• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Behav Genet. Author manuscript; available in PMC Mar 1, 2011.
Published in final edited form as:
PMCID: PMC2830280

Brain electrophysiological endophenotypes for externalizing psychopathology: A multivariate approach


Abnormalities in electrophysiological measures of stimulus-evoked brain activity (including the P3 event-related potential (ERP) and its associated delta and theta time-frequency (TF) components), and intrinsic, resting state brain activity (including EEG in the beta frequency band) have each been associated with biological vulnerability to a variety of externalizing (EXT) spectrum disorders, such as substance use disorders, conduct disorder, and antisocial behavior. While each of these individual measures has shown promise as an endophenotype for one or more aspects of EXT, we proposed that the power to identify EXT-related genes may be enhanced by using these measures collectively. Thus, we sought to explore a multivariate approach to identifying electrophysiological endophenotypes related to EXT, using measures identified in the literature as promising individual endophenotypes for EXT. Using data from our large twin sample (634 MZ and 335 DZ, male and female same-sex pairs), and fitting multivariate biometric Cholesky models, we found that these measures 1) were heritable, 2) showed significant phenotypic and genetic correlation with a general vulnerability to EXT (which is itself highly heritable), 3) showed modest phenotypic and genetic correlation with each other, and 4) were sensitive to genetic effects that differed as a function of gender. These relationships suggest that these endophenotypes are likely tapping into neurophysiological processes and genes that are both common across them and unique to each – all of which are relevant to a biological vulnerability to EXT psychopathology.

Keywords: endophenotype, antisocial behavior, substance use, P3, beta, time-frequency


Decades of twin, family, and adoption research have established that genetic influences underlie substance use disorders (SUDs; e.g alcohol, nicotine, and drug use/dependence) (Goldman, Oroszi, & Ducci, 2005). Elucidating the genetic etiology of SUDs would help to inform intervention and treatment strategies, greatly easing the public health impact of these disorders (Merikangas & Risch, 2003). SUDs, however, rarely occur in isolation; disorders that commonly occur comorbidly with SUDs include a spectrum of disorders characterized by behavioral disinhibition, e.g. conduct disorder and antisocial behavior. Large-scale epidemiological studies with twins have shown that the common comorbidity among these disorders can be accounted for by an underlying externalizing (EXT) factor that is highly heritable (Kendler, Prescott, Myers, & Neale, 2003; Krueger et al., 2002; Young, Stallings, Corley, Krauter, & Hewitt, 2000), and that the source of the comorbidity in these externalizing disorders can be largely attributed to common genetic factors (Hicks, Krueger, Iacono, McGue, & Patrick, 2004; Kendler et al., 2003). The high heritability and common genetic influence across these disorders suggest that a behavioral/clinical phenotype based on this EXT factor may be a better target in the search for genes underlying SUDs and related behavioral disinhibitory disorders in this spectrum than any of the disorders individually.

Efforts to discover genes underlying EXT spectrum disorders have reported significant results implicating genes for alcohol dependence and related phenotypes (e.g. ADH, GABRA2, GABRR1, GABRR2; Bice et al., 2008; Covault, Gelernter, Jensen, Anton, & Kranzler, 2008; Dick et al., 2006; Edenberg et al., 2004; Schuckit, 2009; Wall, Shea, Luczak, Cook, & Carr, 2005; Xuei et al., 2009), other substance use disorders (Agrawal et al., 2006; Hopfer et al., 2007; Stallings et al., 2003), and behavioral disorders (Amin et al., 2009; Burt & Mikolajewski, 2008; McGough et al., 2008; Rommelse et al., 2008). These findings, however, are tempered by the fact that psychopathological disorders are etiologically complex and multifactorial in origin, reflecting the combined influence and interactions of genetic, epigenetic, stochastic, and nongenetic factors (Cowan, Kopnisky, & Hyman, 2002; Gottesman & Gould, 2003; Schork, Greenwood, & Braff, 2007). This complexity in etiology, coupled with an incomplete understanding of how best to identify and diagnose psychological disorders (e.g. Krueger, Watson, & Barlow, 2005), has hampered the search for psychopathology-related genes. This genetic and phenotypic complexity belies the need for phenotypes that are genetically simpler than clinical diagnostic categories, and that are less reliant on manifest behavior for diagnosing affected individuals (Gottesman & Gould, 2003; Iacono, 1998). Such traits, known as endophenotypes, are putatively less complex correlates of the disorder that, because they tap the neurobiological basis of gene effects, are closer to gene action, and can be powerful aids in discovering the disorder’s genetic etiology.

