Shared genetic influences do not explain the association between parent-offspring relationship quality and offspring internalizing problems: results from a Children-of-Twins study

BACKGROUND

Associations between parenting and child outcomes are often interpreted as reflecting causal, social influences. However, such associations may be confounded by genes common to children and their biological parents. To the extent that these shared genes influence behaviours in both generations, a passive genetic mechanism may explain links between them. Here we aim to quantify the relative importance of passive genetic v. social mechanisms in the intergenerational association between parent-offspring relationship quality and offspring internalizing problems in adolescence.

METHODS

We used a Children-of-Twins (CoT) design with data from the parent-based Twin and Offspring Study of Sweden (TOSS) sample [909 adult twin pairs and their offspring; offspring mean age 15.75 (2.42) years], and the child-based Swedish Twin Study of CHild and Adolescent Development (TCHAD) sample [1120 adolescent twin pairs; mean age 13.67 (0.47) years]. A composite of parent-report measures (closeness, conflict, disagreements, expressions of affection) indexed parent-offspring relationship quality in TOSS, and offspring self-reported internalizing symptoms were assessed using the Child Behavior Checklist (CBCL) in both samples.

RESULTS

A social transmission mechanism explained the intergenerational association [r = 0.21 (0.16-0.25)] in our best-fitting model. A passive genetic transmission pathway was not found to be significant, indicating that parental genetic influences on parent-offspring relationship quality and offspring genetic influences on their internalizing problems were non-overlapping.

CONCLUSION

These results indicate that this intergenerational association is a product of social interactions between children and parents, within which bidirectional effects are highly plausible. Results from genetically informative studies of parenting-related effects should be used to help refine early parenting interventions aimed at reducing risk for psychopathology.

The relationship between parental depressive symptoms and offspring psychopathology: evidence from a children-of-twins study and an adoption study

BACKGROUND

Parental depressive symptoms are associated with emotional and behavioural problems in offspring. However, genetically informative studies are needed to distinguish potential causal effects from genetic confounds, and longitudinal studies are required to distinguish parent-to-child effects from child-to-parent effects.

METHOD

We conducted cross-sectional analyses on a sample of Swedish twins and their adolescent offspring (n = 876 twin families), and longitudinal analyses on a US sample of children adopted at birth, their adoptive parents, and their birth mothers (n = 361 adoptive families). Depressive symptoms were measured in parents, and externalizing and internalizing problems measured in offspring. Structural equation models were fitted to the data.

RESULTS

Results of model fitting suggest that associations between parental depressive symptoms and offspring internalizing and externalizing problems remain after accounting for genes shared between parent and child. Genetic transmission was not evident in the twin study but was evident in the adoption study. In the longitudinal adoption study child-to-parent effects were evident.

CONCLUSIONS

We interpret the results as demonstrating that associations between parental depressive symptoms and offspring emotional and behavioural problems are not solely attributable to shared genes, and that bidirectional effects may be present in intergenerational associations.

Genetic overlap between episodic memory deficits and schizophrenia: results from the Maudsley Twin Study

BACKGROUND

Visual and verbal episodic memory deficits are putative endophenotypes for schizophrenia; however, the extent of any genetic overlap of these with schizophrenia is unclear. In this study, we set out to quantify the genetic and environmental contributions to variance in visual and verbal memory performance, and to quantify their genetic relationship with schizophrenia.

METHOD

We applied bivariate genetic modelling to 280 twins in a classic twin study design, including monozygotic (MZ) and dizygotic (DZ) pairs concordant and discordant for schizophrenia, and healthy control twins. We assessed episodic memory using subtests of the Wechsler Memory Scale - Revised (WMS-R).

RESULTS

Genetic influences (i.e. heritability) contributed significantly to variance in immediate recall of both verbal memory and visual learning, and the delayed recall of verbal and visual memory. Liability to schizophrenia was associated with memory impairment, with evidence of significant phenotypic correlations between all episodic memory measures and schizophrenia. Genetic factors were the main source of the phenotypic correlations for immediate recall of visual learning material; both immediate and delayed recall of verbal memory; and delayed recall of visual memory that, for example, shared genetic variance with schizophrenia, which accounted for 88% of the phenotypic correlation (rph=0.41) between the two.

