Effects of the Family Environment: Gene–Environment Interaction and

Passive Gene–Environment Correlation

Thomas S. Price and Sara R. Jaffee

University of Pennsylvania

The classical twin study provides a useful resource for testing hypotheses about how the family
environment influences children’s development, including how genes can influence sensitivity to
environmental effects. However, existing statistical models do not account for the possibility that
children can inherit exposure to family environments (i.e., passive gene– environment correlation). The
authors introduce a method to simultaneously estimate the effects of passive gene– environment corre-
lation and gene– environment interaction and use it to investigate the relationship between chaos in the
home and verbal ability in a large sample of 4-year-old twins.

Keywords: twins, epidemiology, gene– environment interaction, gene– environment correlation

Supplemental materials: http://dx.doi.org/10.1037/0012-1649.44.2.305.supp

Developmental psychologists have long been interested in in-

vestigating whether and how children’s family environments in-
fluence their cognitive and behavioral development. Researchers
have demonstrated that environmental factors measured at the
family level, such as socioeconomic status (SES) and geographical
location, are associated with children’s outcomes independent of
individual-level parent or child factors (e.g., Duncan & Brooks-
Gunn, 1997; Leventhal & Brooks-Gunn, 2000) and are often
mediated by more proximal processes (Collins, Maccoby, Stein-
berg, Hetherington, & Bornstein, 2000).

Although many studies have demonstrated associations between

family environmental factors and children’s adjustment, it is
equally true that the effects of family environment can vary from
child to child, even for children raised in the same family. For
example, extreme privation in childhood can cause profound cog-
nitive deficits, but children raised under such conditions vary
widely in terms of their cognitive and socioemotional functioning
(Rutter, O’Connor, & the English and Romanian Adoptees Study
Team, 2004). One possibility is that the effects of adversity are
conditioned by individual differences in children’s resilience or
adaptivity to environmental risk, which may have a genetic basis

(Rutter, 2003). Gene– environment interactions (G

⫻ E) occur

when genetic factors influence sensitivity to environmental effects.
An alternative way of conceptualizing the interaction is to say that
environmental exposure moderates the effect of genetic risk fac-
tors. The existence of such moderating effects would suggest
greater scope for environmental intervention to alter heritable traits
such as cognitive abilities.

However, many challenges remain to identify G

⫻ E. It has

been demonstrated in animal experiments by researchers showing
differential effects of environmental conditions in groups of ani-
mals stratified by their genetic background (e.g., Bennett et al.,
2002). For obvious reasons, it has been much harder to demon-
strate G

⫻ E in humans. Whereas in animal studies both genotype

and environmental exposure can be manipulated for experimental
purposes, in human studies interactions must generally be sought
between naturally occurring variations in genotype and environ-
ment. Epidemiological approaches to G

⫻ E can offer greater

validity than experimental studies, but they do so at the expense of
statistical power and experimental control. One of the drawbacks
of an epidemiological study of G

⫻ E is the possibility of corre-

lation between genotype and environment: the phenomenon of
gene– environment correlation (rGE). Statistical methods for de-
tecting G

⫻ E in human populations need to allow for the possible

lack of independence between genetic and environmental risk
factors (Etheredge, Christensen, Del Junco, Murray, & Mitchell,
2005; Liu, Fallin, & Kao, 2004).

Genotype and environment can correlate for various reasons

(Jaffee & Price, 2007), but studies of the effects of family envi-
ronments on children’s outcomes are particularly subject to con-
founding due to passive rGE (Kendler & Eaves, 1986; Plomin,
1986). Passive rGE occurs when the family environment depends
on heritable parental characteristics, so that parents pass on to their
children an environment that correlates with the parental genotype.
Biological parents also pass on the genotype to their children.
When this genotype also influences children’s behavioral or cog-
nitive outcomes, the result is a spurious association between en-

Thomas S. Price, Institute for Translational Medicine and Therapeutics,

University of Pennsylvania; Sara R. Jaffee, Department of Psychology,
University of Pennsylvania.

We have no financial interests or conflicts of interest related to the

material reported in the article. This work was supported by Grant P50
HL81012 from the National Heart, Blood, and Lung Institute to Thomas S.
Price and Grant R01 HD050691 from the National Institute of Child Health
and Human Development to Sara R. Jaffee.

We thank the participants in the Twins Early Development Study

(TEDS), Robert Plomin, and the TEDS research team, especially Andy
McMillan for providing data management support.

Correspondence concerning this article should be addressed to Thomas

S. Price, Institute for Translational Medicine and Therapeutics, University
of Pennsylvania, Room 807 BRB II/III, 421 Curie Boulevard, Philadelphia,
PA 19104. E-mail: [e-mail]

Developmental Psychology

Copyright 2008 by the American Psychological Association

2008, Vol. 44, No. 2, 305–315

0012-1649/08/$12.00

DOI: 10.1037/0012-1649.44.2.305

305

vironment and outcome (Plomin, DeFries, & Loehlin, 1977; Scarr
& McCartney, 1983). In this way, an association between a mea-
sure of the family environment and a childhood outcome can be
partially or totally accounted for by the effects of parental geno-
type. In the absence of passive rGE this association is attributed to
the influence of the family environment on the outcome. Failure to
rule out the possibility that passive rGE accounts for some portion
of the association may result in its misattribution to environmental
causes.

When the association between measured environment and out-

come is accounted for by unobserved genetic factors (namely,
parental genotype), then the association is said to be genetically
mediated. When the association is accounted for by unobserved
environmental factors, then it is said to be environmentally medi-
ated. This terminology is admittedly somewhat confusing. The
terms genetic mediation and environmental mediation as used by
behavioral geneticists do not imply that genes or environments are
intervening variables in the association between the measured
environment and the outcome; clearly, parental genotype is caus-
ally prior to both measured environment and childhood outcome.

Studies of twin children can exploit differences in genetic re-

latedness between monozygotic (MZ) and dizygotic (DZ) pairs to
quantify the degree to which the effects of environmental exposure
at the level of the individual are genetically mediated and envi-
ronmentally mediated, even in the presence of G

⫻ E (Eaves,

Silberg, & Erkanli, 2003; Rathouz, Van Hulle, Rodgers, & Lahey,
2007). However, current statistical methods for studies of twin
children are uninformative about whether the effects of family-
wide environments are environmentally or genetically mediated
(Turkheimer, D’Onofrio, Maes, & Eaves, 2005). In fact, the meth-
ods that are currently in use implicitly assume the absence of
passive rGE. This problem extends to studies that investigate
possible interactions between genetic influences and the family
environment.

The goal of the current article is to introduce an analytical

method for twin studies that simultaneously estimates G

⫻ E and

passive rGE for measures of the family environment. The motiva-
tions for developing this method are twofold. First, we have
identified a problem with the statistical methodology that is cur-
rently used to investigate the moderating effects of the family
environment, namely, the assumption that there is no passive rGE.
Therefore, we wish to develop an alternative method that does not
suffer the consequences of violating this assumption. The second
motivation is the prospect that passive rGE can be estimated using
data from child twins under specific circumstances: namely, when
genetic influences on the phenotype both correlate with and are
moderated by a measure of the family environment.

In this study, we analyze simulated datasets to investigate

whether these motivations are justifiable. First, we analyze the
simulated datasets using the existing method for detecting G

⫻ E

to quantify the problems that arise when passive rGE is present.
Second, we reanalyze the simulated data using the new model to
outline the range of circumstances under which the simulated
parameter values are accurately recovered.

We illustrate the model with an application to the trait of

childhood verbal ability. Below, we review twin studies that have
attempted to demonstrate how family-wide environments such as
SES and parental education moderate genetic influences on chil-
dren’s cognitive outcomes. We highlight methodological problems

in these studies that originate in their failure to account for possible
effects of passive rGE and explain how the new statistical model
may overcome these shortcomings. Finally, we apply the method
to a large sample of twins and show for the first time that such data
can be used to distinguish true environmental effects from passive
rGE.