Psychophysiological measures, particularly electroencephalography (EEG) based measures of neural activity, represent, in part, endogenous, biological attributes whose measurement relies neither on clinical diagnosis nor manifest symptoms of a disorder. EEG-based indices of both stimulus-evoked, task-related brain activity (namely, the P3 event-related potential [ERP] and its associated time-frequency [TF] components) and intrinsic, resting state activity (specifically, activity in the beta frequency band) have been associated with EXT spectrum disorders (e.g. Iacono, Carlson, Malone, & McGue, 2002; Porjesz et al., 2005), and have met criteria (e.g. Frederick & Iacono, 2006; Iacono, 1998) proposed to be necessary for status as an endophenotype: 1) heritability (van Beijsterveldt, Molenaar, de Geus, & Boomsma, 1996; van Beijsterveldt & van Baal, 2002; Yoon, Iacono, Malone, & McGue, 2006), 2) association with multiple disorders in the EXT spectrum (Gilmore, Malone, Bernat, & Iacono, in press; Iacono et al., 2002; Porjesz et al., 2005; Rangaswamy et al., 2002), 3) presence in unaffected relatives of those with a disorder (Begleiter, Porjesz, Bihari, & Kissin, 1984; Hill, Muka, Steinhauer, & Locke, 1995; Rangaswamy et al., 2004), and 4) in currently unaffected individuals, association with later development of an EXT spectrum disorder (Berman, Whipple, Fitch, & Noble, 1993; Carlson, Iacono, & McGue, 2004; Hill, Steinhauer, Locke-Wellman, & Ulrich, 2009; Iacono et al., 2002).

P3 Event-Related Potential

The most robust association has been found between externalizing disorders and the P3 ERP. P3 is elicited by detection of a stimulus that deviates from the ongoing context of the task, and it is theorized to reflect context updating in working memory and the generation of subjective expectancies related to the probability of target occurrence (Donchin & Coles, 1988; Duncan-Johnson & Donchin, 1977; Soltani & Knight, 2000). Reduced peak amplitude of the P3 has long been associated with alcoholism and familial risk of developing alcoholism (Begleiter et al., 1984; Hill, 2004; Hill et al., 1995; Polich, Pollock, & Bloom, 1994). Recent work, however, has expanded P3 amplitude reduction’s (P3-AR’s) association to other EXT spectrum disorders (e.g. Iacono et al., 2002), as well as to the underlying EXT factor (Patrick et al., 2006) – an association that is accounted for by shared genetic effects (Hicks et al., 2007).

These findings, however, have mainly involved only male subjects. Studies investigating the relationship between P3 and SUDs/EXT in females are fewer and the results equivocal. Some studies find reduced P3 in female alcoholics (Hill & Steinhauer, 1993; Prabhu et al., 2001) and females at high-risk of developing alcoholism (Porjesz et al., 1998; Van Der Stelt, 1999), while other studies find no significant P3-AR in female alcoholics (Hill, Locke, & Steinhauer, 1999; Parsons, Sinha, & Williams, 1990). Across genders, when P3-AR is present, it tends to be enhanced more for males than females (e.g. Iacono & McGue, 2006; Porjesz et al., 1998; Yoon et al., 2006). Thus, evidence suggests a differential association between P3 and SUDs/EXT as a function of gender. Nonetheless, P3 has been shown to be strongly heritable (van Beijsterveldt & van Baal, 2002; Yoon et al., 2006), and it has been linked to genes putatively underlying EXT spectrum disorders, particularly alcoholism. Williams et al. (1999) found that reduced P3 amplitude and an alcoholism diagnosis were jointly linked to a region on chromosome 4 near the alcohol dehydrogenase gene (ADH3), and Hill et al. (1998) found an association between lower P3 amplitude and presence of the dopamine receptor A1 allele in children from alcoholic families.