CONCLUSIONS

Verbal memory and visual learning and memory are moderately heritable, share a genetic overlap with schizophrenia and are valid endophenotypes for the condition. The inclusion of these endophenotypes in genetic association studies may improve the power to detect susceptibility genes for schizophrenia.

Neuroticism, recall bias and attention bias for valenced probes: a twin study

BACKGROUND

Prior research on the nature of the vulnerability of neuroticism to psychopathology suggests biases in information processing towards emotional rather than neutral information. It is unclear to what extent this relationship can be explained by genetic or environmental factors.

METHOD

The genetic relationship between a neuroticism composite score and free recall of pleasant and unpleasant words and the reaction time on negative probes (dot-probe task) was investigated in 125 female twin pairs. Interaction effects were modelled to test whether the correlation between neuroticism and cognitive measures depended on the level of the neuroticism score.

RESULTS

The only significant correlation was between neuroticism and the proportion of recalled unpleasant words (heritability is 30%), and was only detectable at the higher end of the neuroticism distribution. This interaction effect seems to be due to environmental effects that make people in the same family more similar (e.g. parental discipline style), rather than genetic factors. An interesting sub-finding was that faster reaction times for left versus right visual field probes in the dot-probe task suggest that cognitive processing in the right hemisphere is more sensitive to subliminal (biologically relevant) cues and that this characteristic is under substantial genetic control (49%). Individual differences in reaction times on right visual field probes were due to environmental effects only.

CONCLUSIONS

There is no evidence that the predisposition of individuals to focus on negative (emotional) stimuli is a possible underlying genetic mechanism of neuroticism.

Computation of individual latent variable scores from data with multiple missingness patterns

Latent variable models are used in biological and social sciences to investigate characteristics that are not directly measurable. The generation of individual scores of latent variables can simplify subsequent analyses. However, missing measurements in real data complicate the calculation of scores. Missing observations also result in different latent variable scores having different degrees of accuracy which should be taken into account in subsequent analyses. This manuscript presents a publicly available software tool that addresses both these problems, using as an example a dataset consisting of multiple ratings for ADHD symptomatology in children. The program computes latent variable scores with accompanying accuracy indices, under a 'user-specified' structural equation model, in data with missing data patterns. Since structural equation models encompass factor models, it can also be used for calculating factor scores. The program, documentation and a tutorial, containing worked examples and specimen input and output files, is available at http://statgen.iop.kcl.ac.uk/lsc .

Brain MRI abnormalities in schizophrenia: same genes or same environment?

BACKGROUND

Structural brain volume abnormalities are among the most extensively studied endophenotypes in schizophrenia. Bivariate genetic model fitting (adjusted to account for selection) was used to quantify the genetic relationship between schizophrenia and brain volumes and to estimate the heritability of these volumes.

METHOD

We demonstrated by simulation that the adjusted genetic model produced unbiased estimates for endophenotype heritability and the genetic and environmental correlations. The model was applied to brain volumes (whole brain, hippocampus, third and lateral ventricles) in a sample of 14 monozygotic (MZ) twin pairs concordant for schizophrenia, 10 MZ discordant pairs, 17 MZ control pairs, 22 discordant sibling pairs, three concordant sibling pairs, and 114 healthy control subjects.

RESULTS

Whole brain showed a substantial heritability (88%) and lateral ventricles substantial common environmental effects (67%). Whole brain showed a significant genetic correlation with schizophrenia, whereas lateral ventricles showed a significant individual specific correlation with schizophrenia. There were significant familial effects for hippocampus and third ventricle, but the analyses could not resolve whether these were genetic or environmental in origin (around 30%each).

CONCLUSIONS

Using genetic model fitting on twin and sibling data we have demonstrated differential sources of covariation between schizophrenia and brain volumes, genetic in the case of whole brain volume and individual specific environment in the case of lateral ventricles.

