted our first mailing consisting of an invitation to join the registry and a family information form. This form comprised demographic questions regarding family composition, family income, and the age, ethnicity, education, and occupation of family members, as well as a number of questions regarding physical similarity that were used to determine twins' zygosity. The registry is socioeconomically and ethnically diverse, with a substantial representation of African-American families. Twins and their families from this registry are participating in a series of research projects on personality and psychopathology in children. We are currently conducting a mailed questionnaire study of personality and temperament and have completed a mailed questionnaire study of DSM-IV externalizing and internalizing symptoms. My graduate students and I are using these twin data to investigate a number of issues in the classification of child psychopathology and its relation to personality. Specific issues include: testing alternative models for sex differences in Attention Deficit Hyperactivity Disorder (ADHD) and its component symptom dimensions; clarifying the causes of comorbidity among ADHD, Oppositional Defiant Disorder (ODD), and Conduct Disorder (CD); clarifying the causes of comorbidity among anxiety disorders and depression and the role of personality therein; examining the role of specific environmental influences and candidate genes within behavior genetic models; and examining convergent and discriminant validity of laboratory measures and behavioral ratings of inattention, impulsivity, and related constructs within a behavior genetic framework.

Address:   mailing address - Department of Psychology, 532 N. Kilgo Circle, Emory University, Atlanta, GA 30322, phone - (404) 727-7430, fax - (404) 727-0372, e-mail - waldman@fs1.psy.emory.edu

¹Department of Psychology, Emory University, Atlanta, GA 30322.

Risk factors for aggression and antisocial behavior in twins

Laura A. Baker¹ & Adrian Raine¹

Human aggression and antisocial behavior are the product of both social and biological risk factors. What is not yet understood is how environment and genes mediate the interrelationships among these risk factors and antisocial outcomes. A study of twins and their families provides the ideal opportunity to answer the critical question in this regard: Do measured social and biological variables relate to antisocial development for environmental or genetic reasons? Our ability to develop effective and efficient interventions for antisocial behavior rests heavily upon the answer to this question. A study of normal variation in antisocial and aggressive behavior is underway for 600 twin pairs (both male and female), age 9-10 years during an initial assessment, and age 11-12 years during a follow-up assessment. Our goal is to investigate the environmental and genetic underpinnings of important social and biological risk factors for unlawful, antisocial, and aggressive behavior in boys and girls on the brink of adolescence. Measures of social risk factors include aspects of the family environment, such as socio-economic status, emotional climate, cohesion, parental warmth and affection, parental supervision, discipline and control. Specific environmental factors for each twin are also studied, including individual relationships with each family member, as well as peer-group characteristics. Biological risk factors include psychophysiological indicators of arousal (both electrodermal and cardiac), as well as neuropsychological and cognitive testing. The social and biological risk factors studied here have previously been shown to distinguish between children at high and low risk for aggressive and antisocial behavior. The twin design allows the unique occasion to estimate the relative contributions of environmental and genetic factors to both antisocial behavior and their risk factors, as well as their interrelationships. DNA samples, extracted through mouthwash procedures, are stored for future analyses of sp ecific genetic loci hypothesized to show associations with antisocial behaviors.

Address:   Department of Psychology, SGM 501, University of Southern California, Los Angeles, CA 90089-1061, Baker Phone: (213) 740-2261 FAX: (213) 746-9082 email: lbaker@usc.edu, Baker internet: http://www-rcf.usc.edu/~lbaker/, Raine Phone: (213) 740-7348 FAX: (213) 746-9082 email: raine@usc.edu, Raine internet: http://www-rcf.usc.edu/~raine/

¹Department of Psychology, University of Southern California, Los Angeles, CA 90089-1061

Parasomnias among adult Finnish twins

J. Kaprio^1,2, C. Hublin³, M. Partinen³, K.Heikkilä¹, & M. Koskenvuo⁴

Parasomias are phenomena that intrude into, or occur during, sleep without becoming primary disorders of the states of sleep and wakefulness, per se. Although many are quite common and well-known to the general population, their epidemiology is poorly understood, and no large, representative twin studies have been carried out. For these reasons, we investigated genetic and environmental effects on sleeptalking, nocturnal enuresis, and nightmares. Similar analyses of sleepwalking and bruxism have been published (C. Hublin, et al, 1997, Neurology, 48,177-181, J. Sleep Res., in press). Methods: A questionnaire sent, in 1990, to Finnish Twin Cohort subjects aged 33-60 years, yielded responses from 11,220 twins with complete data for all items on occurrence of common parasomnias during childhood, and as adults. Included were responses from both members of 1,298 MZ and 2,149 like-sex DZ twin pairs. Mx models were fit to these data, in most models, separately by gender. Results: Genetic factors accounted for 67-70% of the phenotypic variance in childhood enuresis; too few subjects reported enuresis in adulthood to permit analyses. For nightmares, additive genetic factors accounted for 44-45% of phenotypic variance in childhood and for 36-38% of variance in nightmare reports by adults. Finally, genetic factors accounted for 51-54% of the variance in sleeptalking in childhood and 37-48% in adulthood. The genetic correlation between childhood and adult sleeptalking was high, but only adult sleeptalking was associated with increased risk of serious psychopathology, assessed using medical register data. Summary: Sleeptalking, nightmares, and nocturnal enuresis in childhood are common parasomnias with modest to strong genetic components.

Address:   Department of Public Health, P.O. Box 41, Mannerheimintie 172, FIN-00014, Helsinki, Finland, 358-9-19127595 (ph), 358-9-19127600 (fax), jaakko.kaprio@helsinki.fi

¹Department of Public Health, University of Helsinki, Helsinki, Finland ²Department of Mental Health and Alcohol Research, National Public Health Institute, Helsinki ³Haaga Neurological Research Centre, Helsinki ⁴Department of Public Health, University of Turku, Turku, Finland

Social support, life-events, and vulnerability to depression among 12,063 adult twins

K. Romanov¹, J. Varjonen², J. Kaprio^1,3, & M. Koskenvuo²

Factors explaining vulnerability to depressive symptoms as assessed by the Beck Depression Inventory (BDI) were studied among 12,063 Finnish adult same-sex twins, aged 33 to 60 years. The effect of stressful life-events on variation of depressive symptoms was analyzed, and quantitative and qualitative social support, somatic health, neuroticism, self-assurance, marital status and socioeconomic class were controlled. Quality and quantity of social support had independent buffering effects on the impact of life events on BDI scores. Multivariate models, analyzed separately for men and women by their somatic health status showed that BDI scores were associated with life events, lack of social support, and chronic somatic disease. Analyses of pairwise twin data are in progress and will be reported.

Address:   Department of Public Health, P.O. Box 41, Mannerheimintie 172, FIN-00014, Helsinki, Finland, 358-9-19127595 (ph), 358-9-19127600 (fax), Kalle.Romanov@helsinki.fi

¹Department of Public Health, University of Helsinki, Helsinki, Finland ²Department of Public Health, University of Turku, Turku, Finland ³Department of Mental Health and Alcohol Research, National Public Health Institute, Helsinki

Peers, teachers, and parents as raters of twins' behavioral and emotional problems and adjustment⁵

Lea Pulkkinen¹, Jaakko Kaprio^2,3, & Richard J. Rose⁴

The Multidimensional Peer Nomination Inventory (MPNI; 30 items), and its parallel teacher- and parental-rating forms (37 items each), were developed and administered to 944 12 year-old Finnish twins drawn from three nation-wide birth cohorts (1983-85) in an ongoing, longitudinal study, FinnTwin12. A total of 12,937 children participated in peer nominations in the 503 school classes in which the targeted twins were enrolled. The twins formed 154 MZ, 132 same-sex DZ, and 127 opposite-sex DZ pairs with complete data and confirmed zygosity. Data from peer nominations, teacher- and parental- ratings, yielded three factors, with scales (and subscales) for Behavioral Problems (Hyperactivity-Impulsivity, Aggression, Inattention), Emotional Problems (Depression, Social Anxiety), and Adjustment (Constructiveness, Compliance, Social Activity). Results from the 944 twin individuals provide evidence of reliability and concurrent validity (against teacher ratings) of peer-referenced assessment, and all scales, except Depression and Social Anxiety, differentiated boys from girls in expected directions. Consistently, across scales and across informants, correlations for MZ cotwins exceeded those for DZ cotwins, but, across informants, significant similarity was found for both SSDZ and OSDZ twins, as well, with little evidence of contrast effects. In peer nominations, all MZ correlations were 0.70 or higher, and DZ correlations ranged from 0.29 to 0.53. The pattern of twins' correlations suggests substantial additive genetic variance in all scales of the MPNI with no effects of sex-limitation at this age; modeling, with Mx, will be used to confirm these suggestions, when subject ascertainment is complete and the twin sample is doubled in size.

Address:   Department of Psychology, University of Jyväskylä, P.O.Box 40351, Jyväskylä, Finland; email, leapulkk@tukki.jyu.fi

¹Department of Psychology, University of Jyväskylä ²Department of Mental Health and Alcohol Research, National Public Health Institute, Helsinki ³Department of Public Health, University of Helsinki ⁴Department of Psychology, Indiana University, Bloomington ⁵FinnTwin12 is supported by AA-09203 and the Academy of Finland

Genes, environments, and precursors of alcoholism: Finnish twin-family studies⁶

Richard J. Rose¹, J. Kapio^2,3, L. Pulkkinen⁴, R .J. Viken¹, M. Koskenvuo⁵

Genetic and environmental influences on behavioral precursors of alcoholism are studied in two ongoing, longitudinal studies of Finnish twins and their parents. FinnTwin16 has baseline data on 2,810 16 year-old twin pairs; questionnaire baseline assessments of FinnTwin12, in progress, are complete for >2,000 twins, and school-based assessments, including peer-nominations for eight dimensions of social behavior, are available for 413 pairs of known zygosity. Abstinence rates at age 16 are influenced by socio-regional variation, sibling interaction effects, and parental drinking patterns, with influences of both sibs and parents greater in some regional environments than others. Among non-abstinent twins at age 16, genetic effects influence frequency of consumption and, among those who intoxicate, of frequency of intoxication. These genetic effects are sex-limited and increase with age. Personality scales predictive of adolescent alcohol use (e.g., MMPI Pd), in between-family analyses of twin individuals, directionally predict differences within highly-discordant MZ twin pairs, inviting a search for environmental sources of personality-discordance. Identified from their parents' screening questionnaires, 11-12 year-old twins at elevated familial risk of alcoholism are rated by teachers and peers as more aggressive and less complaint; at age 14, they report that more of their peers use tobacco and alcohol, and they more likely report drinking to intoxication. Such risk-relevant behavioral differences invite genetic analyses as possible precursors of alcohol abuse/dependency.

Address:   Department of Psychology, Indiana University, Bloomington, IN, 47405-1301; 812-855-8779, (ph) 812-855-4691 (fax); rose@indiana.edu

¹Department of Psychology, Indiana University, Bloomington, Indiana ²Department of Public Health, University of Helsinki, Finland ³Department of Mental Health and Alcohol Research, National Public Health Institute, Helsinki ⁴Department of Psychology, University of Jyväskylä, Finland ⁵Department of Public Health, University of Turku, Finland ⁶Supported by NIAAA (AA 08315, AA 09203, and AA 00145)

Migrant-discordant twins: A natural experiment on effects of environmental change

Niklas Hammar^1,2, J. Kaprio^3,4, L. Alfredsson,¹ M. Koskenvuo⁵, T. Rönnemaa⁶, T. Hammar⁷, U.Hagström¹, & L. Viikari⁶

Migrant studies offer a naturalistic experiment to enhance our understanding of how changes in environment and life-style can affect health habits, disease-risks and behavioral outcomes. Long-term international migration may significantly alter environmental exposure to factors relevant to disease-risk and trait-development, and comparisons of migrants with non-migrants have a long history in medicine and psychiatry. But such comparisons are confounded with uncertain matching of migrants and non-migrants on genetic and familial background, and the most meaningful contrasts are of migrants and non-migrants within-sibships, even more so within pairs of twins. In an ongoing collaborative study, a postal questionnaire is distributed to all twin pairs of the older Finnish Twin Cohort Study in which one or both cotwins have migrated to Sweden. We project the sample will include 2,700 twins or 1,350 pairs. Data-collection began, early 1998, with questionnaire content including medical history and physical symptoms, alcohol and tobacco use, dietary and exercise habits, educational, employment, and residential history. A clinical examination will be made of the most resident-discordant twin pairs. Extensive baseline information on these migrant-discordant twins was collected in earlier postal surveys, and the matched analysis of migrant-discordant twins will offer unusual, if not unique, power in understanding how environmental changes affect behavioral development and disease outcomes.