Event-Related Time-Frequency Components

While these findings support P3-AR’s potential as an endophenotype for EXT psychopathology, at least for males, there is evidence that the P3 ERP is composed of multiple neurophysiological processes (e.g. Dien, Spencer, & Donchin, 2003; Mantini, Corbetta, Perrucci, Romani, & Del Gratta, 2009). An emerging methodological approach that serves to represent the ERP in a more detailed way than standard measures (e.g. peak amplitude), and endeavors to delineate these multiple processes is time-frequency (TF) decomposition. For instance, TF decomposition has been used to show that ERP activity during the P3 can be characterized by TF components in the delta (0–3 Hz) and theta (3–7 Hz) frequency bands (Basar-Eroglu, Basar, Demiralp, & Schurmann, 1992; E. M. Bernat, Malone, Williams, Patrick, & Iacono, 2007; Jones et al., 2006; Yordanova, Devrim, Kolev, Ademoglu, & Demiralp, 2000), suggesting that TF representations may present a more detailed characterization of event-related brain activity.

Indeed, some recent efforts in the search for alcoholism-related endophenotypes have shifted focus away from time-domain P3 amplitude and toward TF measures of ERP activity. P3-related delta and theta power was shown to be reduced in adult alcoholics (Jones et al., 2006) and in high-risk adolescent and adult offspring of alcoholics (Kamarajan et al., 2006; Rangaswamy et al., 2007). Most recently, our group has extended these findings by utilizing a novel, data-driven, principal components analysis based TF analysis method (TF-PCA; E.M. Bernat, Williams, & Gehring, 2005) to reveal a more detailed representation of the multiple overlapping processes related to P3 activity (E. M. Bernat et al., 2007; Gilmore et al., in press). In Gilmore et al. (in press), TF-PCA revealed five time- and frequency-specific event-related delta and theta components, each of which was able to discriminate those subjects with one of six externalizing spectrum disorders from those with no disorder (based on reduced amplitudes in the diagnostic groups) as well as or better than could P3 amplitude. These results demonstrated that TF-PCA derived components may capture more externalizing-related variance in the ERP than is indexed by P3 peak amplitude alone, suggesting that using these TF components in combination with P3 measures may optimize the measurement of this endophenotype.

Further evidence for TF activity’s utility as an endophenotype comes from studies linking P3-related delta and theta activity to specific genes: CHRM2 (Jones et al., 2004; Porjesz & Rangaswamy, 2007), a cholinergic muscarinic receptor gene, and GRM8 (Chen et al., 2009), a metabotropic glutamate receptor gene, both on chromosome 7. Both CHRM2 and GRM8 have been associated with alcohol dependence (Chen et al., 2009; Luo et al., 2005; Wang et al., 2004), while CHRM2 has also been implicated in drug dependence (Luo et al., 2005; Wang et al., 2004), and, most recently, in externalizing disorders (Dick et al., 2008).

Resting State EEG Beta

In addition to these indices of evoked brain activity, measures of intrinsic, “resting state” brain activity as measured using EEG have also shown associations with alcoholism and other EXT-spectrum disorders. The spontaneous neural activity of the resting state is composed of dynamic combinations of delta (0–3 Hz), theta (3–7 Hz), alpha (7–12 Hz), beta (12–30 Hz), and gamma (>30 Hz) band frequencies (Buzsáki, 2006). Activity in the beta frequency band, a putative index of the level of cortical arousal, has consistently shown associations with EXT-spectrum disorders. These associations may be affected, however, by the nature of the specific disorder and by gender. Increased beta power has been found in those with alcoholism (Bauer, 1994, 2001; Costa & Bauer, 1997; Rangaswamy et al., 2002), cocaine dependence (Costa & Bauer, 1997), and those at high risk for developing alcoholism (Pollock, Earleywine, & Gabrielli, 1995; Rangaswamy et al., 2004), although the specific beta frequencies involved and their scalp topographies have been shown to differ as a function of gender (Rangaswamy et al., 2002; 2004). Beta power has been shown to be reduced, however, in adolescents with ADHD (Hobbs, Clarke, Barry, McCarthy, & Selikowitz, 2007; Mann, Lubar, Zimmerman, Miller, & Muenchen, 1992), ADHD with comorbid oppositional defiant disorder (Clarke, Barry, McCarthy, & Selikowitz, 2002), and young adults exhibiting impulsive and antisocial personality characteristics (Houston & Stanford, 2005). Further, while boys’ beta power is reduced more than that of girls relative to controls (Clarke, Barry, McCarthy, & Selikowitz, 2001), girls with ADHD show reduced beta more consistently (i.e. with less variability) than boys (Clarke et al., 2003).

