J. Hudziak

Netherlands Twin Register: From Twins to Twin Families

In the late 1980s The Netherlands Twin Register (NTR) was established by recruiting young twins and multiples at birth and by approaching adolescent and young adult twins through city councils. The Adult NTR (ANTR) includes twins, their parents, siblings, spouses and their adult offspring. The number of participants in the ANTR who take part in survey and / or laboratory studies is over 22,000 subjects. A special group of participants consists of sisters who are mothers of twins. In the Young NTR (YNTR), data on more than 50,000 young twins have been collected. Currently we are extending the YNTR by including siblings of twins. Participants in YNTR and ANTR have been phenotyped every 2 to 3 years in longitudinal survey studies, since 1986 and 1991 for the YNTR and ANTR, respectively. The resulting large population-based datasets are used for genetic epidemiological studies and also, for example, to advance phenotyping through the development of new syndrome scales based on existing items from other inventories. New research developments further include brain imaging studies in selected and unselected groups, clinical assessment of psychopathology through interviews, and cross-referencing the NTR database to other national databases. A large biobank enterprise is ongoing in the ANTR in which blood and urine samples are collected for genotyping, expression analysis, and metabolomics studies. In this paper we give an update on the YNTR and ANTR phenotyping and on the ongoing ANTR biobank studies.

Individual Differences in Aggression: Genetic Analyses by age, gender, and informant in 3-, 7-, and 10-year-old Dutch Twins

Aggression in humans is associated with substantial morbidity and mortality. In this study we report on the aggressive behavior syndrome (AGG) in young children as defined by the Child Behavior Checklist (CBCL) and the Teacher Report Form (TRF). We assessed aggression in a large sample of Dutch twins at ages 3, 7, and 10 years. The purpose of this study was three-fold. First, we determined the number of children who are “clinically deviant” on the AGG scale. Second, we assessed the genetic and environmental contributions to AGG for the maternal, paternal, and teacher ratings at each age, for boys and girls. Third, we explored issues of rater bias by analyzing parental and teacher data simultaneously. CBCL data were available from mothers on 6436 three-year-old, 5451 seven-year-old, and 2972 ten-year-old twin pairs and CBCL data from fathers on 4207 three-year-old, 4269 seven-year-old, and 2295 ten-year-old twin pairs. Teacher report data from the TRF were collected for 1036 seven-year-old and 903 ten-year-old twin pairs from the Netherlands Twin Registry. Structural equation modeling was employed to obtain genetic and environmental estimates at each age. Analyses were conducted separately by age and informant, as well as simultaneously, for all informants. Differences in raw scores across gender were found, with boys being rated as more aggressive than girls by all informants. Mothers reported more symptoms than fathers, who reported more symptoms than teachers. Evidence for moderate to high genetic influence (51%–72%) was seen for AGG by all three informants at all ages with only small sex differences in heritability estimates. Best fitting models for AGG by parent reports also included a small contribution of common environment. The largest sex differences in heritabilities were seen at age 10. Contributions of common (13%–27%) and unique (16%–31%) environment were small to moderate. There was some evidence of genetic dominance by teacher report for 10-year-old girls.

Co-occurrence of Aggressive Behavior and Rule-Breaking Behavior at Age 12: Multi-Rater Analyses

Aggressive Behavior (AGG) and Rule-Breaking Behavior (RB) are two of the eight CBCL syndromes. The phenotypic correlation between AGG and RB ranges from .48 to .76, and varies depending on the rater and the sex of the child. Prevalence of AGG and RB (i.e., T ≥ 67) is in the range of 6%–7% in both boys and girls. Fifty percent to 60% of the children who are deviant on AGG are also deviant on RB and vice versa. Why so many children show problem behavior in the clinical range for both syndromes is unclear. This co-occurrence could be due to genetic factors influencing both traits, to environmental factors influencing both traits, or to both. The purpose of this study is to use a genetically informative sample to estimate genetic and environmental influences on AGG and RB and to investigate the etiology of the co-occurrence of both behaviors. We do this using multiple informants to take into account underlying sources of parental agreement and disagreement in ratings of their offspring. To this end, mother and father ratings of AGG and RB were collected by using the Child Behavior Checklist in a large sample of 12-year-old twins. Parental agreement is represented by an interparent correlation in the range of .53–.76, depending on phenotype (AGG or RB) and sex of the child. Genetic influences account for 79% and 69% of the individual differences in RB and AGG behavior (defined as AGG and RB on which both parents do agree) in boys. In girls 56% and 72% of the variance in RB and AGG are accounted for by genetic factors. Shared environmental influences are significant for RB in girls only, explaining 23% of the total variance. Eighty percent of the covariance between AGG and RB, similarly assessed by both parents, can be explained by genetic influences. So, co-occurrence in AGG and RB is mainly caused by a common set of genes. Parental disagreement seems to be a combination of so-called rater bias and of parental specific views.