Haplotype association analysis of discrete and continuous traits using mixture of regression models

We present a regression-based method of haplotype association analysis for quantitative and dichotomous traits in samples consisting of unrelated individuals. The method takes account of uncertain phase by initially estimating haplotype frequencies and obtaining the posterior probabilities of all possible haplotype combinations in each individual, then using these as weights in a finite mixture of regression models. Using this method, different combinations of marker loci can be modeled, to find a parsimonious set of marker loci that are most predictive and therefore most likely to be closely associated with the a quantitative trait locus. The method has the additional advantage of being able to use individuals with some missing genotype data, by considering all possible genotypes at the missing markers. We have implemented this method using the SNPHAP and Mx programs and illustrated its use on published data on idiopathic generalized epilepsy.

Genetic and environmental influences on psychological distress in the population: General Health Questionnaire analyses in UK twins

BACKGROUND

The General Health Questionnaire (GHQ) is the most popular screening instrument for detecting psychiatric disorders in community samples. Using longitudinal data of a large sample of UK twin pairs, we explored (i) heritabilities of the four scales and the total score; (ii) the genetic stability over time; and (iii) the existence of differential heritable influences at the high (ill) and low (healthy) tail of the distribution.

METHOD

At baseline we assessed the GHQ in 627 MZ and 1323 DZ female pairs and at a second occasion (3.5 years later) for a small subsample (90 MZ and 270 DZ pairs). Liability threshold models and raw ordinal maximum likelihood were used to estimate twin correlations and to fit longitudinal genetic models. We estimated extreme group heritabilities of the GHQ distribution by using a model-fitting implementation of the DeFries-Fulker regression method for selected twin data.

RESULTS

Heritabilities for Somatic Symptoms, Anxiety, Social Dysfunction, Depression and total score were 0.37, 0.40, 0.20, 0.42 and 0.44, respectively. The contribution of shared genetic factors to the correlations between time points is substantial for the total score (73%). Group heritabilities of 0.48 and 0.43 were estimated for the top and bottom 10% of the total GHQ score distribution, respectively.

CONCLUSION

The overall heritability of the GHQ as a measure of psychosocial distress was substantial (44%), with all scales having significant additive genetic influences that persisted across time periods. Extreme group analyses suggest that the genetic control of resilience is as important as the genetic control of vulnerability.

Analytic power calculation for QTL linkage analysis of small pedigrees

Power calculation for QTL linkage analysis can be performed via simple algebraic formulas for small pedigrees, but requires intensive computation for large pedigrees, in order to evaluate the expectation of the test statistic over all possible inheritance vectors at the test position. In this report, we show that the non-centrality parameter for an arbitrary pedigree can be approximated by the sum of the variances of the correlations between all pairs of relatives, each variance being weighted by a factor that is determined by the mean correlation of the pair. We show that this approximation is sufficiently accurate for practical purposes in small to moderately large pedigrees, and that large sibships are more efficient than other family structures under a range of genetic models.

Life events and depression in a community sample of siblings

BACKGROUND

The overall aim of the GENESiS project is to identify quantitative trait loci (QTLs) for anxiety/depression, and to examine the interaction between these loci and psychosocial adversity. Here we present life-events data with the aim of clarifying: (i) the aetiology of life events as inferred from sibling correlations; (ii) the relationship between life events and measures of anxiety and depression, as well as neuroticism; and (iii) the interaction between life events and neuroticism on anxiety/depression indices.

METHODS

We assessed the occurrence of one network and three personal life-event categories and multiple indices of anxiety/depression including General Health Questionnaire, Anhedonic Depression, Anxious Arousal and Neuroticism in a large community-based sample of2150 sib pairs, 410 trios and 81 quads. Liability threshold models and raw ordinal maximum likelihood were used to estimate within-individual and between-sibling correlations of life events. The relationship between life events and indices of emotional states and personality were assessed by multiple linear regression and canonical correlations.

RESULTS

Life events showed sibling correlations of 0-37 for network events and between 0-10 and 0.19 for personal events. Adverse life events were related to anxiety and depression and, to a less extent, neuroticism. Trait-vulnerability (as indexed by co-sib's neuroticism, anxiety and depression) accounted for 11% and life events for 3% of the variance in emotional states. There were no interaction effects.

CONCLUSIONS

Life events show moderate familiality and are significantly related to symptoms of anxiety and depression in the community. Appropriate modelling of life events in linkage and association analyses should help to identify QTLs for depression and anxiety.