Supporting its role as a potential endophenotype for EXT, resting state beta activity has been found to be highly heritable (McGuire, Katsanis, Iacono, & McGue, 1998; van Beijsterveldt et al., 1996), and has been linked to the GABRA2 gene, a GABA(A) neurotransmitter receptor gene on chromosome 4p (Edenberg et al., 2004; Porjesz et al., 2002), which itself has also been associated with a number of disorders in the EXT spectrum: alcohol dependence (Edenberg et al., 2004; Enoch, 2008; Soyka et al., 2008), drug dependence (Agrawal et al., 2006; Dick et al., 2006), and conduct disorder (Dick et al., 2006). Thus, there is evidence that the GABRA2 gene may affect both beta frequency EEG power and a predisposition to develop various disorders in the EXT spectrum.

Multivariate Approach to Identifying Endophenotypes for Externalizing

Taken together, these results demonstrate that multiple stimulus-evoked and resting state electrophysiological measures show promise as endophenotypes for EXT psychopathology. Each of these measures has been shown to be heritable (e.g. van Beijsterveldt et al., 1996; van Beijsterveldt & van Baal, 2002; Yoon et al., 2006), to be associated with genetic risk for one or more EXT disorders (e.g. Begleiter et al., 1984; Hill et al., 1995; Rangaswamy et al., 2004), and with specific, although different, genes associated with these disorders (e.g. Chen et al., 2009; Edenberg et al., 2004; Porjesz & Rangaswamy, 2007; Williams et al., 1999). Thus, each of these brain electrophysiological measures, which putatively tap the neural bases of genetic risk, has demonstrated significant utility in the search for genes related to EXT psychopathology.

We propose that the power to identify EXT-related genes may be enhanced by using these measures in concert, effectively creating a multivariate endophenotype (e.g. Arya, Duggirala, Williams, Almasy, & Blangero, 2001; Frederick & Iacono, 2006; Iacono, Carlson, & Malone, 2000; Price et al., 2006), by capitalizing on the covariance among the multiple measures and EXT. The goal of this report was to explore a multivariate approach to identifying electrophysiological endophenotypes related to EXT psychopathology, using measures identified in the literature as promising individual endophenotypes for EXT. We specifically sought to address the following aims: 1) assess the magnitude of the phenotypic and genetic associations between each endophenotype and EXT, 2) determine if there is significant shared genetic variance among the multiple endophenotypes and EXT, and 3) determine if any of the endophenotypes share unique genetic variance with EXT (i.e. variance which is not also shared with the other endophenotypes). Also, because there is evidence for differential relationships between the brain activity measures and EXT disorders, we assessed potential gender effects. Finding significant relationships among the electrophysiological endophenotypes and EXT would suggest that, used in combination, they may be more optimal targets than any of the individual measures for studies investigating the genetic etiology of externalizing spectrum disorders.



Participants were adolescent male and female twins who participated in the Minnesota Twin Family Study (MTFS), a longitudinal and epidemiological study investigating the development of substance use disorders and related psychopathology. A comprehensive description of the MTFS is found in Iacono & McGue (2002). Two age cohorts compose the MTFS sample: a younger cohort initially recruited the year the twins turned 11 years old, and an older cohort recruited the year the twins turned 17 years old. Participants were identified through public birth records of twin pairs born in the years 1972–1978 in the case of the older male cohort, 1975–1979 for the older female cohort, 1978–1982 for the younger male cohort, and 1981–1984 for the younger female cohort. Zygosity (MZ=monozygotic; DZ=dizygotic) was determined by the agreement of three estimates: (1) a standard zygosity questionnaire completed by parents, (2) evaluation by MTFS staff of physical similarity based on eye, hair, face, and ear characteristics, and (3) similarity on fingerprint ridge count, cephalic index, and ponderal index. A serologic analysis was conducted if the 3 estimates did not agree. Twins return at roughly 3-year intervals for follow-up assessments.

For the present report, data were collected when members of both cohorts were at a similar age, specifically, the intake assessment for the 17-year-old cohorts (males: M=17.5, SD=0.4; females: M=17.5, SD=0.5) and the second follow-up assessment for the 11-year old cohorts (males: M=18.0, SD=0.7; females: M=18.3, SD=0.7). All study protocols and data collection methods remained consistent across the intake and second follow-up assessments. The current report included those participants for whom data on the psychophysiological measures of interest (see below for details) were available: 303 MZ and 153 DZ male twin pairs, 331 MZ and 182 DZ female twin pairs. Consistent with the demographics of Minnesota during the years the twins were born, almost all are Caucasian (98%). Written informed assent or consent (from the participant or from the parent of a minor child) was obtained as appropriate.