Address:   Niklas Hammar, Department of Epidemiology, Division of Public, Health,Karolinska Hospital, Norrbackabyggnaden, S- 171 76 Stockholm, Sweden. E-mail:, niklas.hammar@imm.ki.se

¹Department of Epidemiology, Institute of Environmental Medicine, Karolinska Hospital, Stockholm ²Department of Epidemiology, Division of Public Health, Karolinska Hospital, Stockholm ³Department of Public Health, University of Helsinki ⁴Department of Mental Health and Alcohol Research, National Public Health Institute, Helsinki ⁵Department of Public Health, University of Turku ⁶Department of Internal Medicine, Turku University Hospital, Turku ⁷Centre for Migration Research, Stockholm University, Stockholm

Linkage studies suggest a possible locus for dyslexia near the Rh region on chromosome 1

E. L. Grigorenko^1,2, F. B. Wood³,M. S. Meyer³, J. E. D. Pauls⁴, L. A. Hart⁵, D. L. Pauls^1,2

Eight extended dyslexic families with at least 4 affected individuals were genotyped with twelve genetic markers spanning the Rh (rhesus factor) locus. Eleven of these markers were located on the short arm and the other was on the long arm of chromosome 1. Four theoretically derived phenotypes were used in the linkage analyses: 1) Phonemic Awareness; 2) Phonological Decoding; 3) Rapid Automatized Naming; and 4) Single Word Reading. Both parametric and nonparametric genetic analyses were completed. The statistical analyses were performed in three steps. First, pairwise linkage analyses were completed. Two chromosomal markers (D1S199 and HYTM) gave lod scores > 1 for three phenotypes: Phonemic Awareness, Decoding, and Word Identification. No positive results were obtained for the phenotype of Rapid Naming. Next, two different sets of nonparametric analyses were performed. Allele-sharing among affected relatives was examined using both identity-by-state and identity-be-descent methodologies. For pairwise analyses, both methodologies revealed significant allele-sharing in for markers D1S199 and HYTM as well as other markers in the region. When the magnitudes of the obtained effects were ordered, the strongest linkage evidence was obtained with the phenotype of Phonological Decoding. Finally, multipoint analyses were completed in an attempt to narrow down the region of significance and localize the areas of possible linkage. These results suggest that the most likely location of the putative locus is in a region near marker D1S199. This region is immediately distal to the Rh-gene.

Address:   Psychology Department, Yale University, 2 Hillhouse Ave, New Haven, CT 06520, Phone: (203) 432-4660, Fax: (203) 432-7172, E-mail: elena.grigorenko@yale.edu

¹Child Study Center, Yale University School of Medicine, New Haven, CT 065 10 ²Department of Psychology, Yale University, New Haven, CT 06520 ³Section of Neuropsychology, Department of Neurology, Bowman Gray School of Medicine, Winston-Salem, NC 27103 ⁴Department of Biochemistry, University of Vermont, Burlington, VT 05405 ⁵Department of Psychology, University of Pittsburgh, Pittsburgh, PA 15260

Quantitative Trait Locus for Specific Language and Reading Deficits on Chromosome 6p⁶.

Javier Gayan¹, S. D. Smith², S. S. Cherny^1,3, L. R. Cardon⁴, D. W. Fulker^1,3, W. J. Kimberling², R. K. Olson¹, B. F. Pennington⁵, & J. C. DeFries¹.

Reading disability (RD), or dyslexia, is a complex cognitive disorder manifested by difficulties in learning to read in otherwise normal individuals. Reading-disabled subjects exhibit deficits in several reading and language skills. Previous research has suggested the existence of a quantitative trait locus (QTL) for RD on the short arm of chromosome 6. In the present study, RD subjects' performance in several measures of word recognition and component skills of orthographic coding, phonological decoding, and phonological awareness, were individually subjected to QTL analysis with a new sample of 126 sib pairs, using a multipoint mapping method and 8 informative DNA markers on chromosome 6 (D6S461, D6S276, D6S258, D6S306, D6S105, D6S439, D6S291, D6S1019). The results indicate that there is significant linkage in the region of the first five markers for deficits in orthographic and phonological skills, providing a confirmation of previous findings.

Address:   Javier Gayan, Institute for Behavioral Genetics, University of Colorado, Boulder, CO 80309-0447, Phone: 303 492 2817, Fax: 303 492 8063, Email: gayan@colorado.edu

¹Institute for Behavioral Genetics, University of Colorado, Boulder, CO 80309-0447. ²Center for Hereditary Communication Disorders, Boys Town National Research Hospital, Omaha, NE 68131. ³Social, Genetic and Developmental Psychiatry Research Centre, Institute of Psychiatry, London, UK SE5 8AF. ⁴Sequana Therapeutics, La Jolla, CA 92037. ⁵Department of Psychology, University of Denver, Denver, CO 80208. ⁶Supported by NICHD Grants HD-11681 and HD-27802, and RO1 HD-22223.

Testing psychological theories of hyperactivity within a twin study

Jonna Kuntsi¹ & Jim Stevenson²

In an attempt to integrate the genetic and cognitive/motivational literatures on hyperactivity, we tested different psychological theories of the disorder within a genetically informative design. Pairs of twins with at least one pervasively hyperactive member and control pairs were identified from general population samples in schools in Southern England. Ratings obtained from teachers and parents of 131 monozygotic and 136 dizygotic twin pairs confirmed previous reports of high heritability for hyperactivity (e.g. J. Stevenson, 1992, Behavior Genetics, 22, 337-344). To investigate which aspects of psychological functioning might be carrying this genetic effect, we assessed a subsample of 46 MZ and 47 DZ pairs on tests related to three theories of hyperactivity: the inhibition deficit, executive function/working memory and delay aversion theories. Comparisons between hyperactive and control children provided little support for the inhibition or executive function theories, whereas the data were consistent with a more motivational account of hyperactivity. There were genetic effects only on some of the measures, however. The strongest candidate for carrying the genetic effect on hyperactivity at the level of psychological functioning relates to the variability in their speed of responding.

Address:   Behavioural Sciences Unit, Institute of Child Health, 30 Guilford Street, London WC1N 1EH, UK, Tel: + 44-171-8310975, Fax: + 44-171-8317050, email:, J.Kuntsi@ich.ucl.ac.uk.

¹Institute of Child Health, University College London Medical School, UK ²University of Southampton, UK

Sex differences in the etiology of inattention and hyperactivity/impulsivity

Soo Hyun Rhee¹, Irwin D. Waldman¹, David A. Hay², & Florence Levy³

Several recent studies have examined whether sex is a moderator of the magnitude of genetic and environmental influences on ADHD. Most of these studies have examined ADHD as a unidimensional construct (e.g., DSM-III-R ADHD) and have not addressed how results may differ for the DSM-IV symptom dimensions of inattention and hyperactivity/impulsivity. In this paper, we contrast several alternative models that test for sex differences in the magnitude of genetic and environmental influences on inattention and hyperactivity/impulsivity. The participants are 2043 3- to 15-year-old twin pairs and 348 3- to 18-year-old sibling pairs from Australia. Zygosity was determined from mothers' responses to questions regarding the physical similarity of the twin pairs. Data on DSM-IV ADHD symptoms were gathered from a questionnaire completed by mothers of the twins. We compare the fit of a number of alternative sex difference models, including the homogeneity model, the heterogeneity model, the scalar model, and the general sex-limitation model. The implications of the best-fitting models are discussed.

Address:   mailing address - Department of Psychology, Emory University, Atlanta, GA, 30322, phone - (404) 727-2741, fax - (404)727-1284, e-mail - soorhee@social-sci.ss.emory.edu

¹Department of Psychology, Emory University, Atlanta, GA 30322. ²Department of Psychology, Curtin University, Perth, Western Australia, Australia. ³Department of Child and Adolescent Psychiatry, The Prince of Wales Children's Hospital, High Street, Randwick, NSW 2031, Australia.

Genetic validation of the DSM-IV subtypes of ADHD⁵

David A. Hay¹, M. McStephen, ^1,2, I. Waldman,³, F. Levy ,⁴

Compared with the DSM-III-R unitary view of Attention Deficit Hyperactivity Disorder (ADHD), DSM-IV made radical change in defining three distinct subtypes of ADHD, namely the Predominantly Inattentive(PI), the Hyperactive/Impulsive (HI) and the Combined (C) subtypes. Given the exceptionally high genetic component to ADHD (F.Levy, D.A. Hay, M. McLaughlin, C. Wood and I. Waldman, 1997, J.Amer. Acad. Child Adol. Psych. 36737-744), MZ and DZ concordances should provide some validation that these three subtypes can be reliably distinguished. The Australian Twin ADHD Project (ATAP) is a longitudinal study of some 2000 families in the Australian NHMRC Twin Registry and the data here come from the 1167 same-sex twin pairs in the 1994-5 second wave when they were aged 8-16. Parents completed our Australian Disruptive Behaviours Scale, which is a conservative measure of ADHD symtomatology and diagnosis (Levy et al. 1997). Proband-wise concordances and cross-concordances between the subtypes indicated that three subtypes were essentially distinct genetically with the MZ cross-concordances being 0.10 or less. DSM-IV requires 6/9 symptoms and the distributions of subtype symptoms in the cotwins of the probands provides additional validation for such cutoff scores. The age distributions for the subtypes were very similar, providing no support for the view (R. Barkley, 1997, Psych. Bull. 121, 65-94) that the HI subtype is a developmental precursor of the C subtype. The phenotypical and genetic relationship of the subtypes to speech and reading problems represented further validation of these distinctions. Rates of intervention for these problems were very low in the HI subtype, but 20% in the PI and 50% in the C subtype. The existence of a genetically discrete subtype combining both Inattention and Hyperactivity/Impulsivity is a challenge for conventional multivariate genetics that seeks to model the relationship between these groups of symptoms.

Address:   School of Psychology, Curtin University of Technology,PO Box U 1987, Perth, WA 6845, Australia Phone 61 8 9266 7025 Fax 61 8 9266 2464, email d.hay@psychology.curtin .edu.au

¹School of Psychology, Curtin University of Technology, Perth, WA 6845,Australia ²Mental Health Research Institute of Victoria, Parkville, VIC 3152, Australia ³Department of Psychology, Emory University, Atlanta,GA, 30322, USA ⁴Avoca Clinic, Prince of Wales Hospital, Randwick, NSW, 2031, Australia ⁵Funded by the National Health and Medical Research Council of Australia

Genetic and Environmental Associations Between Attention Problems and Other Problem Behaviors³

Stephanie Schmitz^1,2 & David W. Fulker²

Both mothers and teachers assessed a sample of same-sex twin with the Child Behavior Checklist (CBCL/4-18; Achenbach, 1991; University of Vermont, Burlington, VT)/Teacher Rating Form (TRF; Achenbach, 1991; University of Vermont, Burlington, V) while the children were in first grade. These ratings showed that attention problems correlated highly with other aspects of problem behavior and to a slightly stronger degree with those from the externalizing than the internalizing domain (rs of .58 and .53 for mother and .62 and .47 for teacher ratings, respectively). For teacher ratings, the hyperactivity aspect of attention problems was slightly stronger associated with externalizing (.71 vs. .38) while inattention correlated with internalizing and externalizing with a similar magnitude (r of .51 and .48, respectively). Univariate analyses showed that parameter estimates for genetic influences tended to be higher for maternal than teacher ratings (.72 vs. .44 for the overall attention problem score) and that while the shared environment influenced hyperactivity to some degree, it was non-significant for inattention. Multivariate analyses showed that the observed phenotypic association of attention problems with other aspects of problem behavior was mainly due to common genetic influences. Shared environmental correlations were often close to unity but did not contribute much to the expected phenotypic correlation. Sex differences in these associations will be explored.

Address:   University of Colorado, Campus Box 447, Boulder, CO 80309-0447, tel. (303) 315-7114/492-0835, fax (303) 315-6606/492-8063, email schmitzs@essx.uchsc.edu or schmitzs@colorado.edu, URL http://ibgwww.colorado.edu/~schmitzs

¹University of Colorado Health Sciences Center, CPH Box C268-69, Denver, CO 80262 ²Institute for Behavioral Genetics, University of Colorado, Campus Box 447, Boulder, CO 80309-0447 ³Supported by NIH grant HD 18426 and a grant from the John D. and Catherine T. MacArthur Foundation. S.S. is partly supported by grant MH 15442.