Studies of G

⫻ E and Children’s Cognitive Abilities

A series of twin studies has attempted to quantify and test the

moderating effects of family environmental variables (e.g., paren-
tal education, SES) on genetic factors that influence individual
differences in children’s verbal or cognitive abilities (Asbury,
Wachs, & Plomin, 2005; Fischbein, 1980; Guo & Stearns, 2002;
Harden, Turkheimer, & Loehlin, 2007; Kremen et al., 2005; Rowe,
Jacobson, & Van den Oord, 1999; Scarr-Salapatek, 1971; Turkhei-
mer, Haley, Waldron, D’Onofrio, & Gottesman, 2003). The results
have been contradictory. An analysis of data from the National
Longitudinal Study of Adolescent Health concluded that the her-
itability of verbal ability was greater in families with highly
educated parents (Rowe et al., 1999), although a reanalysis as-
cribed the moderating effect to employment status and race (Guo
& Stearns, 2002). Other studies have also found that the genetic
influences on cognitive abilities are stronger in families in which
parents have more education (Kremen et al., 2005) or higher SES
(Harden et al., 2007; Turkheimer et al., 2003). In contrast, a large
study of 4-year-old twins did not find that heritability estimates
varied as a function of SES (Asbury et al., 2005) but that herita-
bility estimates were higher in high-risk families characterized by
high levels of chaos and poor parent– child communication—
aspects of the environment that typically correlate with low SES
(Asbury et al., 2005; Evans, 2004). Moreover, these reports of G

⫻

E have not been confirmed in studies of nontwin families (Nagoshi
& Johnson, 2005; Van den Oord & Rowe, 1997).

The twin studies reviewed above used either a structural equa-

tion modeling framework (e.g., Turkheimer et al., 2003), a mixed
model (Guo & Stearns, 2002), or a DeFries-Fulker regression
model (e.g., Rowe et al., 1999) to test hypotheses about the effects
of the family environment on cognitive outcomes. These methods
estimate the overall association between family environment and
the phenotype but do not distinguish between environmentally
mediated effects and passive rGE (Turkheimer et al., 2005). In
effect, the influences of latent genetic and environmental factors
are estimated from the variation in the phenotype that remains after
estimating a main effect of the measured family environment
(Purcell & Koenen, 2005; Turkheimer et al., 2005). This is equiv-
alent to assuming that the association between the measured family
environment and the child phenotype is mediated entirely through
the shared environment. In the presence of passive rGE this im-
plicit assumption is violated, so that these procedures not only
misspecify the effect of the measured environment but also mis-
specify the effects of the latent shared environmental factor. A
further consequence of the presence of passive rGE is that the
phenotypic variance cannot be resolved into separate genetic and
environmental components in the usual way (Rathouz et al., un-
published manuscript). Crucially, at least two studies have dem-
onstrated that rGEs are likely to account for part of the association
between SES or parental education and offspring cognitive abili-
ties (Neiss, Rowe, & Rodgers, 2002; Tambs, Sundet, Magnus, &

306

PRICE AND JAFFEE

Berg, 1989), suggesting that estimates of G

⫻ E as predictors of

cognitive abilities might be biased. The same problem may apply
to twin studies of G

⫻ E in other phenotypes: For example, the

relationship between family dysfunction and children’s antisocial
behavior may be genetically rather than environmentally mediated
(Button, Scourfield, Martin, Purcell, & McGuffin, 2005). On the
other hand, studies that have been careful to measure environments
that are not likely to be genetically correlated with the outcome,
such as geographical region, are less vulnerable to this criticism
(e.g., Dick, Rose, Viken, Kaprio, & Koskenvuo, 2001).

Need for New Statistical Methods

In this article we suggest a methodological innovation that

addresses the problem we have identified. We introduce a statis-
tical model for the classical twin design that estimates both the
environmentally mediated effects of the family environment and
passive rGE in the presence of G

⫻ E. The influence of the

measured family environment is modeled as a random effect that
may correlate with genotype rather than a fixed main effect, an
approach that allows both the genetically mediated effects and the
environmentally mediated effects of the measured environment to
be estimated from the data. Similar analytic strategies have been
suggested previously in relation to child-specific environments
(Eaves et al., 2003; Purcell, 2002). Simulation studies have been
performed that quantify the deficiencies of the existing method and
validate the proposed analytical procedure.

We illustrate the new method using data from a large twin study

of early cognitive development. A previous report from this study
found evidence that chaos in the family home and features of
parent– child communication style moderated genetic influences
on verbal ability at age 4 (Asbury et al., 2005). We selected family
chaos for our analysis to facilitate comparison with the existing
literature on the moderating effects of distal family environments
over genetic influences on cognitive development.

Chaos reflects the child’s physical microenvironment, including

the child’s exposure to noise, crowding, and patterns of environ-
mental traffic (Matheny, Wachs, Ludwig, & Phillips, 1995). Fam-
ily chaos may retard cognitive development by causing children to
filter out useful environmental stimuli along with unwanted noise
(Evans, 2006). Parent– child interactions in noisy or crowded
homes are also less conducive to cognitive development because
parents are less responsive to their children (Evans, 2006). Chaos
correlates with SES and may function as a proximal mediator of its
effects (Asbury et al., 2005; Pike, Iervolino, Eley, Price, & Plomin,
2006). It has previously been shown that this measure of the family
environment not only correlates with verbal ability (Petrill, Pike,
Price, & Plomin, 2004; Pike et al., 2006) but also moderates the
effect of genetic influences on verbal ability (Asbury et al., 2005).
Although these findings are premised on the assumption that chaos
has environmentally mediated effects on children’s verbal ability,
an alternative hypothesis is that parents who raise their children in
chaotic home environments also pass along genetic variants that
are associated with poor verbal ability and that, in fact, the asso-
ciation between chaos and children’s verbal ability is partly ge-
netically mediated. In this study, we use a new statistical method
to show that a previously reported association between chaos at
home and early verbal ability cannot be explained by passive rGE.

Method

Statistical Model

In this section we first outline a model for the effects of a

measured family environment on twin phenotypes that parameter-
izes the effects of passive rGE and explains why it cannot be
successfully estimated. Next, we extend the model to account for
the effects of both G

⫻ E and passive rGE and show that the

existence of G

⫻ E allows the main effect of the measured family

environment to be distinguished from passive rGE.

Let us say that we are interested in understanding the sources of

variation in a phenotype like children’s verbal ability. In the
standard biometric model for twin data, the phenotype Y

ij

for twin

j in family i is determined by the population mean

␮ and the values

of the random variables A

ij

, C

i

, and E

ij

that represent additive

genetics, shared environment, and nonshared environment, respec-
tively. We assume that the latent variables A, C, and E are
independently normally distributed and load on the phenotype with
coefficients a, c, and e, respectively. In order for the model to be
identified, it is necessary to provide an arbitrary location and scale for
the latent genetic and environmental variables. In accordance with
convention, we scale these latent variables to zero mean and unit
variance. Because MZ twins have the same genomic DNA, whereas
DZ twins share half their segregating genes, the genetic factors for
twins in the same family, A

i1

and A

i2

, are correlated with Coefficient

1 for MZ twins and with Coefficient 0.5 for DZ twins. The nonshared
environments E

ij

are uncorrelated within members of the same family.