Diagnostic Assessment

For participants in each of the two cohorts, lifetime presence of DSM-III-R (American Psychiatric Association, 1987) disorders, the diagnostic system in place at the time the MTFS was initiated, was assessed via in-person, structured interviews by trained clinical interviewers. Members of each twin pair were assessed concurrently by separate interviewers. Symptoms of nicotine dependence, alcohol dependence, and illicit drug dependence (which included amphetamines, cannabis, cocaine, inhalants, opioids, PCP, and sedatives) were assessed using an expanded version of the Substance Abuse Module of the Composite International Diagnostic Interview (Robins, Babor, & Cottler, 1987). The substance for which the participant reported the maximal number of symptoms was used for their drug dependence variable. An interview designed by MTFS staff (Holdcraft, Iacono, & McGue, 1998) was used to assess symptoms of conduct disorder and adult antisocial behavior (i.e., the adult criteria for Antisocial Personality Disorder). Mothers reported on the presence of conduct disorder and substance use disorder symptoms in her twin children through interviews using the parent version of the Diagnostic Interview for Children and Adolescents (Reich, 2000). A symptom was considered present if reported by either the mother or twin. Symptoms were assigned on the basis of a consensus, “best-estimate” approach (Leckman, Scholomskas, Thompson, Belanger, & Weisman, 1982) combining mother and twin interview data.

Our measure of Externalizing consisted of the first principal component (which accounted for 59% of the variance) of the log-transformed symptom counts of conduct disorder, adult antisocial behavior, alcohol abuse/dependence, nicotine dependence, and illicit drug abuse/dependence (c.f. Hicks et al., 2007; Patrick et al., 2006).

Psychophysiological Assessment

Resting EEG

Five minutes of resting EEG were collected while participants were instructed to relax with their eyes closed. The participants were seated in a comfortable chair in a dark room while listening to 55 dB white noise through headphones.

ERP task

A rotated-heads visual oddball task (Begleiter et al., 1984) was used. Subjects viewed 240 stimuli consisting of either an oval (two-thirds of trials – “standards”) or a superior view of a stylized head (one-third of trials – “targets”), in which a nose and one ear were depicted on the oval. Subjects were required to respond to target trials by pressing a button on either the left or right armrest of their chair, corresponding to the side of the head on which the ear appeared. On half the target trials the nose pointed up (such that the left ear appeared on the left side of the screen; an easy discrimination), while on the other half of target trials the head was rotated 180° so that the nose pointed down (left ear appeared on the right side of the screen; a hard discrimination). Stimulus duration was 98 ms, and the inter-trial interval, during which subjects fixated on a dot in the center of the screen, varied randomly between 1 and 2 seconds.

Recording procedure

A Grass model 12A Neurodata Acquisition System recorded the EEG. Impedances were kept below 5 kΩ. Data were recorded at a sampling rate of 256 Hz during the ERP task and 128 Hz for the resting EEG, and filtered from 0.01 – 30 Hz (half-amplitude, 6 dB/octave rolloff) for both tasks. ERP task data were recorded from three parietal electrodes: on the midline at Pz, and over left and right hemispheres at P3 and P4, respectively, referenced to linked earlobes. Resting EEG was recorded using a bipolar montage from parietal-occipital sites, O1-P7 and O2-P8, over left and right hemispheres, respectively, and monopolarly, referenced to linked earlobes, from central-midline electrode Cz. Additionally, during both tasks, electrooculographic (EOG; used to screen for eye movement artifacts) data were recorded using a pair of biopotential electrodes arranged in a transverse montage, one electrode placed superior to the eye and the other at the outer canthus.

Data processing

Resting EEG

The resting EEG data were divided into 2 s half-overlapping epochs (i.e. the first second of each epoch overlapped with the last second of the preceding epoch). Epochs in which EOG amplitude >60µV were eliminated from further processing. Fast Fourier Transform (FFT) was then applied to each artifact-free epoch using a Tukey-Hanning window, resulting in power spectra with 0.5 Hz resolution. The mean absolute power within the Beta frequency range (13–30 Hz) was then calculated for each of the three recording sites. Because the distributions of power within these bands were skewed, the data were log transformed.