Philosophy of Human Behavior Genetics

S. Wieder¹

As a physician, and a biological scientist I am very much involved in the scientific investigation of human behavior and the heredity of mental and physical traits. I also consider myself a philosopher, however, and from this point of view I find that behavior genetics research is very much constrained and restricted in its perspective, lacking a firm philosophical foundation. In this paper, I delineate important ethical and philosophical principles which should guide future education and research in behavior genetics. A detailed analysis is made demonstrating that of some of the most intractable theoretical problems in the field of behavior genetics require highly sophisticated methods of philosophic investigation. This paper formulates some solid rules which should safely support any future structures of thought and methods of empirical investigation. Scientists and philosophers must agree to jointly pursue genetic knowledge based on the following guiding principles of education and research: Unswervingly support the intellectual freedom of those conducting any form of investigation in the field of behavior genetics as long as they break no just laws. Pursue knowledge of behavior genetics in all its aspects, assumptions and implications without concern for where this knowledge leads and oppose all censorship and control of genetic knowledge. Make every effort to avoid and correct any errors or distortions of knowledge of genetics at the earliest possible stage. Oppose all dogmatic viewpoints on genetic knowledge whether pro or con. Examine the moral and ethical issues and implications of genetics research, and work to publish and otherwise broadcast such observations and findings. Consider with neutrality the proposition that heredity subtracts from human freedom as well as the contrary view that inborn capacities are the source of all human innovation, spontaneity, creativity and freedom. Examine and attempt to anticipate the risks, dangers and potential abuses of genetic knowledge while defending t he absolute right to pursue such knowledge without restriction. Organize meetings of scholars and researchers from any field who wish to pursue knowledge of behavior genetics. Form a committee of eminent researchers and scholars to evaluate and defend the research and theories of behavior geneticists innocent of scientific fraud but who merely espouse unpopular ideas or conduct controversial research. Behavior geneticists should join with philosophers to form a Society for the Study of Philosophical Issues in Behavior Genetics.

Address:   P. O. Box 120, Rock Hill, NY 12775-0120, Tel: (914)-374-2008, Fax: (914)-374-3807, e-mail: emagine@sprynet.com

¹B.W.S. Foundation, Inc.

Genes and Environment: A New Theory of Freedom

Mark Ast¹

The present paper examines the relative behavioral effects of genes and environment and offers a new theory which serves to dissolve their supposed opposition. Paradoxically, both genes and environment may be said to completely determine (different aspects of) human behavior. This new theory introduces a distinction between intensive and extensive magnitudes of determination, varying independently. For example, the rose and the weed are determined to be such, not by reference to some vague and indefinite "mixture" or "interaction" between genes and environment; but each is determined to be what it is (either rose or weed) by genes alone, and at the same time each is determined to be how much or how fully what it is by its environment alone. A nourishing environment will never transform what is genetically a weed into a rosebush, and a deprived environment will never turn a rose bush into a weed—though the rosebush may be so malnourished as to die--or fail to flower and therefore appear weed-like. This distinction is employed to explain the basis of environmental and genetic determination as complete and yet mutually consistent. The complete articulation of this theory includes a new definition of freedom and determination according to which that which is "genetic" may be free of determination in more than the metaphorical sense. The theory is demonstrated to be consistent with, though in important particulars distinct from: Kant's distinction between: (1) empirical determination vs (2) transcendent(al) freedom; and Spinoza's definition of freedom according to which: (1) a certain form of causal determination (genetic or otherwise) is identified with freedom, (2) all things are thoroughgoingly determined, and yet (a) freedom is determination in conformity to the "nature" of the thing determined, (b) that action is free which is in accord with the "nature" of the thing determined, and (c) the free mind is the mind which determines external things and experiences rather than being determined by them to th ink or to act the way it does (the "active intellect").

Address:   Hillside Hospital, CENORR, AT Building B-15, 75-59 263rd Street, Glen Oaks, NY 11004 USA, Tel: 718-470-8393, Fax: 718-962-2742, Email: dialogue@ibm.net

¹Hillside Hospital - Center for Neuropsychiatric Outcome and Rehabilitation Research

Twins' ratings on the "Sibling Inventory of Differential Experience" (SIDE): a three-year follow-up study

Ruth Guttman¹ & Nitzan Goldschmidt¹

The "Sibling Inventory of Differential Experience" (SIDE) aims to assess nonshared experiences in the same family. Siblings are asked to rate their relations to each other in a relative way ("I am more--or less--......than my sib"). Jerusalem MZ and DZ twin pairs (aged 9 to 14) were administered a shortened version of the SIDE, with questions limited to 'sibling closeness', 'sibling antagonism', sibling care taking', and 'sibling jealousy'. Each child was interviewed in three successive years in a cross-sequential design. Average absolute (unsigned) differences in replies were calculated for the MZ and DZ groups. MZ's were found to be more similar than DZ's on each of the scales, indicating genetic influence. Absolute differences between the two groups of twins decreased from year 1 to year 3, i.e.similarities in perception of experiences in the family increased over time in both kinds of twins, but at different levels.

Address:   Department of Psychology, Hebrew University,91905 Jerusalem, Israel, Tel. 972-2-6751589, Fax. 972-2-6413313, e-mail msruthg@mscc.huji.ac.il

¹Scheinfeld Center for Human Genetics in the Social Sciences, Department of Psychology, The Hebrew University of Jerusalem

Discrepancies in life event reports with multiple raters and multiple time points⁴.

J. M. Lessem¹, J. K. Hewitt¹, L. J. Eaves², J. L. Silberg², M. Rutter³, E. Simonoff³.

The Virginia Twin Study of Adolescent Behavioral Development includes self report data and parental report data on life events for adolescent twins. Though generally children and parents agree on which life events occurred and which did not occur, children and parents on average disagree on the occurence of two events out of 39 possible events on a modified form of the Johnson and McCutcheon Life Event Checklist. The variance of discrepancies can be broken down into shared and non-shared environmental components, though the shared environmental component may be due to bias, because only one mother and father responds for both twins, while the children answer only for themselves. This report examines the nature of the discrepancies between multiple respondents to the same life events questionnaire at two time points, 18 months apart.

Address:   University of Colorado, Campus Box 447, Boulder, CO 80309-0447, 303-492-2843, 303-492-8063 (fax), Jeff.Lessem+BGA@Colorado.EDU, http://ibgwww.Colorado.EDU/~lessem

¹Institute for Behavioral Genetics, University of Colorado, Boulder, CO 80309-0447. ²Department of Human Genetics, Medical College of Virginia, Richmond, VA \ 23298. ³MRC Child Psychiatry Unit, Institute of Psychiatry, University of London, London, UK. ⁴Supported by PHS Grant MH45268.

Iris Patterns & Personality: Does a Relationship Exist That Can Be Useful For The Five Factor Model And Behavioral Genetics?

Mats Larsson¹

The Rayid Method of Iris Interpretation (Rayid) is a personality model that has two basic assumptions; (i) that specific fiberpatterns in the iris of the eye are inherited, and (ii) that those in turn can be associated with specific personality traits. The model's first basic assumption has received some support. Twinstudies have documented that it is possible to diagnose genetic similarity between two individuals just by comparing the visual similarity of irispatterns (A. Burkhardt, 1992, Antrop. Anz. 50(3), 235-170). The model's second basic assumption has only been tested once showing a possibility towards its support (J. J. Bruno, 1990 A dissertation presented to the faculty of the California Institute of Integral Studies, San Francisco, 152 pages). The main purpose with this study is to test if specific fiber structures in the iris are stable genetic markers for specific personality traits in The Five Factor Model (FFM). The iris factor was investigated by creating 4 experiment groups (n=10), in accordance with the criteria's for Rayids four basic personality types, and a control group (n=29) which was randomized. The personality factors that measure the basic personality typed behaviors according to the Rayid Model were operationalized by choosing 34 items from the FFM personality inventory for each experimental group. The result showed significant differences between the mental and the emotional type, and between the extreme type and the control group. Furthermore, the items that were operationalized to each type could, in two of three cases, predict the subjects to their theoretical type. The results indicate the possibility for significant differences between Rayid's basic personality types in a large sample. The possibility to map the DNA sequences that are responsible for the fiber patterns in the iris, as well as using irispatterns as a methodology to identify candidate genes in personality research are discussed.

Address:   Mats Larsson, University of Orebro, Institution for Social Sciences, Psychology section, 701 82 Orebro, Sweden. Phone: +46 19 303185. Fax: +46, 19 301484. E-mail: mats.larsson@hoe.se

¹University of Orebro, Institution for Social Sciences, Psychology section, 701 82 Orebro, Sweden.

Genetic influences on personal assessments of stressful situations

Andreas Busjahn¹, Hans-Dieter Faulhaber¹, Friedrich C. Luft¹

External stressors do not exert their influence on humans directly, but are filtered by personal assessment as part of the coping process. Situational characteristics like significance, emotional quality (e.g. pleasant vs. unpleasant, anger), challenge, threat, controllability are important factors in the network of the stress-response. Cardiovascular reactions to stress are heritable, independent of the genetic factors influencing resting blood pressure (A. Busjahn et al., 1996, J.Hypertens. 14, 1195-1199). Beside physiological systems regulating blood pressure psychological processes are involved, thus genetic influences on personal assessment may contribute to the heritability of the stress response. We tested the hypothesis of genetic influences on assessment in the context of a twin study on a variety of cardiovascular phenotypes, including reaction to a psychological stress test. 122 MZ twin pairs and 65 DZ twin pairs are included so far. After completing three different types of task (Stroop-test, mental arithmetics, psychomotor task) probands were asked to assess the test situation on a number of 5-point Likert Scales. The two extremes were verbalized (e.g. Did you experience the situation as pleasant ... unpleasant). Heritability estimates for the different dimensions of assessment are in the range of 0 to 0.30 and thus smaller than those of more generalized personality dimensions. Nevertheless there is a significant genetic influence on this relevant psychological process. The hypothesis of shared genetic variance of cardiovascular stress response and personal assessment of stressful situations will be tested in a larger sample at the end of our study.

Address:   FVK, 13125 Berlin, Wiltbergstr. 50, Phone: +49-30-94172277, Fax: +49-30-94172335, e-mail: busjahn@fvk-berlin.de

¹Franz-Volhard-Klinik at the Max-Delbrück-Center for Molecular Medicine

Differential genetic etiology of reading disability as a function of IQ³

S. J. Wadsworth¹, R. K. Olson¹, B. F. Pennington^1,2, & J. C. DeFries¹

In order to test the hypothesis that the genetic etiology of reading disability differs as a function of IQ, composite reading performance data from 223 pairs of identical twins and 169 pairs of same-sex fraternal twins in which at least one member of each pair was reading-disabled were subjected to multiple regression analysis (J. C. DeFries and D. W. Fulker, 1985, Behav. Genet.15, 467-473; J. C. DeFries and D. W. Fulker, 1988, Acta Genet. Med. Gemellol, 37, 205-216). Heritability of the group deficit in reading performance (h²_g) across the full range of IQ scores was .58. When the basic regression model was fitted separately to data from twin pairs with full-scale IQ scores below 100 or 100 and above, resulting estimates of h²_g were .43 and .72, respectively, a significant difference (p < .03, one-tailed). Results of fitting extended regression models to reading performance and IQ data from the entire sample provided evidence that the genetic etiology of reading disability differs as a linear function of IQ (p < .007, one-tailed). These results suggest that IQ is relevant for the diagnosis of reading disability, and that environmental influences may be more salient as a cause of reading difficulties in children with low IQ.

Address:   Sally J. Wadsworth, Institute for Behavioral Genetics, Campus Box 447, University of Colorado, Boulder, CO 80309, Telephone: 303-492-6795, FAX: 303-492-8063, e-mail: wadswort@ibg.colorado.edu

¹Institute for Behavioral Genetics, University of Colorado, Boulder, Colorado 80309 ²University of Denver, Denver Colorado, 80208 ³Supported by NICHD grants HD-11681 and HD-27802

Discrimination of odor differences in MHC-deficient mice.