Influences on achieving motor milestones: A twin-singleton study

In order to determine if twinning impacted achievement of motor milestones the attainment of early motor milestones in twins was examined and compared to published data from singletons of the same age from the same culture and birth years. We examined the influence of twinning, sex, zygosity and birth cohort (1987-2001) on the motor development of twins aged 0 to 24 months. Data on the attainment of motor milestones (turn, sit, crawl, stand and walk) of twins were collected from maternal reports. All data were corrected for gestational age. Data from the twin sample were compared to normative data from singletons, which were available from Child Health Clinics (CHC). Analyses across twin data and the CHC singleton data revealed no differences between twins and singletons in achievement of motor milestones. Girls were able to sit without support slightly earlier than boys, otherwise there were no other sex differences. Birth-order analyses revealed minimal but significant differences in turning over from back to belly and for sitting without support between the first- and second-born. Dizygotic (DZ) twins were faster than monozygotic (MZ) twins in achieving the moment of sit, crawl, stand and walk. Twins born in earlier cohorts were faster in reaching the moment of crawl, sit and walk. It is concluded that there are no differences in time of reaching motor milestones between twins and singletons within the normal range. Sex has minimal to no effect on motor development in early childhood. DZ twins achieve motor milestones sooner than MZ twins. Attainment of gross motor milestones (crawl, stand and walk) is delayed in later birth cohorts.

Can Genetics Help Psychometrics? Improving Dimensionality Assessment Through Genetic Factor Modeling

In the present article, we discuss the role that quantitative genetic methodology may play in assessing and understanding the dimensionality of psychological (psychometric) instruments. Specifically, we study the relationship between the observed covariance structures, on the one hand, and the underlying genetic and environmental influences giving rise to such structures, on the other. We note that this relationship may be such that it hampers obtaining a clear estimate of dimensionality using standard tools for dimensionality assessment alone. One situation in which dimensionality assessment may be impeded is that in which genetic and environmental influences, of which the observed covariance structure is a function, differ from each other in structure and dimensionality. We demonstrate that in such situations settling dimensionality issues may be problematic, and propose using quantitative genetic modeling to uncover the (possibly different) dimensionalities of the underlying genetic and environmental structures. We illustrate using simulations and an empirical example on childhood internalizing problems.

Short- and Long-Term Effects of Child Care on Problem Behaviors in a Dutch Sample of Twins

This study examined the association between early child care on the development of behavior problems. At the age of 5 years, child care information was collected from parents on a large group of twins who were born between 1985 and 1997. Mothers and fathers rated the behavior of the child at ages 3, 7, and 10 years using the Child Behavior Checklist (CBCL) and teachers reported on the same childrens behaviors using the Teachers Report Form (TRF) at ages 7 and 10. At the age of 3 years, children with nonparental child care experiences had more externalizing problems than children with only parental child care. The long-term effects of quantity of child care were mixed and were only significant for mother ratings and for children from families with a low socioeconomic status. Overall, the effect sizes of child care were very small (effect sizes were between .12 and .23). Children with a larger amount of child care did not show more behavior problems, therefore it was questioned whether the increased levels of behavior problems could be attributed to quantity of child care.

Moderation of genetic factors by parental divorce in adolescents' evaluations of family functioning and subjective wellbeing

Adolescents evaluations of family functioning may have a significant impact on their subjective well-being and adjustment. The aim of the study was to investigate the degree to which genetic and environmental influences affect variation in evaluations of general family functioning, family conflict, and quality of life and the overlap between them. We assessed whether genetic and environmental influences are moderated by parental divorce by analyzing self-report data from 6,773 adolescent twins and their non-twin siblings. Genetic, shared, and nonshared environmental influences accounted for variation in general family functioning and family conflict, with genetic influences being relatively more important in girls than boys in general family functioning. Genetic and nonshared environmental influences accounted for variation in quality of life, with genetic influences being relatively more important in girls. Evidence was found for interaction between genetic factors and parental divorce: genetic influence on general family functioning was larger in participants from divorced families. The overlap between general family functioning and quality of life, and family conflict and quality of life was accounted for the largest part by genetic effects, with nonshared environmental effects accounting for the remaining part. By examining the data from monozygotic twins, we found evidence for interaction between genotype and nonshared, non-measured, environmental influences on evaluations of general family functioning, family conflict, and quality of life.

A genome-wide association study for subjective wellbeing

with significantly increased rates of anhedonia (58%) in individuals who also experienced childhood physical abuse. While our results are highly consistent with the rodent literature, we cannot determine (as the function of rs1049353 is not well understood) whether this interaction reflects the action of the minor allele in buffering the effects of abuse on anhedonia or whether the presence of the major allele confers risk for anhedonia in the presence of abuse.