Twin Design and Role of Environment Flashcards
What is the twin design? (and citation)
The twin design disentangles the relative contributions of genes and environment, across a variety of human traits. The classical twin design is based on contrasting the trait resemblance of monozygotic and dizygotic twin pairs. Monozygotic twins are genetically identical and derived from one fertilised egg (zygote), dizygotic twins develop from separately fertilised eggs and are 50% similar genetically (Polderman et al 2015).
If genetic factors are important for a trait, MZ individuals must be more similar than first-degree relatives or DZ twins.
Using data on twin similarity we decompose the total phenotypic variance of a trait into Additive genetic (A), shared environment (C), and non-shared environment (E), with both MZ and DZ having 100% shared environment and 0 non-shared. Twin data enable the different variance components to be estimated, because MZ and DZ twins have different degrees of correlation for the genetic components (A) but the same degrees of correlation for the environmental components (C and E)
most studies focus on same sex DZ as they provide a more appropriate comparison to MZ twins, who are always of the same sex.
MZ may have more similar experiences but this is not a violation as differences are not caused environmentally CITE
and although some suggest MZ are treated more similarly CITE
this differential treatment does not significantly affect twin similarity for behaviours such as cognitive abilities (CITE),
Eaves et al 2003
Scarr 1968
Morris-Yates, et al, 1990,
what is heritability? how can we estimate it using twins? give example and cite
• Heritability (h2) is an index for the relative contribution of genetic effects to the total phenotypic variance. It tells us about that trait, measured in that specific population, at that specific time, but doesn’t tell us about individual risk The heritability (h2 ) of the phenotype can be estimated from twice the difference between MZ and DZ correlations • For example, typical MZ and DZ correlations for depression are about 0.4 and 0.2 (Kendler et al., 1992) and therefore heritability is estimated at ~40%
Heritability and intelligence
There is substantial heritability for intelligence, with genetics typically accounting for 50% of the variability (Deary, 2013), although estimates have ranged from 30% to 80% (Deary et al., 2009). Longitudinal twin research has shown that heritability rises across the lifespan (Davis, Haworth & Plomin, 2009), increasing from 41% in childhood, to 66% in young adulthood (Haworth et al., 2010). In contrast, influences from the shared environment become less predominant with age, accounting for 74% of the variance in early childhood, and 33% in middle childhood (Davis et al., 2009). These changes in aetiological influences could explain the range of heritability estimates reported by Deary et al., (2009).
casuality from twin design - 2 studies
In addition to confirming genetic and environment influences, twin studies can also be used to confirm relationships and causality by establishing whether they are due to genetic or environmental confounding. For example:
Stadler et al., (2012). - explored the association between subjective wellbeing and increased longevity using twin pair analyses. As expected, at the individual level (without regard to twin pair membership), SWB predicted increased longevity. Exposure effects were also present in unadjusted and adjusted within-pair analyses of dizygotic (DZ) pairs and monozygotic (MZ) pairs, indicating that SWB is associated with increased longevity independent of familial factors of genes and shared environment, consistent with a causal link between SWB and longer lifespan.
De Moor, et al., 2008 - tested causality in the association between regular exercise and symptoms of anxiety and depression. In genetically identical twin pairs, the twin who exercised more did not display fewer anxious and depressive symptoms than the co-twin who exercised less. = Regular exercise is associated with reduced anxious and depressive symptoms in the population at large, but the association is not because of causal effects of exercise.
another assumption of twin design is that results can be generalised to the rest of the population. Specifically, the twin method assumes that twins are similar to singletons. There are many ways in which twins have been found to differ from singletons (CITE) eg
birth weight, lower IQ than singletons, however.. ?
(Evans and Martin, 2000);
those studies that have found differences between twins and singletons have been conducted on young twins, and studies on older twins confirm that these differences have all but disappeared by early to middle childhood
SEM - evaluative point from..
• STRUCTURAL EQUATION MODELLING (SEM), is an approach in which genotypic and environmental effects are modelled as the contribution of unmeasured (latent) variables to the potentially multivariate phenotypic differences between individuals. The latent factors represent the effects of many unidentified influences.
Bulik et al., 2000 – in their review argue that SEM is quite flexible and rather complex models can be easily specified. It provides a scientifically rigorous framework that allows previously articulated hypotheses to be evaluated by a clearly specified set of statistical rules and competing hypotheses can be directly compared. SEM can be depicted graphically in the form of a path diagram which is convenient and often clarifying.
Haworth, Kovas, Dale & Plomin 2008- cholesky
investigated the etiology of academic performance in Science and its etiological links with other academic abilities and general cognitive ability (g). = Our multivariate results show that Science shares genetic influences with English, Math and g. Nonetheless, Science is more than just g, as there are specific genetic and environmental influences on Science. In addition, there are common environmental influences shared by the three school subjects that are not shared with g — a school environment effect.
These findings were revealed using a Cholesky decomposition model which allowed for measurement of the unique and common genetic and environmental influences.
=In this case, the first factor assesses genetic, shared and non-shared environmental influences on g, some of which may also influence English, Math and Science. The second factor estimates influences on English, some of which may also influence Math and Science, and the final factor estimates influences that are unique to Science.
=genetics unqiue to academic perfomance, and science
=shared e specific to g
Haworth et al., 2013
Multivariate twin analyses were used to investigate the genetic and environmental links between environment and outcome. This study used data from the Twins’ Early Development Study (TEDS; Oliver & Plomin, 2007).
To assess the science-learning environment we used items from the Classroom, Home and Peer Environment Influences Scale. A test was used to examine scientific enquiry skills, these are skills needed to design and evaluate scientific evidence, and are a key component in the UK National Curriculum.