Kunio Yamazaki¹, G.K. Beauchamp¹, J. Bard², & E.A. Boyse²

We have shown that mice exhibit body odors of exquisite individuality (odortypes) representing polymorphism of Major Histocompatibility Complex (MHC) genes and less potently, of other chromosomal sites (reviewed by E.A. Boyse et al. 1991, in R. Ader, D.L. Felter and N. Cohen, eds., Psychoneuroimmunology-II, Academic Press, San Diego, pp831-846). Perception of odortype likely underlies mate selection, pregnancy success and parental behavior, serving to maintain normal social interactions and to promote genetic diversity through avoidance of inbreeding. Although several lines of evidence suggest MHC genes themselves code for odortypes, incontrovertible evidences of this would focus the search for mechanism directly on these known genes. Thus the purpose of this study was to determine whether mice with targeted gene disruptions (knock-out mice) that do not express one class of MHC genes (termed class I) differ in body odor from otherwise identical animals expressing these genes. Mice trained in a Y maze served as the bioassay. We tested the ability of 5 trained mice to distinguish the urine odor of C57BL/6J-₂m-deficient (-/-) male mice from the urine of control C57BL/6J (+/+) male mice. The results demonstrated that animals significantly distinguished (+/+) from (-/-) and generalized this response to distinguish (+/-) from (-/-). However, in generalization test they exhibited no tendency to discriminate the heterozygote (+/-) from the homozygote (+/+) suggesting class I effects are not quantitative. These data provide additional strong evidence that it is the MHC (class I) genes themselves that mediate MHC-determined differences in odortype.

Address:   Monell Chemical Senses Center 3500 Market Street Philadelphia PA 19104 USA, 215-898-3094, fax 215-898-2084, yamazaki@monell.org

¹Monell Chemical Senses Center, Philadelphia, PA 19104. ²University of Arizona, Tucson, AZ 85724.

Update on IQ Similarity of Unrelated Same Age Siblings²

Nancy L. Segal¹

Genetically unrelated same age siblings, reared together from early infancy, offer a novel research design for studying genetic and environmental influences on behavior. An initial report on IQ findings from this ongoing project appeared in the Journal of Educational Psychology (1997). Participants ranged in age from 4-26 years. The IQ correlation for 21 pairs was .17, the verbal IQ correlation was -.01, the performance IQ correlation was .29, and the mean within-pair IQ difference was 15.38 (SD= 11.45). Interestingly, an increase in performance IQ similarity with age was observed. Additional data have now been collected for 56 pairs, who range in age from 4 - 56 years. IQ data will be reported for the larger sample, and future plans for this ongoing study will be described.

Address:   714-278-2142, 714-278-7134, nsegal@fullerton.edu

¹Psychology Department, California State University, Fullerton, California 92834. ²Supported by the National Science Foundation (SBR-9712875) and a faculty research award from California State University, Fullerton.

A Genetic Analysis of the Relationship Between Low Academic Performance and Early-Onset Substance Use in Adolescent Males²

L.N. Legrand¹, Matt McGue¹, & W.G. Iacono¹

Early initiation into substance use is a consistent, strong predictor of substance-related problems (e.g., D.B. Kandel, 1982, J. Amer. Acad. Child Psychiat., 21, 328-347). As delaying substance-use initiation may thus be an effective preventative measure, establishing the antecedents of early substance use and understanding their etiological components are important first goals. In the present study, we seek to determine the relative genetic and environmental influences upon the relationship between one such antecedent -- low academic performance -- and subsequent early initiation. As part of the Minnesota Twin Family Study, academic performance, intelligence, and level of substance use were assessed in 588 male twins (370 MZ; 188 DZ). Those boys who would go on to try tobacco, alcohol, or marijuana by age 14 had received significantly lower academic grades three years earlier. Effect sizes for academic grades, between nonusers and early users, ranged from .32 to .46 standard deviation units. This effect was independent of IQ. Univariate genetic analyses indicate that additive genetic factors account for 35.2% of the variance in academic performance at age 11 and 21.7% of the variance in substance use at age 14; shared environmental factors account for 17.0% and 40.0% of the variances, respectively. The results of bivariate analyses designed to determine whether the relationship between academic performance and early drug use is environmentally or genetically mediated will be presented. The implications of these results for effective, early intervention will be discussed.

Address:   Department of Psychology, University of Minnesota, 75 East River, Road, Minneapolis MN 55455, llegrand@tfs.psych.umn.edu

¹Department of Psychology, University of Minnesota, Minneapolis, MN 55455 ²Supported in part by NIH grant DAO5147 and the Eva O. Miller Fellowship from the University of Minnesota.

A study of genetic and environmental influences on prosocial and antisocial behaviours.

Christine Pett¹ & Jim Stevenson¹

There has been a tendency for behaviour genetic studies to concentrate on disruptive and antisocial aspects of children's behaviour. By contrast work on more positive and prosocial aspects has been limited. Indeed the issue of whether prosocial and antisocial are influenced by the same or different genetic and environmental issues has yet to be explored. One of the difficulties in undertaking studies on these more prosocial aspects of behaviour has been the lack of well validated measures. Recently Goodman (R. Goodman, 1997, J. Child Psychol. & Psychiat. 38, 582-586.) has published a Strengths and Difficulties scale which meets this need. The parent of a sample of 147 MZ, 133 DZ same sex and 102 opposite sex twin pairs were given the Strengths and Difficulties questionnaire. The data were analysed to examine whether there were genetic influences on prosocial behaviour, whether these effects were more or less marked than for antisocial behaviour or conduct problems and whether prosocial behaviour was influenced by the same factors that impact on more negative aspects of behaviour. Multivariate model fitting was undertaken to test these relationships. Preliminary analyses indicate that there are both shared and independent genetic influences of prosocial and antisocial aspects of behaviour in childhood and adolescence.

Address:   cp5@crpd.psy.soton.ac.uk, Tel: +44 (01703) 594593, Fax: +44 (01703) 594597

¹Department of Psychology, University of Southampton, Hampshire. SO17 1BJ United Kingdom.

The Y chromosome, sexual behavior and pregnancy block in Mus musculus³

Tara L. McLaughlin^1,2, & Benson E. Ginsburg¹

Previous research in our laboratory has focused on strain differences in male sexual behavior and has explored their influence on the occurrence of pregnancy block in newly inseminated females (T.L.McLaughlin and Benson E. Ginsburg,1997,Behav.Genet.26,300). DBA1/J males have been observed to induce higher levels of pregnancy block, elicit greater levels of female defecation and vocalization and display higher levels of attempted mounts than have C57BL6/J males. The present study extends this line of research by examining the role that the Y chromosome may play in these strain differences. Males of the DBA1/c and DBA1/c x DBA1/c.C57BL10 Y congenic strains were tested for their ability to induce pregnancy block and were observed in dyadic encounters with newly inseminated females. Males of these strains, unlike DBA and C57 males, did not differ on any of the behavioral measures. Further, the results suggest that Y-chromosomal differences within Mus musculus may not be sufficient to induce pregnancy block, as measured in this study. These observations, when considered in light of the previously described strain differences between DBA and C57 males, suggest that male behaviors may interact with olfactory cues to influence the occurrence of pregnancy block. The contribution that the Mus musculus Y chromosome makes to this interaction may be too subtle to measure using the present paradigm.

Address:   Department of Psychiatry, Washington University School of Medicine, 40 North Kingshighway #1, St. Louis, MO. 63108 Ph:314-286-2302, Fax:314-286-2213 Email:Tara@matlock.wustl.edu

¹Biobehavioral Sciences Graduate Degree Program, University of Connecticut, Storrs,CT 06269. ²Currently with the Department of Psychiatry, Washington University School of Medicine, St.Louis, MO 63108. ³Supported in part by a Sigma Psi Grant-in-Aid-of-Research.

Obstetrical complications in twin pregnancies

Joan T. Bihun¹, Jackie S. Roessler¹, & H. Hill Goldsmith¹

Prenatal influences and complications can affect twins differently than singletons. The purpose of this presentation is to document the normative developmental course of twin pregnancies and deliveries, and to investigate the role complications might play in determining genetic and environmental influences on subsequent development. Medical records information was analyzed in some detail for 111 relatively healthy mothers (mean= 30.5 years; 95% received some prenatal care) and their twins (mean GA 36.5 weeks, mean weight was 2611 grams) participating in a larger longitudinal study. Extensive norms were collected and analyzed including aspects such as bed rest, blood pressure, preterm labor, types of delivery, and placentation. An obstetrics composite based on a number of pregnancy complications was not related to ratings of infant temperament at 3 months or maternal depression at 3 or 9 months postdelivery. However, specific components of this composite did relate to pregnancy outcomes and indices of later development. For instance, mothers on bedrest delivered, on average, 1.4 weeks sooner than those who were not on bedrest. However, mothers on bedrest were more likely to deliver both twins vaginally. Bedrest, especially when prescribed in the third trimester, tended to relate to greater maternal depression at 9 months postpartum. Multiparous mothers were more likely to rate their twins as more fearful, less likely to orient for longer periods of time, and tended to rate them as less likely to smile or laugh at 3 months postpartum than did primiparous mothers. Interestingly, there were no temperament differences at 3 months between pre- and full-term infants. The results reveal the importance of specific obstetrical complications rather than composites in predicting maternal and infant behavioral outcomes in relatively healthy samples. The data will be used to make more precise estimations of genetic and environmental influences on infant temperament, cognition, and mother-infant interaction.

Address:   1202 W. Johnson St, Madison, WI 53706; 608-265-6250; jtbihun@facstaff.wisc.edu

¹Department of Psychology, University of Wisconsin, Madison, WI 53706

Using Discordant Sibling Pairs to Replicate Behavioral Associations Within Families²

Johnson, J.K.¹, Viken, R.J.¹, & Rose, R.J.¹

Previous studies of singletons have shown that certain personality traits, such as sensation seeking, are associated with alcohol consumption. Such associations are usually interpreted as evidence that the personality traits are individual risk factors for alcohol consumption or abuse. But singleton studies cannot rule out the possibility that such associations are due to between familial factors rather than a functional relationship between personality and alcohol use. We investigated the relationship between Sensation Seeking Scale scores (M. Zuckerman, 1979, Sensation Seeking: Beyond the Optimal Level of Arousal, Lawrence Erlbaum Associates, Hillsdale, NJ) and alcohol consumption in a sample of 1,700 adults, replicating the same associations in individuals between families and in sibling pairs within families. Individuals above the 70th percentile for sex-specific sensation seeking score distributions reported higher quantity, frequency, and density of drinking, as well as more alcohol problems than those scoring below the 30th percentile on sensation seeking. We then attempted to replicate this finding within sibships, providing a more powerful test of the association between these variables. Within the sample were 750 full-sib pairs, from which pairs of discordant sibs (one sib above the 70th percentile and the other below the 30th on sensation seeking scores) were then selected. The resulting sample of discordant pairs consisted of 46 full-sib pairs. Paired t-tests showed that the differences in reported alcohol behaviors between high and low sensation seekers persists in within family comparisons. High sensation seekers reported significantly higher quantity, frequency, and density of alcohol consumption than those with low sensation seeking scores. The finding that the within family association between sensation seeking and alcohol use replicates the association observed in singletons argues against shared familial factors as a full explanation for the association.

Address:   Psychology Building, Indiana University, Bloomington, IN 47405, (812) 855-2311, (812) 855-4544 FAX, jennjohn@indiana.edu

¹Department of Psychology, Indiana University, Bloomington, IN 47405. ² Supported by NIAAA: AA-07611 and AA-00145.

Do genetic and environmental influences on childhood conduct disorder vary across environmental contexts?

Jessica L. Bar¹, Irwin D. Waldman¹, David A. Hay², & Florence Levy³

Conduct disorder (CD) is a serious childhood disorder, characterized by extreme and persistent antisocial behavior. Relatively few behavior genetic studies on childhood CD per se have been conducted. Behavior genetic studies on childhood antisocial behavior and delinquency have yielded mixed results, with many finding substantial genetic influences and some finding appreciable shared environmental influences. One mechanism by which genetic influences may operate is by predisposing a child to be more susceptible to environmental influences. Sociological studies have found correlations between antisocial behavior and various demographic factors (e.g., urban vs. rural residence, low SES, high crime), but the causal status of these factors is ambiguous. Behavior genetic designs can help elucidate the causal mechanisms by which such variables operate. In this study we examined demographic characteristics of neighborhoods and communities to examine the extent to which the genetic and environmental influences on CD vary across environmental contexts. The participants were 2183 4-14 year old male and female same- and opposite-sex twin pairs and their siblings (962 MZ pairs, 963 DZ pairs, and 258 sibling pairs) recruited from the Australian NHMRC Twin Registry, a nationwide, population-based, volunteer registry. Information on zygosity was based on mothers' reports of their twins' physical similarity. The mothers of twins filled out a questionnaire assessment of 6 DSM-III-R CD symptoms. We used multiple regression analyses of twin and sibling data to estimate genetic and environmental influences on CD symptoms and their moderation by environmental context variables. The implications of our results for the causal role of demographic factors in childhood antisocial behavior will be discussed.