The genetic basis of the relation between speed-of-information-processing and IQ

The relationship of speed-of-information-processing (SIP), as derived from reaction times (RTs) on experimental tasks, and intelligence has been extensively studied. SIP is suggested to measure the efficiency with which subjects can perform basic cognitive operations underlying a wide range of intellectual abilities. Observed phenotypic correlations between RT and IQ typically are in the -0.2 to -0.4 range, and the question is addressed to what extent this relationship is determined by genetic or environmental influences. In a group of Dutch twins the heritabilities for RT tasks at age 16 and 18 years were estimated longitudinally and the nature of the RT-IQ relationship was investigated. At age 16 years heritabilities for a simple reaction time (SRT) and choice reaction time (CRT) were 64 and 62% and the average phenotypic correlations between the RTs and IQ, assessed by the Raven standard progressive matrices, was -0.21. At the second test occasion lower heritabilities were observed for the RTs, probably due to modifications in administration procedures. The mean correlations between the RTs and WAIS verbal and per formal subtests were -0.18 and -0.16. Multivariate genetic analyses at both ages showed that the RT-IQ correlations were explained by genetic influences. These results are in agreement with earlier findings (Baker et al., Behav Genet 1991;21:351-67; Ho et al., Behav Genet 1988;18:247-61) and support the existence of a common, heritable biological basis underlying the SIP-IQ relationship.

Genetic mediation of the correlation between peripheral nerve conduction velocity and IQ

Variation in peripheral nerve conduction velocity (PNCV) and intelligence was studied in 18-year-old Dutch twins. It has been suggested that both brain nerve conduction velocity and PNCV are positively correlated with intelligence (Reed, 1984) and that heritable differences in nerve conduction velocity may explain part of the well-established heritability of intelligence. The relationship among IQ, obtained with the Wechsler Adult Intelligence Scale, and median nerve PNCV was examined in 159 twin pairs. Genetic analyses showed a heritability of 81% for IQ and 66% for onset PNCV. The small but significant phenotypic correlation between IQ and onset PNCV (.15) was entirely mediated by common genetic factors. Analyses of differences scores for PNCV of this study and PNCV from the same subjects collected at age 16 suggest that there might still be development in PNCV in this age interval. This maturation is highly controlled by genetic factors. It is suggested that variation in IQ that is associated with nerve conduction velocity becomes apparent only after the developmental processes in peripheral nerves are completed. This is in line with the suggestion of increasing heritability of IQ in adulthood.

Genetic analysis of peripheral nerve conduction velocity in twins

We studied variation in peripheral nerve conduction velocity (PNCV) and intelligence in a group of 16-year-old Dutch twins. It has been suggested that both brain nerve conduction velocity and PNCV are positively correlated with intelligence (Reed, 1984) and that heritable differences in NCV may explain part of the well established heritability of intelligence. The Standard Progressive Matrices test was administered to 210 twin pairs to obtain IQ scores. Median nerve PNCV was determined in a subgroup of 156 pairs. Genetic analyses showed a heritability of 0.65 for Raven IQ score and 0.77 for PNCV. However, there was no significant phenotypic correlation between IQ score and PNCV.

Effect of depression on psychomotor skills, eye movements and recognition-memory

In this study 12 depressed outpatients were compared to 12 healthy controls with respect to their performance on a number of cognitive tasks, including a recognition-memory task, and their eye movements and pupil size were recorded while watching a traffic film. The recognition-memory task consisted of words with intermediate hedonic tone (neutral words), words with high hedonic tone ('good' words) and words with low hedonic tone ('bad' words). Patients performed slower on perceptual-motor tasks which could be characterized as effort-demanding, while no difference between groups was found on effortless tasks. In addition, the range of horizontal eye movements, an indication of visual span, was found to be less in patients. Signal-detection analysis on the recognition-memory data showed an impairment of 'pure' memory in depressives. Analyses on response bias indicated that patients had more false alarms, but only with respect to good words. It is concluded that patients exhibit cognitive deficits, including memory impairment, a narrower visual span and a risky response strategy on good words, which may be an indication of the trouble patients have in processing emotionally toned words.