We supplement this model with an additional random variable X

representing a measured family environment—that is, a variable
measured at the family level whose value differs between families
but not within families (e.g., family chaos). X is normally distrib-
uted with zero mean and variance

␴

X

2

, has a main effect x on the

phenotype, and is correlated with A due to a passive rGE, such that
Cor(X

i,

A

i1

)

⫽ Cor( X

i,

A

i2

)

⫽ r. Under this model, the value of the

phenotype Y

ij

for twin j in family i is given by

Y

ij

⫽ aA

ij

⫹ cC

i

⫹ eE

ij

⫹ xX

i

⫹ ␮.

(1)

Because we are interested in the effects of an environmental factor
that differs between families but not within families, it is illumi-
nating to rearrange the model for the phenotypic scores into
separate terms for the half sum (the within-family mean, corre-
sponding to the component of the twins’ phenotypic scores that
differs between families) and the half difference (the component of
the twins’ phenotypic scores that varies within families):

1

2

共Y

i1

⫹ Y

i2

兲 ⫽

1

2

a

共A

i1

⫹ A

i2

兲

⫹ cC

i

⫹

1

2

e

共E

i1

⫹ E

i2

兲 ⫹ xX

i

⫹ ␮, (2)

1

2

共Y

i1

⫺ Y

i2

兲 ⫽

1

2

a

共A

i1

⫺ A

i2

兲 ⫹

1

2

e

共E

i1

⫺ E

i2

兲

Note that the sum of these terms is Y

i1

(i.e., the score for Twin

1 in family i) and the difference between them is Y

i2

(i.e., the score

for Twin 2 in family i). The variance in the phenotypic scores can
be partitioned into a component of variance due to differences
between families,

␴

b

2

, accounted for by factors that differ between

307

SPECIAL SECTION: EFFECTS OF THE FAMILY ENVIRONMENT

families such as the main effect of X, and a component of variance
due to differences within families,

␴

w

2

, which cannot be accounted

for by the main effect of X because X takes the same value for all
children in a family. Let us assume that the variables follow a
multivariate normal distribution and there are no systematic effects
of birth order. The latter is not generally considered a controversial
assumption for behavioral data simply because twins are so close
in age, although there is some evidence that perinatal risk is
elevated in second-born twins (Armson et al., 2006). Under these
assumptions the following equality holds:

Var

共Y

i1

兲 ⫽ Var 共Y

i2

兲 ⫽ ␴

b

⫹ ␴

w

⫽ Var

冋

1

2

共Y

i1

⫹ Y

i2

兲

册

⫹ Var

冋

1

2

共Y

i1

⫺ Y

i2

兲

册

.

(3)

The variances of the half-sums and half-differences correspond,
respectively, to the between-family variance

␴

b

2

and the within-

family variance

␴

w

2

, and are uncorrelated. The values of

␴

b

2

and

␴

w

2

depend on the zygosity of the twin pair, because genetic differences
between twins can contribute to within-pair differences for DZ pairs
but not for MZ pairs. It can be shown that the between- and within-
family phenotypic variances for MZ and DZ pairs are given by:

␴

bMZ

2

⫽ a

2

⫹ c

2

⫹

1

2

e

2

⫹ ␹

2

␴

X

2

⫹ 2ax␴

X

r,

(4)

␴

wMZ

2

⫽

1

2

e

2

,

␴

bDZ

2

⫽

3

4

a

2

⫹ c

2

⫹

1

2

e

2

⫹ x

2

␴

X

2

⫹ 2ax␴

X

r,

␴

wDZ

2

⫽

1

4

a

2

⫹

1

2

e

2

.

We can see that a positive association between the measured

family environment and the phenotype increases the variance of
the phenotype between families. This association has two compo-
nents: an environmental component that is mediated solely by the
latent environmental variable X, and a genetic component due to
the passive rGE between A and X. The phenotypic variance ac-
counted for by the environmental component of the association is
simply the square of the main effect of the environmental influence
on the phenotype, x

2

␴

X

2

. The phenotypic variance due to passive

rGE equals twice the covariance between A and X

共2ax␴

X

r

兲. Note

that the within-family phenotypic variance does not depend on
either x or r. The lack of dependence of within-family variance on
x is not surprising, because we model the effect of the family
environment the same way for both twins in a pair. The reason that
passive rGE makes no contribution to within-family differences
can be understood intuitively as follows. Genetic differences
within DZ twin pairs arise from recombination during meiosis, a
random process that is uncorrelated with the parental genotypes
and hence with the processes that cause passive rGE. As a final
point, it is important to note that the environmentally and geneti-
cally mediated effects of the measured family environment on the
between-family and within-family variances are confounded in
this model. There are four pieces of information, and there are five

parameters (namely a, c, e, x and r) to be estimated (the variance
␴

X

can be estimated directly from the data). This means that x and

r are not identified: It is not possible to estimate unique values for
these parameters from the data.

Let us now extend the model with additional terms describing

the moderating effects of the measured family environment. We
can model G

⫻ E by allowing a moderating effect of the measured

environment on the genotype so that the coefficient of genetic
influence on the phenotype is given by a

共1 ⫹ m

A

X

兲, where m

A

is

a linear moderation term. Nonzero values for this term imply that
the genetic influences on the phenotype vary across levels of the
measured environmental variable. We can also allow linear mod-
eration of the paths for the shared environment, c

共1 ⫹ m

C

X

兲, and

nonshared environment, e

共1 ⫹ m

E

X

兲. The variances of the phe-

notypic sums and differences are now given by:

␴

bMZ

2

⫽ a

2

共1 ⫹ m

A

X

兲

2

⫹ c

2

共1 ⫹ m

C

X

兲

2

⫹

1

2

e

2

共1 ⫹ m

E

X

兲

2

⫹ x

2

␴

X

2

⫹ 2a共1 ⫹ m

A

X

兲x␴

X

r,

(5)

␴

wMZ

2

⫽

1

2

e

2

共1 ⫹ m

E

X

兲

2

,

␴

bDZ

2

⫽

3

4

a

2

共1 ⫹ m

A

X

兲

2

⫹ c

2

共1 ⫹ m

C

X

兲

2

⫹

1

2

e

2

共1 ⫹ m

E

X

兲

2

⫹ x

2

␴

X

2

⫹ 2a共1 ⫹ m

A

X

兲x␴

X

r,

␴

wDZ

2

⫽

1

4

a

2

共1 ⫹ m

A

X

兲

2

⫹

1

2

e

2

共1 ⫹ m

E

X

兲

2

.

The covariance between the measured environment X and the
phenotype Y, c

XY

, comprises terms relating to the environmentally

and genetically mediated effects of the measured environment:

c

XY

⫽ x␴

X

2

⫹ a共1 ⫹ m

A

X

兲␴

X

r.

(6)

The covariance between measured environment and phenotype
due to passive rGE, a

共1 ⫹ m

A

X

兲␴

X

r, is a linear function of X,

whereas the covariance due to the environmentally mediated ef-
fect, x

␴

X

2

, is constant with respect to X. Consequently, if it is

known (or can be shown in advance) that there is moderation of the
genetic variance such that m

A

and a are nonzero, then a model that

estimates the covariance between measured environment and phe-
notype as a linear function of X is identified and allows unique
values for x and r to be estimated from the data. In other words, the
existence of G

⫻ E allows one to distinguish between the envi-

ronmentally and genetically mediated effects of the measured
family environment. For example, if the environmental exposure X
is binary with values corresponding to exposure/nonexposure, then
the equations in (5) will provide eight pieces of information—four
variances for each value of X—sufficient to estimate the eight
parameters in the model. Naturally, the power to discriminate
between the main effect of the environment and passive rGE will
depend strongly on the values of the genetic path parameter a and
the genetic moderation parameter m

A

: Power will increase as the

magnitude of these parameters increases.