Address:   Department of Psychology, Emory University, Atlanta, GA, 30322, phone - (404) 727-2741, fax - (404) 727-1284, e-mail - jbar@emory.edu

¹Department of Psychology, Emory University, Atlanta, GA 30322. ²Department of Psychology, Curtin University, Perth, Western Australia, Australia. ³Department of Child and Adolescent Psychiatry, The Prince of Wales Children's Hospital, High Street, Randwick, NSW 2031, Australia.

Home cage activity in mice selected for high- and low open-field ambulation³

Norman D. Henderson¹, T. Tritto², D. W. Fulker², and J. C. DeFries².

DeFries, Gervais, and Thomas, 1978, Behavior Genetics,8, 3-13, used two replicate sets of F3 mice derived from a BALB/cJ ´ C57BL/6J crossto select for High- or Low open-field (OF) activity. Following 30 generations of bi-directional selection, the resulting replicate lines were maintained by within-line random mating for generations 31-48 and then inbred by brother-sister mating. At the time of this study the lines were at generation 90. Using a newly developed infra-red detector motion detector, we continuouslymeasured gross locomotor activity in the home cages of mice from both sets of High- and Low-OF Activity lines and their F1 crosses. In one set of replicate lines, the High-OF mice showed significantly higher levels of motor activity in their home environment during the dark phase of the diurnal cycle than did the Low-OF line. The High and Low lines differed by only one phenotypic SD, however, which was about 1/7 the size of the OF activity difference observed at generation 30. This H-L home cage activity difference is well within the limits expected by genetic drift alone among inbred lines (Henderson, 1997, Behavior Genetics, 27, 145-154). In the second replicate, High- and Low-OF lines showed nearly identical levels of home cage activity, whereas their OF activity at generation 30 differed by over 8 phenotypic SDs. In both replications, the F1 hybrid cross was more active in home cages during the dark phase than either of their parent lines. No significant line differences in home cage activity were observed during light-phase. These results suggest that the level of locomotor activity exhibited in a brightly lit OF is largely genetically independent of activity levels in familiar home-cage environments.

Address:   Department of Psychology, Oberlin College, Oberlin, OH 44074, Phone: (440) 775-7695, FAX: (440) 775-8356, e-mail: fhenders@oberlin.edu

¹ Department of Psychology, Oberlin College, Oberlin, OH 44074, ² Institute for Behavioral Genetics, University of Colorado, Boulder, CO 80309, ³ Supported by NIMH Grant MH-53480.

Genetic disposition for global maturity: An explanation for genetic effects on parental reports of ADHD³

Birgitta Steffensson¹, Jan-Olov Larsson², Ingegärd Fried², Eman ElSayed², Per-Anders Rydelius², & Paul Lichtenstein¹

Attention Deficit Hyperactivity Disorder (ADHD) has been shown to be substantially influenced by genetic factors, however it is still not known how these effects are mediated. The aim of this study was to evaluate whether genetic disposition for global maturity could be a mediator of the well-known findings of genetic factors for ADHD-symptoms.Through the population based Swedish Twin Registry, 1,480 pairs of twins aged 8-9 years were identified. A mailed questionnaire with items from DSM-III-R for ADHD and items concerning global maturity were completed by the parents. Genetic effects explained 68% of the variance in ADHD for girls, and 35% for boys. Global maturity mediated a portion of genetic effects for ADHD for both boys and girls. There were also shared environmental, as well as non-shared environmental effects unique to ADHD not in common with global maturity.The results show that there are at least two pathways in which genetic effects can influence ADHD; one through predisposition for maturational lag, indicating that maturational lag could be of importance for treatment and prognoses of ADHD, and one unique to ADHD, possible mediated by personality.

Address:   Birgitta Steffensson, Division of Genetic Epidemiology, Institute of Environmental Medicine, Karolinska Institute, Box 210 S-171 77 Stockholm, Sweden. Telephone:+46-8-728 74 30. Telefax: +46-8-30 45 71. E-mail: Birgitta.Steffensson@imm.ki.se

¹Division of Genetic Epidemiology, Institute of Environmental Medicine, The Karolinska Institute, Stockholm, Sweden. ² Department of Child & Adolescent Psychiatry, The Karolinska Hospital, Stockholm, Sweden. ³Supported by the Swedish Society of Medicine.

The importance of environmental influences on scholastic achievement in Bedouin full and half siblings

Yoon-Mi Hur¹, S. Elbedour², Hasan Abu Saad³

Behavioral genetic studies of scholastic achievement in European samples suggest that genetic factors account for 30 to 70% of variations in scholastic achievement. Few studies, however, have investigated the genetic and environmental influences on academic achievement in non-European samples. The present study examined three measures (Arabic, English, and Mathematics) of academic achievement in 612 Bedouin full- and half-siblings in the Negev desert of Israel, ranging in age from 8 to 20 years. The Bedouin Arabs in the Negev region are a nomadic and semi-nomadic population. They have a long tradition of multiple marriages, and both full- and half-siblings live in intact families under the same conditions. School education is a low priority in this society; the illiteracy rate is known to be high. The majority of the children spend their time working in the home and caring for the family's sheep stock. Scores on achievement tests in Arabic, English, and mathmatics were corrected for gender. To determine the importance of genetic and environmental influences on the three measures, maximum likelihood correlations of full- and half-siblings were compared, and a standard behavioral genetic model was applied to the raw pedigree data using an estimation procedure for variable pedigree sizes in Mx. Contrary to the findings from European samples, variation in the three measures of scholastic abilities were influenced predominantly by environmental factors and not by genetic factors. A previous study of this population (S. Elbedour, T. J. Bouchard, Jr., & Y-M Hur, 1998, Intelligence, 25, 71 - 82) indicated that variation in general mental ability is also largely environmental in origin. It is speculated that under their extremely disadvantaged environmental conditions, genetics are not an important causal factor for individual differences in IQ and scholastic achievement.

Address:   Department of Psychology, Elliott Hall, 75 East River Road, University of Minnesota, Minneapolis, MN 55455-344

¹Department of Psychology, University of Minnesota, Twin Cities, MN 55455, ²Ben-Gurion University of the Negev, Israel, ³Tel-Sheva high school, Israel

Sources Of Individual Variation Contributing To Change In Cognitive Abilities³

Chandra A. Reynolds¹, Margaret Gatz¹, Nancy L. Pedersen²

Both stability and change in cognitive abilities in older adults have been documented in recent twin studies. The present study extends previous work by applying biometrical growth models that account for changes in level and rate of change in a sample of 788 twins from the Swedish Adoption/Twin Study of Aging (Pedersen, N. L., McClearn, G. E., Plomin, R., Nesselroade, J. R., Berg, S., and deFaire, U., 1991, Acta Genetica Medicae et Gemmellologiae, 40, 7-20) and the Study of Dementia in Swedish Twins (Gatz, M., Pedersen, N. L., Berg, S., Johansson, B., Johansson, K., Mortimer, J. A., Posner, S. F., Viitanen, M., Winblad, B., and Ahlbom, A.,1997, Journal of Gerontology: Medical Sciences, 52A, M117- M125). The models considered allow for testing effects and make use of methods that allow for missing data. Cognitive tests were administered to each individual from one to six times between 1985 and 1995 with either three years or 18 months between testing occasions. Initial analyses indicate lower familiality for rates of change than for level across six cognitive tests: WAIS Information, Koh's Block Design, Digit Span (Forward and Backward), Thurstone's Picture Memory, and Symbol Digit. Of the tests considered, Symbol Digit exhibited the largest familiality for rate of decline.

Address:   Department of Psychology, University of Southern California, Los Angeles, CA 90089-1061., E-mail: reynold@rcf.usc.edu; Phone: (213) 740-0864; FAX: (213) 746-5994.

¹Department of Psychology, University of Southern California, Los Angeles, CA 90089-1061. ²Division of Epidemiology, Institute of Environmental Medicine, The Karolinska Institute, P.O. Box 210, 17177 Stockholm ³Supported by NIH grants AG10175, AG04563 and AG08724.

Common Genetic Influences on General Cognitive Ability in Males and Females: A Twin Study²

Valerie S. Knopik¹ & John C. DeFries¹

Although the average scores of males and females are highly similar on various tests of general cognitive ability, variances tend to be somewhat higher in males (L. V. Hedges and A. Nowell, 1995, Science 269, 41-45). Thus, the etiology of individual differences in general cognitive ability may differ in males and females. In the present study, this hypothesis was assessed by fitting a structural equation model of sex-limitation to full-scale IQ data from monozygotic (MZ), same-sex dizygotic (SSDZ), and opposite-sex dizygotic (OSDZ) twin pairs tested as control subjects for the Colorado Learning Disabilities Research Center. A total of 426 twin pairs (92 MZ males, 119 MZ females, 68 SSDZ males, 60 SSDZ females, and 87 OSDZ pairs) were used in the analysis. When the full model was fitted to the data, resulting estimates of heritability did not differ significantly in males and females (0.69 and 0.41, respectively), and corresponding estimates of the variance due to shared-environmental influences (0.10 and 0.32) were not significantly different from zero. Moreover, the estimate of the between-sex genetic correlation was 1.00 (0.61-1.00, 95% confidence interval). Thus, individual differences in general cognitive ability appear to be due substantially to common genetic influences in males and females.

Address:   Valerie S. Knopik, Institute for Behavioral Genetics, Campus Box 447, University of Colorado, Boulder, CO 80309, Telephone: 303/492-7362, FAX: 303/492-8063, E-Mail: Valerie.Knopik@Colorado.EDU

¹Institute for Behavioral Genetics, University of Colorado, Boulder, Colorado 80309. ²Supported in part by NICHD grant HD-27802, HD-11681, and NIMH grant MH-16880.

Genetic and environmental influences on the relationship between attachment to family, attachment to school, and adolescent depressed mood

Kristen C. Jacobson¹, & David C. Rowe²

The present study examined genetic and environmental influences on the relationship between attachment to family, attachment to school, and adolescent depressed mood. Data are from the National Longitudinal Study of Adolescent Health (Add Health), a large, nationally representative study of adolescent health and health related behaviors. The Add Health dataset includes a subsample of sibling pairs (e.g., MZ twins, DZ twins, full siblings, half siblings, and unrelated siblings). The statistical package Mx was used to estimate the genetic and environmental influences on attachment to family, attachment to school, adolescent depressed mood and the correlations between these variables. Furthermore, sex differences were also examined. Results from the best fitting model revealed some sex differences. Specifically, genetic factors accounted for approximately one-half the variation in attachment to family and attachment to school among females, but explained only one-third of the variance for males. In contrast, genetic factors were somewhat larger for adolescent depressed mood for males than for females, although non-shared environment explained the majority of variation in depressed mood for both sexes. Significant shared environmental influences were found for boys' reports of attachment to family and attachment to school, but not for females. For both males and females, genetic factors explained approximately one-half of the covariation between attachment to family and attachment to school, with non-shared environmental factors explaining the remainder. Likewise, genetic factors and non-shared environment contributed equally to the covariation of attachment to family and attachment to school with depressed mood for females. However, the covariation between these variables for males could be explained entirely by common non-shared environmental factors.

Address:   221-B FCR Building, PO Box 210033, The University of Arizona, Tucson, AZ 85721-0033, (520) 621-7127, (520) 621-3401 (fax)

¹Department of Human Development & Family Studies, The Pennsylvania State University, University Park, PA 16802 ²Department of Family Studies, The University of Arizona, Tucson, AZ 85721

IQ and Academic Achievement in Twins: A Longitudinal Study².

Marcie Chambers¹ & D. W. Fulker¹.