For the ERP data, blinks and other ocular artifacts were corrected using the method of Gratton, Coles, and Donchin (1983). Trials with activity >100 µV were excluded from further processing. ERP epochs, consisting of 2 s of data, including a 500 ms prestimulus baseline, were averaged over the easy and hard target conditions at each of the three electrode sites. P3 amplitude was defined as the point between 280 and 600 ms at which amplitude of the average waveform was maximal.

Time-frequency PCA decomposition of ERPs

The time-frequency PCA (TF-PCA) method has been detailed in previous reports (E. M. Bernat et al., 2007; E.M. Bernat et al., 2005; Gilmore et al., in press). Here, the primary features are outlined. Averaged target ERP data from each of the three electrodes, for each participant, were subjected to TF decomposition using the Cohen’s class reduced interference distribution (RID) transform. This TF transform resulted in averaged TF surfaces, which are representations of the overall energy of the event-related activity, within a frequency range of 0 – 7.5 Hz (through the upper limit of theta) and a time range from stimulus onset (0 ms) to 1000 ms post-stimulus. PCA, using the covariance matrix and varimax rotation, was then performed on these average TF surfaces to decompose the surfaces into separate TF-PCA components. The resulting TF surfaces represented each TF-PCA component’s matrix of rotated component loadings for each TF point. Five components were retained, based upon inspection of the scree plot of singular values, representing the relative variance accounted for by each component, for a break, or elbow. In a final step, each participant’s average TF surface was weighted using the extracted TF-PCA components. To weight the original TF data, each time-frequency point is multiplied by the corresponding point in the matrix of rotated loadings for each component. This produces weighted data surfaces, for each participant for each TF-PCA component, whose data points represent energy in units weighted by the component loadings. For subsequent analyses, component scores representing the peak energy on the weighted TF-PCA surface were used.

Data Analysis

Our main approach consisted in fitting multivariate biometric models using a Cholesky parameterization. This permitted us to determine whether adding candidate endophenotypes to our model would account for significant unique variance in EXT. Because our approach to decomposing time-frequency surfaces differs from that used in previous studies relating TF energy to externalizing psychopathology, yielding relevant dimensions not necessarily readily evident to the naked eye (Gilmore et al., in press), it was not immediately apparent which features would be most relevant. We therefore conducted principal component analysis of the 15 TF-PCA measures (5 components, 3 electrodes) and EXT to identify TF-PCA components that are related to (load with) EXT (c.f. Patrick et al., 2006). We used the components that loaded with EXT in subsequent biometric models. PCA and all data manipulations were conducted using R (version 2.8.1; R Foundation for Statistical Computing).

We freely estimated means and covariances of P3 amplitude, beta power, the time-frequency components that emerged from the above analysis, and EXT without imposing any latent structure on the data (i.e., without the additive genetic (A), shared environmental (C), and non-shared environmental (E) latent factors). This served as a baseline model. We subsequently fit an ACE model using a Cholesky parameterization. Based on the strength of prior evidence regarding associations between the candidate endophenotypes and genetic risk for EXT, we entered them in the following order: P3 amplitude, time-frequency components, beta power, and EXT (see Figure 1 for an illustration of the Cholesky model). We tested several submodels by fixing parameters at 0 in order to conduct significance tests of the importance of different parameters. Specifically, we constrained to 0 paths from EXT to the first three latent A factors to determine whether genetic influences on the endophenotypes account for significant genetic variance in EXT. We subsequently assessed whether genetic factors beyond A1 (which represent genetic variance in externalizing shared with all the endophenotypes) would be significant, and particularly whether there would be gender differences in this regard. We used several fit statistics to evaluate model fit: likelihood-ratio tests, Akaike’s Information Criterion (AIC; Akaike, 1974) and Draper’s Information Criterion (DIC; Draper, 1995).

Figure 1
Path diagram for the AE Cholesky model. Standardized path coefficients are shown for both males and females (males / females); significant paths (p≤.01) are printed in bold. All paths should be squared to estimate the proportion of variance accounted ...


Before analyzing the data, we examined them for outliers using an R package (Fox, 2009) that implements the regression diagnostic procedures of Cook and Weisberg (1982). We examined the bivariate distributions of twin scores, and deleted pairs producing significant residuals by Bonferroni-corrected p value. We also deleted pairs producing Cook's distance scores above the threshold suggested by Cook and Weisberg. This resulted in removing 20 pairs (2.1% of the total sample).