Because the covariance between X and Y is entirely due to the

covariance between X and the half-sums, which also equals c

XY

,

308

PRICE AND JAFFEE

the covariances between the variables in our model can be de-
scribed completely by inserting the values from the equations in
(5) and (6) into the following structural equations:

MZ: Var

冉

1

2

关Y

i1

⫹ Y

i2

兴,

1

2

关Y

i1

⫺ Y

i2

兴,X

i

冊

⫽

冉

␴

bMZ

2

0

c

XY

0

␴

wMZ

2

0

c

XY

0

␴

X

2

冊

,

(7)

DZ: Var

冉

1

2

关Y

i1

⫹ Y

i2

兴,

1

2

关Y

i1

⫺ Y

i2

兴,X

i

冊

⫽

冉

␴

bDZ

2

0

c

XY

0

␴

wDZ

2

0

c

XY

0

␴

X

2

冊

.

The expected vector of means for, respectively, the half-sums,
half-differences, and measured environment is (

␮, 0, 0) for both

zygosity groups. However, in practice it may be better to estimate
the means for MZ and DZ groups without imposing any con-
straints to capture any mean effects due to zygosity and birth order.

A path diagram corresponding to this model of environmental
mediation and moderation is shown in Figure 1. For simplicity of
presentation, the means model is omitted. The model can either be
implemented within a structural equation modeling paradigm and
estimated by maximum likelihood—an example script is provided
as part of the supplementary materials online in Supplemental
Appendix I, for the freely distributed software package Mx (http://
www.vcu.edu/mx/; Neale, Boker, Xie, & Maes, 1999)— or imple-
mented as a Bayesian model and estimated by Markov chain
Monte Carlo methods. Bayesian models allow very flexible pa-
rameterization: A script for the freely distributed program win-
BUGS 1.4.1 (www.mrc-bsu.cam.ac.uk/bugs/; Spiegelhalter,
Thomas, Best, & Lunn, 2003) is provided online in Supplemental
Appendix II.

Data Analysis

As mentioned previously, the statistical model described by the

equations in (7) can only be estimated when it is already known
that the measured family environment moderates the genetic vari-
ance (i.e., m

A

⫽ 0). For this reason, the full model—containing

both r and m

A

parameters— cannot, by itself, provide a test for the

(1+m

C

X)

2

A

b

½(Y

1

+Y

2

)

a

C

b

E

b

c

√½ e

√½ e

E

w

X

X

1

x

r

A

b

√¾ a

C

b

E

b

√½ e

E

w

X

X

1

x

½(Y

1

+Y

2

)

½(Y

1

+Y

2

)

A

w

√¼ a

½(Y

1

–Y

2

)

√

4

/

3

r

½(Y

1

+Y

2

)

½(Y

1

–Y

2

)

(1+m

A

X)

2

(1+m

A

X)

2

(1+m

C

X)

2

c

(1+m

E

X)

2

(1+m

E

X)

2

√½ e

(1+m

E

X)

2

(1+m

E

X)

2

(1+m

A

X)

2

s

X

2

s

X

2

MZ

DZ

Figure 1.

Path diagrams for monozygotic (MZ) and dizygotic (DZ) twin pairs, showing observed variables

(square boxes), latent variables (circles), regression paths (single-headed arrows) and correlations (double
headed arrows). The means model has been omitted. X

⫽ measured family environment; ␴x

2

⫽ the variance of

the measured family environment;

1

⁄

2

(Y

1

⫹Y

2

)

⫽ half sum of twin phenotypes;

1

⁄

2

(Y

1

–Y

2

)

⫽ half difference

between twin phenotypes; A

⫽ additive genetics; C ⫽ shared environment; E ⫽ nonshared environment; b

suffix

⫽ between-family variance; w suffix ⫽ within-family variance; m

C

⫽ linear moderation of shared

environmental path; r

⫽ correlation due to passive rGE; m

A

⫽ linear moderation of genetic path; a ⫽ additive

genetic path parameters; c

⫽ shared environment path parameters; e ⫽ nonshared environment path parameters;

x

⫽ environmentally mediated effect of the measured environment.

309

SPECIAL SECTION: EFFECTS OF THE FAMILY ENVIRONMENT

statistical significance of the genetic moderation parameter m

A

.

The usual way to test whether or not a parameter in a structural
equation model is statistically significant is to estimate an alter-
native model in which the parameter is fixed at zero. The param-
eter is considered significant if the reduced model provides a
significantly worse fit to the data, as measured using the likelihood
ratio test or by other methods of hypothesis testing such as the
comparison of Akaike’s information criterion fit statistics (Akaike,
1974). As we have shown, the alternative model with r freely
estimated and m

A

⫽ 0 is not identified. The existence of G ⫻ E

can, however, be established by testing the significance of m

A

in a

reduced model in which r is fixed to zero. This reduced model is
essentially a reparameterization of the existing structural equation
model used in, for example, Turkheimer et al. (2003).

We therefore recommend a two-step strategy for data analysis.

The first step is to test for genetic moderation by the measured
environment using the statistical model with r fixed to zero. If the
results of the first analysis indicate significant effects of genetics
and G

⫻ E, the model can be rerun with r freely estimated in order

to estimate the effects of passive rGE in addition to the other
parameters.

Simulation Studies

A series of simulations was conducted to validate the statistical

methods that we have described. Four simulations were conducted,
each containing 500 datasets consisting of 500 pairs each of MZ
and DZ twins, using the parameter values a

⫽

冑

0.5, c

⫽

冑

0.25, e

⫽

冑

0.25, x

⫽ 0, r ⫽ k, and m

A

⫽ 0, indicating no G ⫻ E for each

of the values k

⫽ 0, 0.1, 0.2, 0.3. (The method by which the

datasets were simulated is available from the corresponding author
on request.) These values correspond to a heritable phenotype with
shared and nonshared environmental influences, with a measure of
the family environment that has no environmentally mediated
effects on the phenotype, correlates nonnegatively with genetic
influences on the phenotype, and does not moderate the genetic
effects. A further three sets of 500 datasets were simulated using
the parameter values a

⫽

冑

0.5, c

⫽

冑

0.25, e

⫽

冑

0.25, x

⫽ k, r ⫽

k, and m

A

⫽ 0 for each of the values k ⫽ 0.1, 0.2, 0.3. These values

correspond to a heritable phenotype with shared and nonshared
environmental influences, with a measure of the family environ-
ment that is positively correlated with the phenotypes due to both
environmentally mediated and genetically mediated effects and
does not moderate the genetic effects. Finally, six sets of 500
datasets were simulated, similar to the others but with x

⫽ 0 or k,

r

⫽ k, and m

A

⫽ k for k ⫽ 0.1, 0.2, 0.3, corresponding to the

presence of both rGE and G

⫻ E.

Results

The simulated data were first analyzed using an existing struc-

tural equation model (http://www.psy.vu.nl/mxbib/mx_show_
script.php?page

⫽rawVCmod1 as used in, e.g., Turkheimer et al.,

2003) in which the absence of passive rGE is implicitly assumed
(Purcell & Koenen, 2005; Turkheimer et al., 2005; see Table 1).
This model was constrained to estimate linear moderation on the
genetic path parameter only and did not estimate moderation of the
shared or nonshared environment paths. The results were reparam-
eterized slightly, because this model estimates a genetic path a

⫹

uX rather than a(1

⫹m

A

X), so that m

A

⫽ u /a for a ⫽ 0. The model

did not falsely identify G

⫻ E when it was absent and recovered

accurate estimates of m

A

. However, as noted previously, this

model cannot distinguish between genetically mediated and envi-
ronmentally mediated effects of the measured family environment
(Purcell & Koenen, 2005; Turkheimer et al., 2005). As a result,
genetically mediated effects of the family environment (i.e., sim-
ulated values of r

⬎ 0) were estimated as a main effect of the

measured environment (i.e., estimated values of x

⬎ 0). In addi-

tion, estimates of the squared shared environmental path parameter
c

2

were increasingly biased down from the simulated value as the

simulated strength of the passive rGE increased. Similar results
can be obtained using augmented DeFries–Fulker analysis (as used
by, e.g., Rowe et al., 1999), although because this model tests for
linear moderation of the standardized variance components h

2

and

c

2

rather than the genetic and environmental path parameters, the

datasets must be simulated in a slightly different way (data not
shown).