A Cholesky decomposition model was used to analyze IQ and academic achievement in twin pairs. IQ was measured by the Bayley Scales of Infant Development (New York: Psychological Corp) at ages 1 (N=717) and 2 (N=661), Stanford-Binet Intelligence Scale (Test Editorial Offices: Iowa City, Iowa) at ages 3 (N=611) and 4 (N=609), and the WISC-R (Wechsler Intelligence Scale for Children - Revised, Psychological Corporation, New York) at age 7 (N=606). Academic achievement was assessed by a teacher report of school performance for reading and math (Manual for the Teacher's Report Form and Teacher Version of the Child Behavior Profile, Department of Psychiatry, University of Vermont, Burlington) at ages 7 (N=425) and 9 (N=210). A multivariate analysis was carried out to estimate the additive genetic, common environmental, and individual environmental effects for the associations between the variables. The results from the most parsimonious models indicate the additive genetic component present at age 1 affects IQ up to age 4, accounting for 14% to 53% of the variance. A second genetic component present at age 3 affects IQ and academic achievement through age 9, accounting for 16% to 83% the variance. A third genetic component comes in at age 7 affection only academic achievement at ages 7 and 9, accounting for 14% to 55% of the variance. Common environment has little or no effect after age 2. Individual environment is best represented by age specific factors. Most, if not all, of the covariation between IQ and academic achievement can be accounted for by genetic affects.

Address:   Campus Box 447 Boulder, CO 80309-0447 phone 303-492-1235 fax 303-, 492-8063 e-mail marcie.chambers@colorado.edu

¹Institute for Behavioral Genetics, University of Colorado, Boulder, CO 80309. ²Supported by NIH Grants HD10333, HD18426, and MH43899, and from the John D. and Catherine T. MacArthur Foundation.

Is there a genetic basis for sex differences in social cognition?

Jim Stevenson¹

It has been suggested that there is a genetic mechanism whereby males may be disadvantaged in social cognition (D.H.Skuse et al.,1997, Nature, 387, 705-708). That study found that girls with Turner's syndrome receiving their single X-chromosome from their fathers had higher scores on a measure of social cognition than those having a maternal X-chromosome. It was suggested that there is an imprinted gene on the X chromosome influencing this ability which is not expressed if transmitted on the maternal X chromosome. If this is correct, it would be expected that the MZ correlation for females on this social cognition measure would be lower than that for MZ males pairs. Using a sample of 147 MZ, 133 DZ same sex and 102 opposite sex twin pairs this proposed X-linked genetic mechanism was examined. Although replicating the male deficit in this aspect of social cognition, there was no difference in the similarities within MZ male (r = 0.823) and female (r = 0.833) pairs. Sex-limitation models were fitted to these data to test whether the genetic correlation between males and females had been lowered by this imprinted X-linked locus. The best model was the common effects sex limitation model (² = 31.73, df = 11, p< .001). This model had common additive genetic and nonshared environmental influences with h² = 0.78 and e² = .22 for both males and females. The significant improvement over an AE model that constrained all terms to be equal for males and females (² = 40.66, df = 13, p< .001) was a consequence of accounting for variance differences between the sexes. There is no evidence from this twin study to support different genetic contributions to social cognition in normal males and females.

Address:   Dept. of Psychology, University of Southampton, Highfield, Southampton SO17 1BJ., Tel: +44 1703 592583 Fax: +44 1703 594719 email : jsteven@psy.soton.ac.uk.

¹Department of Psychology, University of Southampton, Southampton, UK.

Genetics of early cancer detection behaviors in Australian female twins

Susan A. Treloar¹, Christine McDonald¹, Nicholas G. Martin¹

Early detection of cervical and breast cancer comprises important components of women's health strategy. Screening programs and preventive behaviors such as breast self-examination provide the means to this end. Our twin study sought to identify the relative influence of environmental and genetic factors on `liability' to early cancer detection behaviors, including use of cervical (Pap) smear tests, mammograms, and breast examination. Additive genetic and random environmental effects models gave the best, most parsimonious fit to the data for each early cancer detection behavior. The heritability of liability to Pap smear use was 66%, mammogram use 50%, breast examination by a doctor 38% and breast self-examination 37%. Genetic influences were behavior-specific; there was no evidence for a common genetic influence on the four behaviors. Potential covariates investigated included age, amount of contact between twins, educational level and personality traits such as harm avoidance, novelty seeking, reward dependence, neuroticism, anxiety, depression, self-esteem, perceived control, interpersonal dependency and ways of coping. None were significant. The importance of women's own agendas is highlighted. The study was carried out before the implementation of media campaigns to encourage awareness and increase participation rates. Hence follow-up investigation, including data on regularity of behaviors, would be informative.

Address:   Queensland Institute of Medical Research, Brisbane QLD 4029, Australia, Phone: + 61 7 3362 0229 Fax: + 61 7 3362 0101 Email: sueT@qimr.edu.au

¹Queensland Institute of Medical Research, Brisbane, Australia

Etiological validation of a developmental taxonomy of antisocial behavior

Irwin D. Waldman¹, David A. Hay², & Florence Levy³

It has long been recognized that there is meaningful heterogeneity in childhood and adolescent antisocial behavior (ASB). Recently, developmental psychopathology researchers have focused on age-of-onset as an important source of the heterogeneity in ASB. Children with early-onset ASB appear to have a more stable and pervasive condition with a worse prognosis than children with late-onset ASB. In addition, it has been posited that early-onset ASB is more biologically-based than late-onset ASB, which is posited to be influenced more by environmental risk factors. In this paper, we examine genetic and environmental influences on four clusters of antisocial behaviors – oppositionality, aggression, property violations, and status violations – in a sample of 1534 twin pairs (790 MZ and 744 DZ) and 298 sibling pairs, ages 4-19. These four clusters were shown to differ in median age-of-onset and were isolated using multi-dimensional scaling in a meta-analysis of factor analytic studies of Oppositional Defiant Disorder (ODD) and Conduct Disorder (CD) symptoms. In univariate behavior genetic analyses, a model containing additive genetic and shared and non-shared environmental influences fit best for all four ASB symptom clusters. Heritability was highest for oppositionality and decreased monotonically for aggression, property violations, and status violations. In contrast, shared environmental influences were lowest for oppositionality and increased monotonically for aggression, property violations, and status violations. Multivariate behavior genetic analyses suggested that the monotonically decreasing heritabilities were due predominantly to a common genetic factor, whereas the monotonically increasing shared environmental influences were due predominantly to a common shared environment factor. The implications of these results for a developmentally-informed taxonomy of antisocial behavior are discussed.

Address:   mailing address - Department of Psychology, 532 N. Kilgo Circle, Emory University, Atlanta, GA 30322, phone - (404) 727-7430, fax - (404) 727-0372, e-mail - waldman@fs1.psy.emory.edu

¹Department of Psychology, Emory University, Atlanta, GA 30322. ²Department of Psychology, Curtin University, Perth, Western Australia, Australia. ³Department of Child and Adolescent Psychiatry, The Prince of Wales Children's Hospital, High Street, Randwick, NSW 2031, Australia.

The heritability of personality factors in zoo chimpanzees Pan troglodytes³

Alexander Weiss¹, J.E. King²

Factor analytic studies on zoo housed chimpanzees have shown six reliable personality factors, five of which resemble the human Big Five while the sixth is a chimpanzee specific dominance factor (King, J.E. and Figueredo, A.J., 1997, Journal of Research in Personality, 31, 257-271). A recent field study showed that dominance is related to higher fertility and infant survival in chimpanzees (A. Pussey, J. Williams, and J. Goodall, 1997, Science, 277, 828-831). Personality factors in humans have a broad-sense h² between .40 and .50. Almost all of the remaining variance is accounted for by the unshared environment (Loehlin and Rowe, 1992 cited in Rowe, D.C., 1994, The Limits of Family Influence, The Guiliford Press, New York). We calculated the heritability of six chimpanzee personality factors. We were especially interested in the dominance factor because of its established relationship to fitness. Personality data were collected on 145 zoo housed chimpanzees. Because of the complex network of relatedness within this population, we used the Symmetric Differences Squared (SDS) method to calculate h², c², and e². SDS regresses the squared diffferences between the phenotypes of animals in all possible pairs onto 1-r_g and any environmental variable. In our case, the latter term was determined by whether the pair were in same or different zoos. Dominance had substantial heritability. This is the first study showing that subjective personality ratings in chimpanzees may be heritable. It also supports our hypothesis that fitness benefits of high dominance would enhance this factor's heritability.

Address:   Alexander Weiss, Department of Psychology, University of Arizona, Tucson, AZ 85721, 520-881-2389, aweiss@u.arizona.edu

¹Department of Psychology, University of Arizona, Tucson, AZ 85721. ²Department of Psychology, University of Arizona, Tucson, AZ 85721. ³Supported by ChimpanZoo and the Jane Goodall Institute.

Developmental genetic analysis of conduct disorder in adolescent twins²

Hermine H.M. Maes¹, Michael C. Neale¹, Judy L. Silberg¹, Lindon J. Eaves¹

The prevalence of conduct disorder (CD) increases during adolescence, particularly in males. Various forms of the disorder have been described, "adolescent-limited" and "life course persistent" (T.E. Moffit, 1993, Psychol Rev, 100:674-701). We hypothesize that their genetic etiologies are distinct. In the Virginia Twin Study of Adolescent Behavioral Development, data on conduct disorder symptoms were collected using the Child and Adolescent Psychiatric Assessment (CAPA) from 1,226 male twins between the ages of 8 and 16. Second wave data were available on 929 twins. Raw data on one or two-year interval data were used to fit developmental genetic models, including common factor, simplex and growth curve models, allowing for age changes in genetic effects on rate of change in CD symptoms over the age range. A linear growth curve model provided the most parsimonious fit to the data. This model included genetic and environmental factors on the intercept and slope, as well as on the residual variances. The overall picture of the genetic control of developmental change in CD is consistent with Moffit's model. We found two genetically distinct components affecting the development of CD. The linear component, has initially small genetic effects that increase with age ("life-course persistent"). The second component creates a burst of genetic variability in mid-adolescence ("adolescent-limited").

Address:   Virginia Institute for Psychiatric and Behavioral Genetics, Virginia Commonwealth University, Box 980003, Richmond, VA 23298. Tel: 804 828 8145 Fax: 804 828 8801 Email: hmaes@hsc.vcu.edu Web URL: http://www.vipbg.vcu.edu

¹Virginia Institute for Psychiatric and Behavioral Genetics, Virginia Commonwealth University, Box 980003, Richmond, VA 23298. ²Supported by NIH Grants MH45268 and RR08123

Developmental isolation and subsequent adult behavior of Drosophila paulistorum: Courtship³

Lee Ehrman¹, & Y.-K. Kim²

In a series of investigations into the development of discriminatory behavior in the D. paulistorum semispecies complex (Y.-K. Kim, L. Ehrman, and H. R. Koepfer, 1992, Behav. Genet. 22, 545-556; Y.-K. Kim, L. Ehrman, and H. R. Koepfer, 1996, Behav. Genet. 26, 15-25; Y.-K. Kim, H. R. Koepfer, and L. Ehrman, 1996, Behav. Genet. 26, 27-37), differences in courtship between control males and socially-isolated males (since egg deposition), representing each of the six D. paulistorum semispecies, were observed, and responses of females toward courtship from these two types of males scored. As controls, courtship elements and sequences did not unequivocally identify any given semispecies. However, for each element, there were consistent quantitative differences among semispecies: These six semispecies were significantly different from each other for all seven male courtship elements and for all four female responses, except standing. Developmental isolation was apparently sufficient enough to change courtship activities and mating performance: Socially-isolated males displayed more total courtship than did controls. Further, when mate competition for virgin females was observed between socially-isolated males and controls, such high levels of sexual activities led to impressive mating successes. Females simply preferred males that were more sexually active, in these instances, the socially-isolated ones. However, socially-isolated males from all six semispecies displayed homosexual behavior toward other males in our observation chambers. These findings suggest the possibility of changes in D. paulistorum brain structure. Thus, experience-dependent modification of brain structure (G. M. Technau, 1984, J. Neurogenet. 1, 113-126; A. Balling, G. M. Technau, and M. Heisenberg, 1987, J. Neurogenet. 4, 65-73; M. Heisenberg, M. Heusipp, and C. Wanke, 1995, J. Neurosci. 15, 1951-1960) during the development of individuals isolated from conspeci fics will be observed.