Stability of Time-Frequency Components

For our TF-PCA components, we first wanted to confirm that the decompositions were stable across our different cohorts of subjects (i.e. across males and females in the 11 and 17 year old cohorts; see Figure 2 which illustrates the ERP and the 5 TF components extracted from the TF-PCA decomposition). We assessed stability of the PCA solution across the independent cohorts of subjects by conducting the TF-PCA decomposition separately for each, and then computing Tucker's congruence coefficients (which measure the similarity of two factorial configurations by ascertaining the degree to which the magnitude of loadings on the two factors have a similar pattern; values in excess of .90 indicate high congruence). Congruence coefficients between matching components were uniformly high, ranging from .93 to .99 across age cohorts and from .93 to .95 across gender cohorts, indicating that the TF-PCA decompositions had highly similar patterns.

Figure 2
Results of TF-PCA analyses. Grand-averaged ERP and TF (Avg) plots are presented at the top. ERP (from 0 ms – 1000 ms) is from electrode Pz, grand averaged over all participants; P3 is the peak in the waveform occurring around 450 ms. The five ...

TF Components Most Relevant for EXT

To determine which time-frequency components to include in our Cholesky model, we conducted PCA on the 15 measures (5 TF-PCA components, 3 electrodes) plus EXT, separately for each gender. The raw data were centered about the variable means and scaled by their respective standard deviations, equivalent to PCA on the correlation matrix. Although most of the variation in the data was due to the 15 time-frequency measures, the second (unrotated) component in both genders included a loading for the EXT composite in addition to certain TF measures, while the third component loaded almost entirely on the EXT composite. The pattern was clear: the TF measures and EXT each loaded on separate components, and one component included loadings for both, thus capturing covariance between the two domains. TF-PC1 (with peak energy centered around 2.5 Hz and 330 ms; see Figure 2) and TF-PC4 (with peak energy also centered around 2.5 Hz, but occurring slightly later in time, around 400 ms; see Figure 2) both loaded with EXT, with very similar loading patterns and magnitude of loadings for the two genders, thus indicating that these were the TF components most relevant to externalizing. In addition to loading with EXT, these components were highly correlated, with correlations between component scores for the different electrodes ranging from .76 to .91. Thus, to simplify our Cholesky model and avoid potential issues with collinearity, we elected to include in our model only TF-PC1, which loaded most strongly on the combined component in both genders.

Endophenotypic Measures

For subsequent analyses, we created composite endophenotype measures. For those measures recorded from one primary scalp area (P3 amplitude and TF-PC1 peak energy, which were recorded from three closely spaced parietal electrodes) we averaged across the three electrode sites. For beta, we created a composite consisting of the first principal component of the covariance matrix of the three electrodes. Loadings for the bipolar occipital leads were approximately twice as great as the loading for the monopolar Cz recording. The scree plot clearly indicated that one component, which accounted for 78% of the variance, was warranted for beta.

Table 1 gives descriptive statistics for the endophenotypes and EXT. As the table indicates, MZ and DZ twins had similar mean levels and variances. Because beta and EXT are principal component scores, their values can be negative. There were small but consistent mean differences between male and female participants, with higher levels of the endophenotypes and lower levels of EXT in females. For instance, with an overall standard deviation of 0.92 on the EXT component, the difference between males and females corresponded to an effect size of d = .51 (not shown in Table 1, which is broken down by zygosity, but offered for illustrative purposes). In addition, although our analysis used the first principal component of the five disorders’ symptom counts as our EXT measure, to insure that we had meaningful variance in EXT in this population-based sample, we determined the proportion of participants with three or more symptoms of each of the individual disorders making up the EXT score. This corresponds to those who are above the symptom threshold required to qualify for a diagnosis. Results showed that, overall, 24% of the sample had three or more symptoms of at least one disorder.

Table 1
Descriptive statistics (Mean and SD) for each phenotype.

The phenotypic correlations among the variables are presented in Table 2. Lack of independence due to the fact that the sample consisted of twin pairs was accounted for in assessing the significance of correlations. For males, all the correlations were significant. For females, all but the P3-EXT correlation were significant. As might be expected from an epidemiological sample, the correlations between the endophenotypes and EXT were modest. Therefore, to further highlight their importance and to link our findings to the literature which often examines clinical samples or offspring of clinical samples compared to healthy controls, we conducted a quintile split on each of the endophenotypes, separately for males and females, and compared EXT scores between those high (top 20%) and low (bottom 20%) on each of the endophenotypes. Differences between groups corresponded to effect sizes (Cohen’s d; the mean difference between EXT scores) in the small to moderate range (P3: males d=.31, females d=.20; TF-PC1: males d=.34, females d=.29; Beta: males d=.37, females d=.18). These effect sizes, considered together with the modest phenotypic correlations, suggest that the correlations may underestimate the endophenotypes’ abilities to differentiate between those high and low in EXT symptoms.