We reanalyzed these simulated datasets using the new structural

equation model with r fixed to zero (Table 2). As for the previous
analyses, the model was constrained to estimate linear moderation
on the genetic path parameter only. The results were virtually
identical, indicating the near-equivalence between these two mod-
els: The only real difference between them is that the new model
estimates an additional parameter for the variance of the measured
environmental variable. It is important that the absence of bias in
the test of G

⫻ E provided justification for the proposed two-step

analytic strategy.

Finally, we analyzed the simulated datasets with both geneti-

cally and environmentally mediated effects of the measured family
environment and G

⫻ E using the full structural equation model.

The mean, median, standard deviation, and 2.5% and 97.5% quan-
tiles (defining an empirical symmetric 95% confidence interval
[CI]) for the parameter estimates, as well as the proportion of
statistically significant estimates of m

A

, r, and x, are shown in

Table 3. The mean parameter estimates were close to the simulated
value, demonstrating that in principle this model can accurately
estimate the effects of rGE, G

⫻ E, and environmentally mediated

effects of the family environment. However, the variability around
these estimates, especially those for c

2

, r, and x in datasets with the

smallest simulated values of m

A

, r, and x, emphasizes the need for

large sample sizes. It is no surprise that there is little power to
distinguish genetic from environmental mediation of the effects of
the measured environment when the true value of m

A

is close to

zero. Indeed, in the absence of G

⫻ E, r and x are completely

confounded, as we have already noted.

Twins Early Development Study

Sample.

We illustrate our analytical approach by using it to

estimate the moderating effects of family chaos on genetic influ-
ences on verbal ability at 4 years. The Twins Early Development
Study (TEDS) is a population-based study of twins born in the
United Kingdom between 1994 and 1996 (Trouton, Spinath, &
Plomin, 2002). The families participating in TEDS are represen-
tative of the U.K. population of parents of young children in terms
of parental education, ethnicity, and employment status, which
were assessed from parental responses to questionnaire booklets
sent to the families in the twins’ 2nd year (Trouton et al., 2002).

310

PRICE AND JAFFEE

Parental ratings of physical similarity were used to determine the
zygosity of the twins, a method that assigns zygosity with more
than 95% accuracy as validated by genotyping (Price et al., 2000).
Families were excluded if English was not the principal language
spoken in the home, if zygosity was uncertain, or if either twin had
severe medical, genetic, or perinatal problems. Data were also
excluded if the test booklets sent to parents when twins were 4
years old were incomplete or were returned more than 6 months
late. Opposite-sex twin pairs were excluded from the current
investigation to simplify the statistical analysis. The final sample
consisted of 943 families with male MZ twins, 1,099 families with

female MZ twins, 1,001 families with male DZ twins, and 1,042
families with female DZ twins.

Verbal ability.

Verbal ability was assessed at age 4 with the

Expressive Vocabulary and Grammatical Complexity subscales of
the MacArthur Communicative Development Inventory (MCDI;
Fenson et al., 1994). These measures were parent administered.
The MCDI is widely used and demonstrates excellent internal
consistency, test–retest reliability, and concurrent validity with
tester-administered measures (Fenson et al., 1994). Expressive
vocabulary is assessed with a multi-item checklist on which par-
ents report on their children’s production of root words. We

Table 1
Results of Simulation Study to Test for the Effects of G

⫻ E With an Existing Structural Equation Model

Simulated value

a

2

c

2

e

2

m

A

x

⌬

⫺2LL

p

⬍ .05 (%)

p

⬍ .01 (%)

m

A

x

r

M

SD

M

SD

M

SD

M

SD

M

SD

M

SD

0

0

0

.50

.07

.24

.06

.25

.02

.00

.04

.00

.03

1.01

1.54

5.0

1.2

0

0

.1

.50

.07

.24

.06

.25

.02

.00

.04

.07

.03

1.01

1.54

5.0

1.2

0

0

.2

.49

.07

.23

.06

.25

.02

.00

.04

.14

.03

0.96

1.37

5.6

0.6

0

0

.3

.50

.07

.20

.07

.25

.02

.00

.03

.21

.03

0.98

1.27

3.4

0.6

0

.1

.1

.50

.07

.24

.07

.25

.02

.00

.03

.17

.03

1.00

1.35

5.4

0.8

0

.2

.2

.49

.07

.23

.07

.25

.02

.00

.04

.34

.03

1.03

1.36

5.4

0.8

0

.3

.3

.50

.07

.20

.07

.25

.02

.00

.03

.51

.03

1.03

1.38

5.8

0.8

.1

0

.1

.50

.07

.24

.06

.25

.02

.10

.04

.07

.03

9.13

5.79

80.8

62.0

.2

0

.2

.50

.07

.23

.07

.25

.02

.20

.05

.13

.03

31.8

11.2

100.0

100.0

.3

0

.3

.49

.07

.23

.06

.25

.02

.30

.07

.17

.03

63.8

16.3

100.0

100.0

.1

.1

.1

.49

.07

.25

.06

.25

.02

.10

.04

.17

.03

8.91

5.81

80.0

59.2

.2

.2

.2

.49

.07

.24

.06

.25

.02

.20

.05

.33

.03

31.3

11.3

100.0

99.6

.3

.3

.3

.48

.07

.23

.06

.25

.02

.30

.07

.48

.03

63.1

16.6

100.0

100.0

Note.

Simulation study tested for the effects of G

⫻ E by use of an existing structural equation model (Turkheimer et al., 2003), including maximum

likelihood estimates of squared additive genetic (a

2

), shared environment (c

2

), and nonshared environment (e

2

) path parameters; linear moderation of genetic

path (m

A

) and environmentally mediated effect of the measured environment (x); change in fit (

⌬

⫺2LL

) and proportion of models with statistically significant

changes in fit at p

⬍ .05 and p ⬍ .01 when the moderation term is dropped from the model. Each value is the result of analyzing 500 simulated datasets.

Simulated values of m

A

, x, and correlation due to passive rGE (r) are given in the first three columns on the left.

Table 2
Results of Simulation Study to Test for the Effects of G

⫻ E With a Reduced Version of the New Structural Equation Model

Simulated value

a

2

c

2

e

2

m

A

x

⌬

⫺2LL

p

⬍ .05 (%)

p

⬍ .01 (%)

m

A

x

r

M

SD

M

SD

M

SD

M

SD

M

SD

M

SD

0

0

0

.50

.07

.24

.07

.25

.02

.00

.04

.00

.03

0.98

1.33

5.0

1.0

0

0

.1

.50

.07

.24

.06

.25

.02

.00

.04

.07

.03

1.05

1.63

5.8

1.8

0

0

.2

.49

.07

.23

.07

.25

.02

.00

.04

.14

.03

0.95

1.33

5.4

0.8

0

0

.3

.50

.07

.20

.07

.25

.02

.00

.04

.21

.03

0.98

1.29

4.2

0.6

0

.1

.1

.50

.07

.24

.07

.25

.02

.00

.04

.17

.03

0.98

1.31

5.2

0.6

0

.2

.2

.50

.07

.23

.07

.25

.02

.00

.04

.34

.03

0.97

1.29

4.8

0.4

0

.3

.3

.50

.07

.20

.07

.25

.02

.00

.04

.51

.03

0.97

1.31

4.4

0.8

.1

0

.1

.50

.07

.25

.06

.25

.02

.10

.04

.07

.03

7.69

5.20

76.6

50.6

.2

0

.2

.49

.07

.23

.07

.25

.02

.20

.05

.13

.03

26.7

10.3

99.8

99.6

.3

0

.3

.49

.07

.23

.06

.25

.02

.30

.07

.17

.03

53.3

15.2

100.0

100.0

.1

.1

.1

.49

.07

.25

.07

.25

.02

.10

.04

.17

.03

7.61

5.42

71.8

49.4

.2

.2

.2

.49

.07

.24

.06

.25

.02

.20

.05

.33

.03

26.2

10.1

99.6

98.8

.3

.3

.3

.48

.07

.23

.06

.25

.02

.30

.07

.48

.03

52.1

15.4

100.0

100.0

Note.