Address:   Division of Natural Sciences, State University of New York, 735 Anderson, Hill Road, Purchase, NY 10577, U. S. A.; TEL: 914-251-6671; FAX:, 914-251-6635; E-mail: lehrman@purvid.purchase.edu

¹Division of Natural Sciences, State University of New York, Purchase, NY 10577, U. S. A. ²Department of Genetics, University of Georgia, Athens, GA 30602, U. S. A. ³Supported by Whitehall Foundation

Psychophysiology and Genetics in Twins

Laura A. Baker¹

Psychophysiological measures have been used extensively as biological markers in the study of mental processes, risk for psychiatric disorders, antisocial behavior, and reactions to stress. Yet, relatively little is known about the genetic and environmental underpinnings of most variables derived from a psychophysiological laboratory testing. The present study of twins attempted to further our understanding of the heritable and non-heritable bases of a number of psychophysiological indicators. We examined electrodermal and cardiac measures of autonomic arousal, during both rest and task situations, in 150 pairs of 9-16 year old twins. Genetic variability was found to be particularly important for baseline (i.e., during rest) measures of heart rate, skin conductance level, and non-specific skin conductance responses. Genetic variability was lower, albeit still significant, for these same variables measured while the children engaged in the continuous performance task or while listening to mild tones (i.e., during tasks). Moreover, reactivity (change between rest and task) was heritable only for electrodermal measures, including SCL and NSCR, but not for cardiac ones. Environmental factors shared by the twins were also important for electrodermal measures, which are especially sensitive to seasonal variations. These results emphasize the need for further research in order understand how genes and environment may mediate psychophysiology/behavior relationships, which will be discussed.

Address:   Department of Psychology, SGM 501, University of Southern California, Los Angeles, CA 90089-1061, Phone: (213) 740-2261 FAX: (213) 746-9082 email: lbaker@usc.edu, internet: http://www-rcf.usc.edu/~lbaker/

¹Department of Psychology, University of Southern California, Los Angeles, CA 90089-1061

Alcohol preference in mice: Replication using genotypic selection and an F₄ advanced intercross.

Lisa M. Tarantino ^1,2,4 & Gerald McClearn^1,2,3,4

A substantial body of literature has addressed genetic factors in predisposition to alcoholism. Genetic studies in humans are, of course, confounded by other factors such as environmental and social circumstances. Animal models offer a more direct way of studying the genetic influences on alcohol-related processes while controlling for these other factors. It was discovered in the 1950s that inbred strains of mice differ in voluntary consumption of alcohol (McClearn & Rodgers, 1959, Q J Stud Alcohol 20, 691-695). Since then, researchers have capitalized on the differences between two of these strains, C57BL/6 and DBA/2, in an effort to gain an understanding of the genetic factors underlying alcohol-related behaviors. Utilizing BXD recombinant inbred (RI) mice, we have previously nominated several chromosomal regions, called quantitative trait loci (QTLs), which may contain genes associated with alcohol consumption (Rodriguez et al., 1995, Alcohol Clin Exp Res 19, 367-373). This study was the first step in a multi-stage research program which included subsequent analysis of an F₂ intercross population to confirm the nominated regions. Four of these regions have been replicated (Tarantino et al., in press, Alcohol Clin Exp Res). We have recently completed a genotypic selection experiment for two of the regions on Chrs 2 and 4. The Chr 2 QTL has replicated again while the Chr 4 QTL, although in the right direction, did not yield a significant result. We have also analyzed 218 F₄ animals in an attempt to narrow the regions implicated in both the RI and F₂ study. Once again we have replicated the results from previous studies while also indicating that even an early generation of an advanced intercross can narrow QTL regions significantly.

Address:   101 Amy Gardner Building, University Park, PA 16802; Ph: 814-865-1717;, Fax: 814-863-4768; e-mail: lisa@maustrap.cdhg.psu.edu

¹Center for Developmental and Health Genetics, ²Intercollege Graduate Program in Genetics ³Department of Biobehavioral Health ⁴The Pennsylvania State University ⁵Funding provided by AA-08125 and DA-07277

Stability and change in cognitive functioning in late adulthood³

Scott M. Hofer¹, S. Berg², B. Johansson², G. E. McClearn¹

The development and aging of cognitive capabilities are sensitive to many influences throughout the lifespan. In a recent study (G. E. McClearn, B. Johansson, S. Berg, N. L. Pedersen, F. Ahern, S. A. Petrill, & R. Plomin, 1997, Science,276n, 1560-1563), specific and general cognitive abilities were found to remain substantially heritable in late life. Previous findings from other longitudinal twin studies indicate that genetic effects contribute more to stability than to change throughout adulthood. In this study, we investigated the relative contributions of environmental and genetic factors to longitudinal change and continuity in specific and general cognitive functioning. The focus of these analyses was on the concordance between twins in intraindividual variability rather than on interindividual differences in cognitive performance. These analyses were performed on data from the OCTO-Twin study of octogenarian and nonagenarian twins measured at three occasions with two-year inter-occasion intervals. Multivariate analyses of cognitive change included health status and health behaviors, sensory acuity, and measures of lifestyle. Models with missing data due to attrition (due mainly to mortality) were estimated assuming the data were "missing at random" (e.g., including covariates and previous outcome variables associated with dropout status). In general, the findings concur with previous results of substantial stability in the genetic factors associated with cognitive functioning.

Address:   Center for Developmental and Health Genetics, The Pennsylvania State University, 101 Amy Gardner House, University Park, PA 16802, P: (814) 865-1717 F: (814) 863-4768, smhofer@cdhg.psu.edu

¹Center for Developmental and Health Genetics, Pennsylvania State University, University Park, PA 16802 ²Institute of Gerontology, University College of Health Sciences, Box 1038, S-551 11 Jonkoping, Sweden. ³Supported by NIA Grant AG08861.

Extremes in cognitive functioning in the oldest-old: A DeFries-Fulker analysis of non-demented octogenarian twins.

Stephen A. Petrill¹, Boo Johansson², Stig Berg²Robert Plomin³, & Gerald E. McClearn⁴

Numerous studies have examined the genetic and environmental influences on cognitive ability in the elderly. Many of these studies have attempted to explain either the individual differences across the range of ability, or the genetic and environmental etiology of groups selected for dementia. The current study examines intermediate groups of non-demented twins selected for low general cognitive ability. The current study is based upon the more than 350 80+ year-old twin pairs participating in the OctoTwin Study. Low-g groups were formed at or below the 10%tile, 20%tile, 30%tile, 40%tile. DeFries-Fulker DF (J. C. DeFries, and D. W. Fulker. 1985. Behav. Genet. 15, 467-473; 1988. Acta Genetica, Med Gemellol, Twin Res. 37, 205-216.) analyses suggest that genetic influence is nonsignficant in groups at or below the 40%tile, dropping from a group heritability of .51 at the 40%tile, to .22 at the 30%tile, to .00 at or below the 20%tile. As a comparison, groups at or above the 60%tile, 70%tile, 80%tile, and 90%tile were also ascertained. In contrast to the low-g results, DF analyses above the population mean suggest statistically significant genetic influences, with group heritability exceeding .74 in all groups selected at or above the 60%tile. These results suggest that general cognitive ability may be genetically influenced in the oldest-old above the population mean, but may be more environmentally influenced below the population mean. Implications are discussed.

Address:   Department of Psychology, Wesleyan University, 207 High Street, Middletown, CT 06459-0408, (860) 685-2602 phone, (860) 685-2761 fax, spetrill@wesleyan.edu

¹Department of Psychology, Wesleyan University ²Institute of Gerontology, Jonkoping, Sweden ³Social Genetic and Developmental Psychiatry Research Centre, Institute of Psychiatry, London ⁴Center for Developmental and Health Genetics, The Pennsylvania State University ¹Supported by NICHD Postdoctoral Fellowship (HD-27088) at the Pennsylvania State University The OCTO Twin Study is supported by a grant form the National Institute on Aging (NIA: AG 08861).

Cognitive impairment and dementia in the oldest old twins

Boo Johansson¹, Birgitta Andersson¹, David Karlsson^1,2, & Gerald E. McClearn³

In a population-based study, like OCTO Twin, we face the entire range of memory and cognitive functioning among the oldest old; from those with preserved abilities to individuals with severe dementia. In addition, the longitudinal design identifies new cases with dementia. This presents problems for standardized psychometric methods in scoring performances (such as zero-scorers') that require considerations in data analysis. So far, we have employed a conservative approach in which OCTO twins suspected of dementia have been excluded in the analyses (e.g., McClearn et.al., 1997). Dementia also present problems for interpretations of heritabilities within the cognitive domain in late life. In OCTO Twin a diagnostic workup was conducted in those suspected for dementia. This routine includes a neuropsychological workup, a review of medical records, and a diagnosis conference. Diagnoses are determined using current standards and cases are followed to autopsy. The outcomes from the diagnostic work-up in the first two waves of the five measurement occasion OCTO Twin study are presented in the current study. In the first wave, including only complete pairs, 89 twins were identified as suspects of dementia. The criteria for a dementia diagnosis (DSM-III-R) were met for eighty-one. For main type of dementia, 53% met criteria for Alzheimer's (DAT) and 30% for dementia of vascular type (VaD). The percentages for incident cases at the 2-year follow-up were nearly identical. At both occasions, concordant pairs were typically found among MZ twins with a DAT diagnosis. The concordance rate for MZ twins with DAT however increased to the follow-up, while it remained in the DZ twins. These preliminary findings demonstrate the need for a longitudinal design in studying genetic influences on dementia in late life.

Address:   Box 1038, SE-551 55 Jonkoping, Sweden, Phone: + 46 - 36 -32 49 00, Fax: +46 - 36 -32 49 16, Boo.Johansson@hhj.hj.se, ACKNOWLEDGEMENT:, The OCTO Twin Study is supported by NIA (AG 08861).

¹Institute of gerontology, University College of Health Sciences, Jonkoping, Sweden ²Department of Geriatric Medicine, Värnamo hospital, County of Jönköping, Sweden ³Center for Developmental and Health Genetics, Pennsylvania State University, USA

Genetic and environmental influences on health-control beliefs in the oldest-old⁵.

Julia D. Grant^{1, 2}, Boo Johansson^{2, 3}, & Gerald E. McClearn^{2, 4}.

Much research, stemming primarily from Rotter's work (J.B. Rotter, 1966, Psychological Monographs, 80, 1-28), has indicated that beliefs about personal control can influence behavior. One topic that has received substantial attention in recent years is perceived control over health. Although numerous researchers have used the Multidimensional Health Locus of Control scale (MHLC; K.A. Wallston, B.S. Wallston, and R. DeVellis, 1978, Health Education Monographs, 6, 160-170) to examine health-control beliefs and their associations with other health and personality measures, little work has been done investigating health locus of control in individuals over the age of 80 years. Furthermore, the influence of genetic and environmental factors on health-control beliefs has not been systematically examined in previous studies. The Swedish OCTO-Twin study provides the opportunity to investigate genetic and environmental influences on attributions of control over health. In the present analyses, we examined health-control beliefs in a sample of 94 octogenarian twin pairs (MZ=44, like-sex DZ=50; mean age=82.6 years). Factor analysis, with a three-factor solution specified, produced the same factors originally proposed by Wallston and colleagues. All three subscales had acceptable internal consistency (alphas from .72 to .76). Associations between health-control beliefs and other health measures (e.g., self-rated health, hospital visits, exercise) were modest. Intraclass correlations indicated that genetic influences were present for the Chance and Powerful Others subscales, but not for the Internal subscale. Quantitative genetic analyses confirmed significant genetic influences on the Chance subscale (h²=.64) as well as significant familiality on the Powerful Others subscale (h²=.29).