To investigate the links between the learning environment and science performance, we conducted bivariate twin model-fitting. Bivariate model-fitting decomposes the covariance between traits, providing estimates of the genetic and environmental correlations between traits.
= The most surprising result was that the science-learning environment was almost as heritable (43%) as performance on the science test (50%), and showed negligible shared environmental influence (3%). This science learning environment included separate subscales of the classroom environment and the peer environment which were strikingly similar.
Bivariate twin analyses indicated that 56% of the phenotypic correlation between the science-learning environment and science performance was explained by genetic influences, indicating gene–environment correlation.
BUT
However, note that because the phenotypic correlation is only 0.225, this means that overlapping genetic factors explain just a small proportion (2.8%) of the total variance in science performance. There are at least two explanations for the low correlation - ?
TEDS evaluative point
The TEDS sample has been shown to be reasonably representative of the general population in terms of parental education, ethnicity and employment status (Kovas et al., 2007
Haworth 2008
= genetic influences account for over 60% of the variance in scientific achievement, with environmental influences accounting for the remaining variance. Environmental influences were mainly of the non-shared variety, suggesting that children from the same family experience school environments differently.
Tucker-Drob and Harden (2012)
provided evidence for gene-environment correlations using longitudinal data from identical and fraternal twin pairs.
• The first set of analyses tested the directionality of the association between parenting and early cognitive ability, using two regression models. The first controlled for cognitive stimulation at 2 years and found that children’s cognitive ability at 2 predicted the quality of stimulation received from their parents at 4 years.
• The second regression model tested the reciprocal association (scores at 2 years predicting parenting behavior at 4 years, controlling for baseline parenting behavior).
• Next, data from both twins in each pair was used to estimate a series of behavioral genetic models. Following the conventions of the classical twin model, variance in each phenotype was decomposed into three components: ACE.
• The associations between the phenotypes were modeled using a Cholesky decomposition, in which each subsequent phenotype is regressed onto the A, C, and E components of all preceding phenotypes
= phenotypic associations between parental cognitive stimulation and child cognitive ability were reciprocal. Notably, the standardized regression coefficients from phenotypic models were approximately equal, indicating that children’s abilities predict their parents’ future behavior as strongly as parents’ behaviors predict their children’s future abilities.
=Parenting influenced cognitive development through an environmental pathway, whereas children’s cognitive ability influenced subsequent parenting through a genetic pathway. These results suggest that genetic influences on cognitive development occur through a transactional process, in which genetic predispositions influence early cognitive development leading children to evoke stimulation of differing levels of quality from their parents, and these early levels of stimulation by parents act as effectual environments in boosting their children’s subsequent cognitive development.
HOWEVER, the data analyzed were limited in some respects. First, parenting and cognition data were only available for a relatively narrow period during early childhood: ages 2 years to 4 years. More longitudinal measurements over an extended age range would be useful to examine how gene-environment transactions unfold over the entire span of child development.
ALSO, while the current twin design was informative about the operation of children’s genes, it was insufficient for making inferences about the operation of parents’ genes. That is, while we found that parenting affected cognitive development through a family-level environmental pathway, our design was not capable of determining the extent to which parenting behaviors were themselves influenced by parents’ genes
MZ twins may have more similar experiences than DZ twins because they are more similar genetically. Such differences between MZ and DZ in experience are not a violation to the equal environment assumption because … (CITE)
the differences are not caused environmentally (Eaves et al., 2003).
twins tend to have lower birth weights and are at greater risk of perinatal complications than singletons (CITE)
In childhood, language develops slower in twins, and twins perform less well on tests of verbal ability and IQ (CITE). These delays are similar for MZ and DZ and tend to be a result of postnatal environment rather than pre-maturity (CITE).
However, most of the deficits are recovered in early school years (CITE) and twins do not appear importantly different from singletons for cog abilities (CITE) personality (CITE) or psychopathology (CITE).
In addition, most studies have not matched twins with singletons in terms of genetic background nor early environmental experiences. A recent study comparing twins with their siblings (who are matched in both genetic background and early environmental experiences) in Dutch WAIS (Wechsler Adult Intelligence Scale) scores, failed to detect any differences between twins and their siblings in cognitive abilities (CITE)
O’Brien et al., 1987 Ronald et al., 2005 Rutter et al, 1991 Christensen et al., 2006 Johnson et al 2002 Robbers et al., 2011 Reiss, et al, 2000. Posthuma et al., 2000
are twins generalisable?
Ronald et al., 2005 Rutter et al, 1991 Christensen et al., 2006 Johnson et al 2002 Robbers et al., 2011 Reiss, et al, 2000. Posthuma et al., 2000
In childhood, language develops slower in twins, and twins perform less well on tests of verbal ability and IQ (Ronald et a., 2005). These delays are similar for MZ and DZ and tend to be a result of postnatal environment rather than pre-maturity (Rutter et al., 1991).
However, most of the deficits are recovered in early school years (Christensen et al., 2006) and twins do not appear importantly different from singletons for personality (Johnson et al 2002) or psychopathology (Robbers et al., 2011).
A host of studies comparing older twins with singletons have failed to find differences in physical characteristics and cognitive abilities and psychological outcomes (Reiss, et al, 2000.)
In addition, most studies have not matched twins with singletons in terms of genetic background nor early environmental experiences. A recent study comparing twins with their siblings (who are matched in both genetic background and early environmental experiences) in Dutch WAIS (Wechsler Adult Intelligence Scale) scores, failed to detect any differences between twins and their siblings in cognitive abilities (Posthuma et al., 2000)