Simulation study tested for the effects of G

⫻ E by use of a reduced version of the new structural equation model, including maximum likelihood

estimates of squared additive genetic (a

2

), shared environment (c

2

), and nonshared environment (e

2

) path parameters; linear moderation of genetic path (m

A

)

and environmentally mediated effect of the measured environment (x); change in fit (

⌬

⫺2LL

); and proportion of models with statistically significant changes

in fit at p

⬍ .05 and p ⬍ .01 when the moderation term is dropped from the model. Each value is the result of analyzing 500 simulated datasets. Simulated

values of m

A

, x, and correlation due to passive rGE (r) are given in the first three columns on the left.

311

SPECIAL SECTION: EFFECTS OF THE FAMILY ENVIRONMENT

calculated a composite score by summing the number of words
checked. The Grammatical Complexity subscale examines
whether and how children combine words. We calculated a total
verbal ability score by summing standardized scores on these two
measures (Spinath, Price, Dale, & Plomin, 2004).

Chaos.

The degree of chaos in the home was assessed at the

same time by parental report using the short version of the Con-
fusion, Hubbub, and Order Scale (Matheny et al., 1995), which
includes items such as “You can’t hear yourself think in our home”
and “We are usually able to stay on top of things.” The score is
derived by summing six items rated on a 5-point scale and has
been validated against direct observations (Matheny et al., 1995).
These items show acceptable internal consistency in the TEDS
sample (Cronbach’s

␣ ⫽ .63). To aid convergence of the analytic

model, we centered and scaled the variable to unit variance.

Descriptive statistics.

Details of the distributions of these vari-

ables in the TEDS sample are available in previous publications on
the TEDS sample (Asbury et al., 2005; Petrill et al., 2004; Pike et
al., 2006; Spinath et al., 2004). The analyses of verbal ability and
chaos in the home presented here employ the standardized resid-
uals after removing the linear effects of age and sex.

Phenotypic and twin correlations.

Chaos and verbal ability

correlated at –.19 ( p

⬍ .01). The intraclass twin correlations are

presented in Figure 2 as a function of zygosity and degree of chaos
in the home. A trend was evident toward lower DZ correlations in
more chaotic homes. Sex differences in twin correlations for the
4-year verbal measure were relatively small, as noted in a previous
analysis of this dataset (Spinath et al., 2004).

Figure 2.

Intraclass twin correlations for verbal ability at age 4 by

zygosity and degree of chaos in the home. MZ

⫽ monozygotic; DZ ⫽

dizygotic.

Table 3
Results of Simulation Study to Quantify the Effects of G

⫻ E in the Presence of Passive rGE

With the New Structural Equation Model

Variable

a

2

c

2

e

2

m

A

x

r

m

A

, x, r

⫽ .1

M

.49

.29

.25

.11

.11

.09

SD

.07

.09

.02

.04

.20

.29

2.5%

.36

.14

.22

.04

⫺.29

⫺.50

50%

.49

.28

.25

.11

.12

.09

97.5%

.62

.48

.28

.20

.51

.64

p

⬍ .05 (%)

86.6

5.6

6.8

m

A

, x, r

⫽ .2

M

.50

.26

.25

.21

.20

.19

SD

.07

.06

.02

.05

.10

.14

2.5%

.37

.13

.22

.13

.01

⫺.06

50%

.49

.26

.25

.21

.21

.19

97.5%

.63

.38

.28

.31

.39

.47

p

⬍ .05 (%)

100.0

57.4

28.8

m

A

, x, r

⫽ .3

M

.50

.25

.25

.31

.30

.29

SD

.07

.06

.02

.06

.06

.09

2.5%

.37

.13

.22

.21

.17

.13

50%

.50

.25

.25

.30

.30

.29

97.5%

.64

.35

.28

.43

.41

.46

p

⬍ .05 (%)

100.0%

98.6

89.0

Note.

Simulation study quantified the effects of G

⫻ E in the presence of passive rGE by use of the new

structural equation model, including linear moderation of genetic path (m

A

) and environmentally mediated effect

of the measured environment (x); correlation due to passive rGE (r); maximum likelihood estimates of squared
additive genetic (a

2

), shared environment (c

2

), and nonshared environment (e

2

) path parameters; change in fit

(

⌬

⫺2LL

); and proportion of models with statistically significant changes in fit at p

⬍ .05 and p ⬍ .01 when the

moderation term is dropped from the model. Statistics and the proportion of models in which dropping the
relevant term results in significant deterioration in fit according to the likelihood ratio test at p

⬍ .05 are given

for each set of estimates resulting from the analysis of 500 simulated datasets.

312

PRICE AND JAFFEE

Test of G

⫻ E. The data were first analyzed using a reduced

version of the model described above, with the parameters for r,
m

C

, and m

E

fixed to zero. Previous explorations of these data have

indicated only small differences in parameter estimates between
males and females (Spinath et al., 2004); for this reason, the path
parameters for males and females were equated in these analyses,
except that different mean values were estimated for males and
females. A second model, further constrained with m

A

fixed to

zero, provided significantly worse fit (

⌬

⫺2LL

⫽ 71.7, ⌬

df

⫽ 1, p ⬍

.0001), indicating substantial moderation of genetic effects by
family chaos. The results suggested that genetic influences on
verbal ability were stronger in more chaotic households.

Estimation of model parameters.

Having established the pres-

ence of significant G

⫻ E, we estimated an enlarged model with

the parameter r free to vary and m

C

and m

E

fixed to zero. This

model estimates the effect of G

⫻ E while distinguishing geneti-

cally from environmentally mediated effects of chaos on children’s
verbal abilities. There was a moderate effect of additive genetics
(a

2

⫽ 0.33; 95% CI 0.29, 0.37), a substantial component of shared

environment (c

2

⫽ 0.49; 95% CI 0.45, 0.54), and a small compo-

nent of nonshared environment (e

2

⫽ 0.12; 95% CI 0.11, 0.13),

consistent with a previous analysis of this dataset (Spinath et al.,
2004). There was significant moderation of the genetic variance
(m

A

⫽ 0.17; 95% CI 0.13, 0.22) by chaos. Each increment (or

decrement) of one standard deviation in the chaos measure was
associated with a 17% increase (or decrease) in the genetic path
parameter. Genetic influences were stronger on children growing
up in chaotic homes. Although the total variability was also greater
in these children, the proportion of genetic variance (heritability)
was also greater in children from chaotic homes, consistent with
the findings reported in Asbury et al. (2005). The 95% CI for the
standardized main effects of chaos on verbal ability, x, extended
from

⫺0.03 to ⫺0.29, with a point estimate of ⫺0.16 that was

statistically different from zero (

⌬

⫺2LL

⫽ 6.16, ⌬

df

⫽ 1, p ⫽ .01).

The 95% CI for the estimate of passive rGE r spanned from

⫺0.24

to 0.18, with a nonsignificant point estimate of

⫺0.03 (⌬

⫺2LL

⫽

0.089,

⌬

df

⫽ 1, p ⫽ .77).