Address:   Department of Psychiatry, Washington University School of Medicine, 40 N., Kingshighway, Suite 1, St. Louis, MO 63108, Phone: (314) 286-2299, Fax: (314) 286-2213, Email: julie@matlock.wustl.edu

¹Department of Psychiatry, Washington University School of Medicine, 40 N. Kingshighway, Suite 1, St. Louis, MO 63108 ²Center for Developmental and Health Genetics, 101 Amy Gardner House, PSU, University Park, PA 16802 ³Institute for Gerontology, Halsohogskolan, Box 1038, S-551 11 Jonkoping, Sweden ⁴Department of Biobehavioral Health, 315 East HHD, PSU, University Park, PA 16802 ⁵OCTO-Twin is supported by Grant AG08861

Monozygotic twins increase power of genetic association studies of complex phenotypes

Michael B. Miller¹

The value of selected pairs of dizygotic (DZ) twins for linkage analysis is widely recognized. Monozygotic (MZ) twins are not useful for linkage analysis because they are genetically identical at every locus. However, the special value of MZ twins for studies of genotype-phenotype association in genetically complex traits has not been adequately appreciated or exploited. For quantitative traits, the strength of genotype-phenotype association (and the statistical power of the association method) is determined by the proportion of variance in the trait that is contributed by the single locus under investigation (i.e., the heritability of the locus = h₁²). We can effectively increase this heritability by using mean scores of pairs of MZ twins as the phenotype instead of using scores of singletons. Assuming no epistasis at the locus under consideration, when the overall heritability of a score is h², the locus heritability of the mean score for a pair of MZ twins is given by h₁² 2/(1+h²+c²) which is always greater than h₁² and less than 2h₁². Thus the power of association studies is improved by using mean scores for MZ pairs instead of scores of singletons, but genotyping costs hold constant. Association analysis of binary (affection status) traits can be made more powerful by using concordant MZ pairs instead of singletons. This is true regardless of method (case control, haplotype relative risk or transmission disequilibrium test [TDT]). Risch and Merikangas (1996, Science, 273, 1516-1517) computed sample sizes (singleton families and affected sib families) needed for the TDT to produce 80% power for detection of a susceptibility locus under a two-allele multiplicative model of epistasis. Under their model, a single parameter () determined the size of the effect for a locus. I demonstrate that when concordant affected MZ twins are used instead of affected singletons, the effect size is increased from to ², and parental heterozygosity is increased for low frequency penetrance-increasing alleles. As a result, the required number of families is less than for singletons or for affected sib pairs. The cost is reduced further with MZ twins because it is only necessary to genotype one twin, just as with singletons, but unlike affected sibs where both must be genotyped. Also, fewer MZ twins than sib pairs would have to be screened phenotypically to achieve any sample-size goal because of the stronger phenotypic correlation of MZ twins.

Address:   Michael B. Miller, Department of Psychology, 210 McAlester Hall, University of Missouri, Columbia, MO 65211, e-mail: mbmiller@taxa.psyc.missouri.edu, web: http://taxa.psyc.missouri.edu/~mbmiller/

¹Department of Psychology, University of Missouri, Columbia, MO

Measurement models for sexual orientation in a large Australian twin sample³

N.G.Martin¹, K.Kirk¹, M.P.Dunne¹, J.M.Bailey²

All previous work on the genetics of sexual orientation has used definitions derived from Kinsey that do not necessarily extract the optimum information from the item data available and therefore potentially cloud the question. In an effort to make this less subjective we are using measurement models in a genetic context (see Hewitt, in Neale & Cardon, 1992, chapter 16) to examine the causes of individual differences in responses to 10 items concerning sexual orientation in 4901 individual twins (1890 complete pairs, 1121 singles) aged 18-50 from the Australian Twin Registry who answered an anonymous questionnaire on sexual behaviour and attitudes. Such multivariate methods have the potential to overcome the low power inherent in the analysis of low frequency categorical items.

Address:   Queensland Institute of Medical Research, Brisbane 4029, Australia, Phone: + 61 7 3362 0278 Fax: +61 7 3362 0101 Email:, nickM@qimr.edu.au

¹Queensland Institute of Medical Research, Brisbane, Australia ²Department of Psychology, Northwestern University, Evanston IL, USA ³Supported by a First Award from NIMH (USA) to JMB and a CARG (Australia) grant to NGM and MPD

Modeling the genetic contribution to smoking and other substance use disorders: quantitative genetic approaches³.

A.C. Heath¹, P.A.F. Madden,¹ A. Todorov,¹ & N.G. Martin²

Results from national twin studies in Sweden, Finland, Australia and the U.S.suggest that genetic factors have an important influence on the onset and course of smoking behavior. We review structural equation model-fitting (SEM) methods that have been used to test for genetic influences on smoking behavior, with illustrations using data from the 1981 survey of the Australian twin panel. The hypothesis that genetic and environmental factors that determine probability of becoming a regular smoker ( initiation') are statistically independent of genetic and environmental influences on probability of becoming a persistent long-term smoker ( persistence') can be rejected; but so also can the hypothesis of a single heritable liability dimension with persistent smokers higher in liability than successful quitters. Relaxation of the assumption of orthogonal liability dimensions can be achieved either via a mediating variable' model that allows for a partial regression of the persistence dimension on the initiation dimension, or via a combined' model under which some successful quitters are assumed to have low liability on the initiation' dimension. We show how tests for genetic effects using model-fitting methods can also be accomplished using a logistic regression model-fitting approach, using dummy variables to model cotwin's smoking status and zygosity. Indeed, when there is no zygosity difference in prevalence, the likelihood-ratio chi-square tests of the hypothesis of no genetic influence are identical in the two approaches. Such a regression approach is readily extended to a survival analysis framework, using Cox regression to model time to successful smoking cessation, and can be applied more easily than SEM to test for mediators of genetic influences on smoking persistence. Using data from the 1981 survey of the Australian twin panel, we show that high heritability of smoking persistence in men (71%) is not explained by associations with personality, education or other sociodemographic variables.

Address:   40 N. Kingshighway, Suite One, St Louis, MO 63108., Phone: 1-314-286-2206; Fax: 1-314-286-2213; Email: andrew@wupsych1.wuSTL.edu

¹Dept. of Psychiatry, Washington University School of Medicine, St Louis, U.S.A. ²Division of Epidemiology and Population Health, Queensland Institute of Medical Research, Australia. ³Supported by NIH grants AA07728, CA75581 and DA00272.

Gene mapping in the presence of gene-environment interactions.

Alexandre A. Todorov¹, E. Genin², K.D. Siegmund ³, P.A.F. Madden¹, A.C. Heath¹

Evidence is accumulating for an important genetic contribution to smoking and other drug involvement, prompting a search for study designs that would be appropriate for the detection of gene(s) underlying vulnerability to drug dependence. Work in the area of linkage analysis has clearly illustrated the difficulty to detect genes of moderate effects using traditional sampling methods. Affected sibpair (ASP) methods lose tremendous power in the presence of gene-environment interactions (A.A.Todorov, K.D. Siegmund, E. Genin, D.C. Rao, A.C. Heath, 1997, Genet. Epid. 14,541). For QTLs, sampling extremely discordant sibpairs does make many genes amenable to analysis, but often proves difficult to implement. In the present, we address the issue of sampling design in the presence of gene-environment correlations. To illustrate the principles outlined in this talk, we first consider the problem of sampling ASPs when the effect of the genes is modulated by the degree of exposure, of which the investigator has an imperfect assessment. As a second illustration, we consider the problem of determining sample sizes for a linkage study of alcohol metabolism, in which the researcher actively manipulates the degree of exposure to the drug (e.g., here, with a challenge dose of alcohol) and has the possibility to oversample some subpopulations defined by important covariates.

Address:   Department of Psychiatry, Washington University School of Medicine, 40N Kingshighway, Suite 1, St.Louis MO 63139, USA, todorov@matlock.wustl.edu, (phone) 314-286-2301, (fax) 314-286-2213

¹Dept. of Psychiatry, Washington University School of Medicine, St.Louis MO 63110 ²Dept. of Integrative Biology, USC School of Medicine, Los Angeles CA 90033 ³Dept. of Preventive Medicine, UC Berkeley, Berkeley CA 94720 Supported in part by grants AA-00728, DA-00272.

New models for survival studies of twins³

Anatoli I. Yashin¹, I.A. Iachine², & A.S. Begun¹

In this paper we discuss applications of survival models to genetic epidemiological studies which further develop the ideas descibed by Yashin and Iachine (A.I.Yashin and I.A. Iachine, 1995, Genet. Epidemiol. 12, 529-538). A new family of correlated frailty models of bivariate survival for the genetic analysis of duration traits is suggested. This family exploits the properties of the three-parametric distribution of frailty which includes gamma, inverse Gaussian and positive stable distributions as particular cases. The semiparametric representations for the bivariate marginal survival functions associated with different frailty distributions are derived. Such representations are used in estimation procedures which do not require parametric assumptions about conditional survival distributions. Several such procedures are suggested and analysed using simulated data. The identifiability and other properties of these models are discussed. The statistical analysis of survival data on Danish twins is performed. The results of analysis for different models are compared. The models are used in the genetic studies of susceptibility to disease and death. The heritability of individual frailty is estimated for different models. The modification of these models in the case of observed covariates in the genetic studies of related individuals is considered. The applications of these models to the analysis of cross-sectional and longitudinal data are discussed.

Address:   Max Planck Institute for Demographic Research, Doberanerstrasse 114, 18057 Rostock, Germany, Phone: (49) 381 208 11 06, Fax: (49) 381 208 12 02, e-mail: yashin@demogr.mpg.de

¹Max Planck Institute for Demographick Research, 18057 Rostock, Germany ²Odense University, Medical School, CHS, Denmark ³supported by NIH/NIA Grant PO1 AG08791

Maximum-likelihood methods of association and linkage²

S. S. Cherny^1,2, D. W. Fulker^1,2, & P. Sham²

Powerful methods currently exist for detecting association in samples of individuals when parental genotypic data are available, allowing simultaneous control for such things as admixture and population stratification, which could otherwise yield false-positive results. However, there are no available methods for detectings allelic associations in unselected, population-based samples of sibships which allow for such statistical control. It is now widely accepted that variance components methods for mapping QTLs using sibships are optimally powerful. We propose an extension of the maximum-likelihood variance-components method of detecting linkage in siblings which incorporates a model on the phenotypic means in addition to the usual model of covariance structure conditional on allelic sharing. The method involves partitioning the mean effect of a locus into a between- and within-family component, thereby controlling for stratification and admixture, while simultaneously modelling linkage, resulting in a further increase in power in cases of weak association due to the trait locus not being in complete disequilibrium with the marker locus. Power is explored for the method under various conditions and the method is compared with a simple analysis-of-variance approach to association data.

Address:   Institute for Behavioral Genetics, Campus Box 447, University of Colorado, Boulder, CO 80309-0447, Phone: +1 303 492 0835, FAX: +1 303 492 8063, Email: Stacey.Cherny@Colorado.EDU, WWW: http://ibgwww.colorado.edu/~cherny/

¹Institute for Behavior Genetics, University of Colorado, Boulder, CO 80309-0447, ² Social, Genetic and Development Psychiatry Research Centre, Institute of Psychiatry, DeCrespigny Park, Camberwell, London SE5 8AF, United Kingdom, ²Supported in part by DA-11015 and a Programme Project grant from the Medical Research Council of Great Britain

The effects of randomized response on the power of the classical twin study

Michael C. Neale¹

Randomized response is commonly used method to increase honest responding to sensitive questions on topics such as substance use or sexual abuse. For example, on every question the respondent tosses three coins in the air and if they are all heads they answer yes, otherwise they answer truthfully. Rates of positive responses have been shown to increase with this method, once corrected for the expected number of artificial positives. In the context of the classical twin study - or any study that focuses on the correlation between responses made in this fashion - some increase in the confidence interval of twin correlations is expected. This effect translates into reduced power to detect heritable and shared environmental variance. The loss of power is shown to be quite drastic for all except abnormally rare response mechanisms requiring a large number of coins. Inferences concerning the confidence intervals on heritability estimates for sensitive topics are drawn.

Address:   Virginia Institute for Psychiatric and Behavioral Genetics, VCU, Box 980126, Richmond, VA 23298. Tel: 804 828 3369, Fax: 804 828 1471, Email: neale@psycho.psi.vcu.edu, Web URL: http://www.vipbg.vcu.edu/~neale/

¹Virginia Institute for Psychiatric and Behavioral Genetics, VCU, Box 980126, Richmond, VA 23298. Supported by grants RR-08123 and MH-01458.

Differential twin-spouse phenotypic assortment, not social homogamy²

Kay Phillips¹

Analyses of twin-spouse data (C. A. Reynolds, L. A. Baker, and N. L. Pedersen, 1996,Behav. Genet. 26,73-88) involved two models, called the factor/delta path model and the delta/delta path model. The results were interpreted as supporting social homogamy as the source of spouse resemblance on a measure of fluid intelligence. The authors favored the factor/delta path model based on fit to the data, but this model is not a delta path model. No delta path-based derivations ca