Page 119 - Read Online

Page 105 De Francesco et al. J Transl Genet Genom 2024;8:102-18 https://dx.doi.org/10.20517/jtgg.2023.51

traits by analyzing two main samples: dyads of parents and children who share the same genetic heritage
but not the environment and, on the opposite, dyads that share the same environment but are not
[11]
genetically related . Despite having proved to be valuable in unraveling the genetic and environmental
contribution to several behavioral traits, adoption studies have some significant limitations. Firstly, the
samples collected may not always be representative of the general population: on the one hand, people who
decide to give children up for adoption are often exposed to severe socioeconomic and environmental
conditions, whereas on the other, in order to ensure the child wellbeing, in most western countries, adoptive
families are selected through intense screening processes aimed at identifying future parents with great
socioeconomic statuses and high levels of education, but these variables obviously cannot be controlled
[13]
when assessing biological families .

Moreover, because of cultural changes such as the increase in the use of hormonal contraceptives, the
decriminalization of abortion, and the increasing rates of occupied women, in the last decades, adoption
[10]
rates have decreased significantly, making it increasingly difficult to retrieve this specific sample .

Twin studies also allow the detection of environmental contributions to phenotypes, but they certainly
represent the best indicator to estimate the extent to which biology influences a trait, through the
comparison between Monozygotic (MZ) and Dizygotic (DZ) twins [10,16] . MZ twins can also be referred to as
identical twins as, by being conceived by the fertilization of a single zygote, they share 100% of their genetic
heritage and are phenotypically identical. DZ twins are instead conceived by the fertilization of two different
eggs and share only 50% of the genetic heritage, just as siblings born from different pregnancies. However,
in twin studies, DZs constitute a better control group than siblings, as they are peers who have also shared
the same intrauterine environment .
[17]

Twin studies
Twin studies evaluate correlations for a given trait between MZ and DZ twins by assuming that if the
phenotype is mostly influenced by genetic variables, correlations between MZs will be higher than those
between DZs . However, these premises are valid only if the Equal Environments Assumption (EEA) is
[11]
met. According to the EEA, MZ and DZ twins share to the same extent all environmental factors that are
relevant for the phenotypic expression of the trait under study. If this requirement is not met, because of the
supposedly increased shared environmental elements among MZ twins, the accuracy of their comparison
with DZ twins would be jeopardized. This could result in an overestimation of the influence of genetic
factors, as also noted by Fagnani et al. . Historically, twin studies have been criticized for a possibly
[18]
reduced representativeness, due to a higher rate of pre-term birth in twins compared to singletons.
However, over the years, twins have been shown to provide a reliable picture of all traits of interest in bio-
psycho-social research, including personality, psychopathology, and the attainment of motor development
milestones . The main statistical methods applied in twin studies are represented by the univariate and
[10]
multivariate models, which can be used depending on the aim of the study. The former is used to address
the genetic and environmental contributions to the variance of a single phenotype, whereas the latter is used
[18]
when two or more phenotypes are considered . In a univariate model, the intraclass correlation is
calculated and compared between monozygotic (MZ) and dizygotic (DZ) twins, whereas in multivariate
models, the cross-twin/cross-trait correlation is the main focus of the analysis, as it regards the comparison
of two or more phenotypes observed in the two twins belonging to the same pair (e.g., anxiety as observed
in twin 1 and anger observed in twin 2) . Structural equation models (SEM) are employed to identify the
[18]
contribution of genetic and environmental factors to phenotypic variance and covariance . In these
[14]
models, the considered phenotypes are included as observed variables, whereas A, C, and E are the latent
variables considered. There are three main models useful for estimating the parameters: the Cholesky
model, the Independent Pathway model, and the Common Pathway model. The Cholesky model [Figure 1]