Discussion

Existing statistical methods for testing hypotheses about the

effects of the measured family environment on children’s devel-
opment using twin data implicitly assume the absence of passive
rGE. These include the methods that have been used to investigate
the effects of G

⫻ E, which—in the presence of passive rGE—are

liable to misspecify both the effects of the measured environment
and the effects of the latent shared environmental factor, although
it is important to note that estimates of the G

⫻ E itself appear to

be unaffected. We have demonstrated a new analytic strategy that
detects and quantifies the effects of passive rGE in the presence of
G

⫻ E. We have presented results from simulated datasets that

attest to the viability of this method over a range of parameter
values.

To illustrate this novel statistical model, we have performed an

analysis that confirmed a previous finding that genetic influences
on children’s abilities are stronger among children in the TEDS
sample who were raised in a chaotic home environment (Asbury et
al., 2005). These results are consistent with a diathesis–stress
model in which a genetic predisposition for some outcome (e.g.,

poor verbal ability) is more strongly expressed in high risk envi-
ronments (Rosenthal, 1963). The new method was sufficiently
powerful to exclude the possibility that the association between
chaos in the home and children’s verbal ability could be explained
solely by passive rGE. Indeed, the estimate of passive rGE was
nonsignificant, suggesting that the association between chaos and
verbal ability may be mediated wholly by the shared environment.
These analyses support the conclusions of a previous report, which
suggested that chaos is associated with poor verbal performance in
3- to 4-year-olds but did not demonstrate environmental mediation
of the association (Petrill et al., 2004).

In this demonstration of our method, there is a possibility of

criterion contamination deriving from reliance on a single infor-
mant for both environmental and outcome variables. Despite ex-
tensive validation of the measures we employed, it could be that
informant error in reporting chaos in the home is correlated with
error or bias in administering tests of offspring’s verbal abilities.
Method variance due to correlated measurement errors would be
estimated as part of the main effect of the measured environment.
Given the overall lack of genetic correlation between chaos and
verbal ability, we consider it unlikely that heritable factors con-
tribute to this method variance to any significant degree. Hence,
our conclusion that the association between chaos and verbal
ability is not mediated by passive rGE remains unchanged.

We have presented the simplest possible version of this tech-

nique, employing single measures for the environment and out-
come. Further explorations of these data could be improved by
incorporating analyses of multiple phenotypic or environmental
measures, especially information from more than one informant.
The model can in principle be adapted for multivariate measures of
the outcome and/or environment by substituting vector and matrix
quantities for the scalar variables in the structural equations. As is
the case for the univariate model, the crucial issue is the size of the
moderating effects that are necessary to estimate the mediation
parameters accurately. We are currently developing extensions of
the method to address this issue in the multivariate case. In
addition, the method can be used to investigate moderating effects
of the family environment on shared and nonshared environment
influences, although the current article does not explore these
possibilities in detail.

These findings reaffirm that the classical twin study is a useful

experimental design for investigating gene– environment interplay.
Although children-of-twins studies and other extended family re-
search designs have been advocated as the best methods for
distinguishing genetically mediated and environmentally mediated
effects of the family environment (Purcell & Koenen, 2005;
Turkheimer et al., 2005), the classical twin design has the double
advantage that suitable datasets are both more readily available
and easier to collect. Phenotypic data are only required for the
twins themselves, rather than for multiple generations in the fam-
ily. This facilitates data collection, especially for phenotypes mea-
sured in adulthood. In addition, the overwhelming majority of
existing family studies have collected information only or predom-
inantly for one generation. The methods we present are also
adaptable to other family study designs: The authors consider that
a balance of different research designs with complementary merits
and assumptions will prove optimally informative (Rutter, Moffitt,
& Caspi, 2006).

313

SPECIAL SECTION: EFFECTS OF THE FAMILY ENVIRONMENT

Finally, we recognize the limitations of all family study designs

to resolve questions of genetic and environmental etiology. The
ultimate significance of G

⫻ E concerns the mechanistic interac-

tion between measured environments and specific genotypes, not
statistical interactions with “black box” variables representing
familial genetic risk (Rutter & Silberg, 2002). Elucidating the
interplay between genetic and environmental risk factors will
promote understanding of the causes of cognitive disability and
psychopathology and may be important for preventative efforts
(Jaffee & Price, 2007).

References

Akaike, H. (1974). A new look at statistical-model identification. IEEE

Transactions on Automatic Control, 19, 716 –723.

Armson, B. A., O’Connell, C., Persad, V., Joseph, K. S., Young, D. C., &

Baskett, T. F. (2006). Determinants of perinatal mortality and serious
neonatal morbidity in the second twin. Obstetrics & Gynecology, 108,
556 –564.

Asbury, K., Wachs, T. D., & Plomin, R. (2005). Environmental moderators

of genetic influence on verbal and nonverbal abilities in early childhood.
Intelligence, 33, 643– 661.

Bennett, A. J., Lesch, K. P., Heils, A., Long, J. C., Lorenz, J. G., Shoaf,

S. E., et al. (2002). Early experience and serotonin transporter gene
variation interact to influence primate CNS function. Molecular Psychi-
atry, 7, 118 –122.

Button, T. M., Scourfield, J., Martin, N., Purcell, S., & McGuffin, P.

(2005). Family dysfunction interacts with genes in the causation of
antisocial symptoms. Behavior Genetics, 35, 115–120.

Collins, W. A., Maccoby, E. E., Steinberg, L., Hetherington, E. M., &

Bornstein, M. H. (2000). Contemporary research on parenting. The case
for nature and nurture. American Psychologist, 55, 218 –232.

Dick, D. M., Rose, R. J., Viken, R. J., Kaprio, J., & Koskenvuo, M. (2001).

Exploring gene– environment interactions: Socioregional moderation of
alcohol use. Journal of Abnormal Psychology, 110, 625– 632.

Duncan, G. J., & Brooks-Gunn, J. (1997). Consequences of growing up

poor. New York: Russell Sage.

Eaves, L., Silberg, J., & Erkanli, A. (2003). Resolving multiple epigenetic

pathways to adolescent depression. Journal of Child Psychology and
Psychiatry, 44, 1006 –1014.

Etheredge, A. J., Christensen, K., Del Junco, D., Murray, J. C., & Mitchell,

L. E. (2005). Evaluation of two methods for assessing gene–
environment interactions using data from the Danish case-control study
of facial clefts. Birth Defects Research Part A: Clinical and Molecular
Teratology, 73, 541–546.

Evans, G. W. (2004). The environment of childhood poverty. American

Psychologist, 59, 77–92.

Evans, G. W. (2006). Child development and the physical environment.

Annual Review of Psychology, 57, 423– 451.

Fenson, L., Dale, P. S., Reznick, J. S., Bates, E., Thal, D. J., & Pethick, S. J.

(1994). Variability in early communicative development. Monographs
of the Society for Research in Child Development, 59, 1–173.

Fischbein, S. (1980). IQ and social class. Intelligence, 4, 51– 63.
Guo, G., & Stearns, E. (2002). The social influences on the realization of

genetic potential for intellectual development. Social Forces, 80, 881–
910.

Harden, K. P., Turkheimer, E., & Loehlin, J. C. (2007). Genotype by

environment interaction in adolescents’ cognitive aptitude. Behavior
Genetics, 37, 273–283.

Jaffee, S. R., & Price, T. S. (2007). Gene– environment correlations: A

review of the evidence and implications for prevention of mental illness.
Molecular Psychiatry, 12, 1432–1442.

Kendler, K. S., & Eaves, L. J. (1986). Models for the joint effect of

genotype and environment on liability to psychiatric illness. American
Journal of Psychiatry, 143, 279 –289.

