Further reading
4.4 Gene–environment correlation
Genetic expression has traditionally been considered a process taking place entirely within the constraints of the physiological milieu of an individual and the associated corollary of this conceptualisation is that any association between the individual and their external environment will be unidirectional, that is genes will have no influence on environmental expo- sures. The last few decades has seen a revision of this view, as it has became clear that some genes have an impact on their external environmental setting through behaviours which are, at least partly genet- ically determined, that is bidirectional relationship between genes and environment is probable [21].
Twin studies suggest genetic factors operate through pathways external to the person, influencing expo- sure to a range of environmental risk factors such as stressful life events, poor social support and marital relationship quality [22–24]. That individuals have an active role in the selection, modification and con- struction of their environmental milieu is now widely accepted by behavioural geneticists and evolutionary biologists. Some environmental exposures may indeed be heritable, a phenomenon referred to in evolutionary biology as theextended phenotype[25].
It has been shown in animal behavioural studies that the way animals modify their physical environment,
Box 4.1 The process to determine the joint gene–environmental interplay influencing the risk for psychiatric disorders.
Descriptive Epidemiology
Determine whether there is familiar aggregation
Segregation analyses
Locate risk gene (genes) main effects / interaction using
Linkage & association studies
Study Design
• generation of hypotheses about the relative contribution of putative genetic and environmental risk factors
• family history studies
• twin & adoption studies
• Cohort
• (nested case control)
• Case Control
• Family Study
• General population control
• Sib pair analysis
• case-parent trios
• multi generational pedigrees
• Family member as control methods employed to measure
genetic exposure
• Proxy (surrogate) indicators of genetic risk
• Molecular genotyping
• family history
• intermediate phenotypes
• sibling correlations
• Candidate gene approach
• GEWIS
• What is the genetic model of transmission?
• single gene
• few genes
• poly gene
for example how they build burrows, dams and webs is not randomly distributed but in part, involves her- itable traits. Furthermore in social animals such as rodents, the provision and development of different parental nurturing behaviours (e.g., feeding routines) is partly predicted by variation in their genetics. Fur- thermore, selection experiments find genes to be asso- ciated with aggressive traits which lead the rodent to experience a different social milieu, compared to that of their less aggressive counterparts [24].
GEr can be both causal (Figure 4.1) and non- causal (Figure 4.2) [26]. Non-causal GEr occurs when the health outcome and the environmental factor under study share a genetic liability, but the environmental factor does not have a real causal relationship to the outcome, that is the observed envi- ronmental health outcome association is confounded by genotype. Behavioural contamination may also result in GEr when genetic/individual factors influ- ence how a person feels or perceives a situation, which will be reflected on how they report envi- ronmental exposures. This information bias will be especially problematic when environmental exposure is collected retrospectively.
Causal GEr occurs when there is (at least in part) a genetically mediated variation in causal environ- mental exposure profiles, that is part of a causal pathway. Genotype is unlikely to code directly for specific environmental exposures, but is more likely to actindirectly, through factors such as behaviour and personality; most studies to date have looked at dimensions of neurotism, extraversion and openness to experience, with shared genetic influences evident in studies examining neurotism scores and parenting and neurotism and stressful life events [26, 27].
Gene
PSYCHIATRIC DISORDER LEVEL OF CAUSAL
ENVIRONMENTAL FACTOR
OBSERVED ASSOCIATION
Fig 4.1 Casual: gene–environment correlation.
Gene
True causal relationship
PSYCHIATRIC DISORDER ENVIRONMENTAL
FACTOR
CONFOUNDED ASSOCIATION
Fig 4.2 Non-casual gene–environment correlation.
Three different GEr mechanisms have been described. Passive GEr occurs as most children inherit their genetic characteristics and early rearing environment from the same biological parents, for example parental maltreatment is associated with childhood conduct disorder; however, antisocial behaviour has been demonstrated to be moderately heritable and parents with such traits are more likely to abuse their children, therefore the maltreatment may be indicative of genetic risk transmitted by the parent to the child rather than a true environmental risk factor for conduct disorder [21]. Evocative (reactive) GErrefers to the impact an individual’s behaviour has on their social environment, especially the reaction of others to their behaviour, such that a shy person and a gregarious person evoke different responses in the same social setting. An association between marital difficulties and depression has been demonstrated; however, this relationship may reflect the conflicts that arise from interacting with someone who is depressed, rather than marital difficulties causing depressionper se[21].Selective (active) GEroccurs when an individual’s genetically mediated personality traits and behaviours influence the environmental ‘niches’ they select to live their life in.
Kenneth and Baker [28] have recently carried out a systematic review of human studies examining the heritability (the proportion of individual differences for a trait in a particular population that results from inter-individual genetic differences) of environmental exposures [28]. They identified 55 independent papers reporting on general and specific stressful life events, parenting, family environment, social
support, peer interactions and marital quality and found weighted heritability estimates ranging from 7 to 39%. This review suggests that all aspects of the environment studied in the literature to date, have a significant genetic influence but with small to moder- ate effect sizes. The review suggests that stressful life events which are more dependent on an individual’s own behaviour (e.g., divorce) are more heritable than events that are out with the control of the indi- vidual (e.g., death of a loved one). It found parenting behaviours reflecting positive emotional quality of the parent-child relationship were more heritable than parenting behaviour related to discipline styles.
Importantly, it seems that the genetic makeup from all parties in a dynamic relationship contribute equally to the quality of the environmental experi- ence. For example, in the child–parent relationship the genes from the parent seem to impact as much as the genes in the child, therefore it is possible that risk of psychiatric disorders may be mediated in part through self selection into pathogenic environment.