Kremen, W. S., Jacobson, K. C., Xian, H., Eisen, S. A., Waterman, B.,

Toomey, R., et al. (2005). Heritability of word recognition in middle-
aged men varies as a function of parental education. Behavior Genetics,
35, 417– 433.

Leventhal, T., & Brooks-Gunn, J. (2000). The neighborhoods they live in:

The effects of neighborhood residence on child and adolescent out-
comes. Psychological Bulletin, 126, 309 –337.

Liu, X., Fallin, M. D., & Kao, W. H. (2004). Genetic dissection methods:

Designs used for tests of gene– environment interaction. Current Opin-
ion in Genetics and Development, 14, 241–245.

Matheny, A. P., Wachs, T. D., Ludwig, J. L., & Phillips, K. (1995).

Bringing order out of chaos: Psychometric characteristics of the Confu-
sion, Hubbub, and Order Scale. Journal of Applied Developmental
Psychology, 16, 429 – 444.

Nagoshi, C. T., & Johnson, R. C. (2005). Socioeconomic status does not

moderate the familiality of cognitive abilities in the Hawaii Family
Study of Cognition. Journal of Biosocial Science, 37, 773–781.

Neale, M. C., Boker, S. M., Xie, G., & Maes, H. H. (1999). Mx: Statistical

modeling (Version 1.64) [Computer software]. Available from Virginia
Commonwealth University: http://www.vcu.edu/mx/

Neiss, M., Rowe, D. C., & Rodgers, J. L. (2002). Does education mediate

the relationship between IQ and age of first birth? A behavioural genetic
analysis. Journal of Biosocial Science, 34, 259 –275.

Petrill, S. A., Pike, A., Price, T., & Plomin, R. (2004). Chaos in the home

and socioeconomic status are associated with cognitive development in
early childhood: Environmental mediators identified in a genetic design.
Intelligence, 32, 445– 460.

Pike, A., Iervolino, A. C., Eley, T. C., Price, T. S., & Plomin, R. (2006).

Environmental risk and young children’s cognitive and behavioral de-
velopment. International Journal of Behavioral Development, 30, 55–
66.

Plomin, R. (1986). Development, genetics, and personality. Hillsdale, NJ:

Erlbaum.

Plomin, R., DeFries, J. C., & Loehlin, J. C. (1977). Genotype-environment

interaction and correlation in the analysis of human behavior. Psycho-
logical Bulletin, 84, 309 –322.

Price, T. S., Freeman, B., Craig, I., Petrill, S. A., Ebersole, L., & Plomin,

R. (2000). Infant zygosity can be assigned by parental report question-
naire data. Twin Research, 3, 129 –133.

Purcell, S. (2002). Variance components models for gene– environment

interaction in twin analysis. Twin Research, 5, 554 –571.

Purcell, S., & Koenen, K. C. (2005). Environmental mediation and the twin

design. Behavior Genetics, 35, 491– 498.

Rathouz, P. J., Van Hulle, C. A., Rodgers, J. L., & Lahey, B. B. (2007).

Specification, testing, and interpretation of gene-by-measured-
environment interaction models in the presence of gene– environment
correlation. Unpublished manuscript.

Rosenthal, D. (1963). The Genain quadruplets. New York: Basic Books.
Rowe, D. C., Jacobson, K. C., & Van den Oord, E. J. C. G. (1999). Genetic

and environmental influences on vocabulary IQ: Parental education level
as moderator. Child Development, 70, 1151–1162.

Rutter, M. (2003). Genetic influences on risk and protection: Implications

for understanding resilience. In S. S. Luthar (Ed.), Resilience and vul-
nerability: Adaptation in the context of childhood adversities (pp. 489 –
509). New York: Cambridge University Press.

Rutter, M., Moffitt, T. E., & Caspi, A. (2006). Gene– environment interplay

and psychopathology: Multiple varieties but real effects. Journal of
Child Psychology and Psychiatry, 47, 226 –261.

Rutter, M., O’Connor, T. G., & the English and Romanian Adoptees Study

Team. (2004). Are there biological programming effects for psycholog-
ical development? Findings from a study of Romanian adoptees. Devel-
opmental Psychology, 40, 81–94.

314

PRICE AND JAFFEE

Rutter, M., & Silberg, J. (2002). Gene– environment interplay in relation to

emotional and behavioral disturbance. Annual Review of Psychology, 53,
463– 490.

Scarr, S., & McCartney, K. (1983). How people make their own environ-

ments: A theory of genotype

3 environment effects. Child Develop-

ment, 54, 424 – 435.

Scarr-Salapatek, S. (1971, December 24). Race, social class, and IQ.

Science, 174, 1285–1295.

Spiegelhalter, D., Thomas, A., Best, N., & Lunn, D. (2003). WinBUGS

user manual Version 1.4. Retrieved from http://www.mrc-bsu.cam
.ac.uk/bugs/

Spinath, F. M., Price, T. S., Dale, P. S., & Plomin, R. (2004). The genetic

and environmental origins of language disability and ability. Child
Development, 75, 445– 454.

Tambs, K., Sundet, J. M., Magnus, P., & Berg, K. (1989). Genetic and

environmental contributions to the covariance between occupational
status, educational attainment, and IQ: A study of twins. Behavior
Genetics, 19, 209 –222.

Trouton, A., Spinath, F. M., & Plomin, R. (2002). Twins Early Develop-

ment Study (TEDS): A multivariate, longitudinal genetic investigation
of language, cognition and behavior problems in childhood. Twin Re-
search, 5, 444 – 448.

Turkheimer, E., D’Onofrio, B. M., Maes, H. H., & Eaves, L. J. (2005).

Analysis and interpretation of twin studies including measures of the
shared environment. Child Development, 76, 1217–1233.

Turkheimer, E., Haley, A., Waldron, M., D’Onofrio, B., & Gottesman, I. I.

(2003). Socioeconomic status modifies heritability of IQ in young chil-
dren. Psychological Science, 14, 623– 628.

Van den Oord, E. J. C. G., & Rowe, D. C. (1997). An examination of

genotype-environment interactions for academic achievement in an U.S.
national longitudinal survey. Intelligence, 25, 205–228.

Received October 31, 2006

Revision received August 13, 2007

Accepted October 19, 2007

䡲

Members of Underrepresented Groups:

Reviewers for Journal Manuscripts Wanted

If you are interested in reviewing manuscripts for APA journals, the APA Publications and
Communications Board would like to invite your participation. Manuscript reviewers are vital to the
publications process. As a reviewer, you would gain valuable experience in publishing. The P&C
Board is particularly interested in encouraging members of underrepresented groups to participate
more in this process.

If you are interested in reviewing manuscripts, please write to the address below. Please note the
following important points:

• To be selected as a reviewer, you must have published articles in peer-reviewed journals. The

experience of publishing provides a reviewer with the basis for preparing a thorough, objective
review.

• To be selected, it is critical to be a regular reader of the five to six empirical journals that are most

central to the area or journal for which you would like to review. Current knowledge of recently
published research provides a reviewer with the knowledge base to evaluate a new submission
within the context of existing research.

• To select the appropriate reviewers for each manuscript, the editor needs detailed information.

Please include with your letter your vita. In the letter, please identify which APA journal(s) you
are interested in, and describe your area of expertise. Be as specific as possible. For example,
“social psychology” is not sufficient—you would need to specify “social cognition” or “attitude
change” as well.

• Reviewing a manuscript takes time (1– 4 hours per manuscript reviewed). If you are selected to

review a manuscript, be prepared to invest the necessary time to evaluate the manuscript
thoroughly.

Write to Journals Office, American Psychological Association, 750 First Street, NE, Washington,
DC 20002-4242.

315

SPECIAL SECTION: EFFECTS OF THE FAMILY ENVIRONMENT