Table 2
Phenotypic correlations among the phenotypes.

The intraclass correlations are more similar than different for the two genders, and indicate substantial genetic influence on all composite measures (Table 3); MZ correlations were significantly greater than DZ correlations in all cases (all z-statistics > 2.80, p-values < .01). Although the DZ correlation for EXT was substantial for both genders, suggesting some influence of shared environment, it was not significantly greater than half the MZ correlation (p > .10).

Table 3
Intraclass correlations for each phenotype

Cholesky Decomposition Model

For the biometric analyses, the Cholesky model provided improved fit relative to the unstructured base model (which estimated only the phenotype means and covariances) by both AIC and DIC, indicating that modeling the phenotypes in terms of latent ACE factors was reasonable, and in particular that the Cholesky version was adequate (results of model-fitting analyses are presented in Table 4). As Table 4 indicates, C influences on all phenotypes could be dropped without significant loss of fit, whereas A could not (i.e. relative to the ACE model, fit statistic values were significantly lower for the “No C” models, but not for the “No A” models). We subsequently adopted an AE model as our base model.

Table 4
Biometric model fit statistics

Univariate heritability estimates obtained from the AE model, given in Table 5, indicate moderately large heritable differences in all phenotypes, with estimates relatively similar for the two genders. Genetic correlations from the AE model are presented in Table 6. These were largely similar for males and females, and indicate substantial genetic overlap among the ERP measures (P3 and TF-PC1). Overlap between these and beta was much more modest, and was approximately twice as great among males as among females. Genetic correlations with EXT are modest but significant with one exception: the P3-EXT correlation in females was not significant.

Table 5
Univariate heritability estimates
Table 6
Genetic correlations from AE model

Figure 1 illustrates the path diagram of the AE Cholesky model. A1 and E1 represent the additive genetic and non-shared environmental variance, respectively, that is shared by all three endophenotypes and EXT. A2 and E2 are the variances shared by TF-PC1, Beta, and EXT, independent of that shared with P3. A3 and E3 are the variances shared by Beta and EXT, independent of that shared with P3 and TF-PC1. A4 and E4 are the variances unique to EXT. Shown are the standardized path coefficients, which, when squared, represent the proportion of variance accounted for by the respective path. To assess the significance of each path, we constrained it equal to 0 and assessed fit by means of likelihood-ratio tests, AIC, and DIC. Significant likelihood-ratio tests and larger values of the information-theoretic fit statistics indicate significance of the path being tested. Results of these tests are presented in Table 4 for the genetic factors. The three fit statistics were consistent in showing a significant effect of A1 on the three endophenotypes, indicating shared genetic variance among them, for both males and females. The A2 influence on beta was not significant for either sex, indicating little genetic overlap between TF-PC1 and beta beyond what is also common to P3. Among males, only the A1 influence on EXT was significant, indicating shared genetic variance among all three endophenotypes and EXT. Among females, by contrast, the A1 influence on EXT was not significant but the influence of A2 and A3 both were. (The latter effect was somewhat marginal, in that setting this path to 0 resulted in a somewhat smaller value of DIC, although AIC was larger and the likelihood-ratio test was significant.)

Table 4 does not include similar statistics for the nonshared environmental effects. Because E includes measurement error as well as other sources of nonshared influence, the paths from each E to its corresponding phenotype (E1 on P3, E2 on TF-PC1, E3 on beta, and E4 on EXT) cannot be deleted. Constraining the remaining paths indicated that E1, the latent factor common to all phenotypes, had a significant influence on TF-PC1 in both males, Δχ2 = 170.5, p < .001, AIC = 11795.5, DIC = 11961.6, and females, Δχ2 = 223.1, p < .001, AIC = 11848.1, DIC = 12014.1. Among males, there was also a marginally significant effect of E1 on EXT, Δχ2 = 4.6, p = .032, AIC = 11629.6, DIC = 11795.6. (The latter actually represented a small decrease, indicating better fit.) None of the other paths was significant, with all chi-squared values less than or equal to 1.4.


In the present report, we explored a multivariate approach to identifying electrophysiological endophenotypes related to EXT psychopathology, using a large sample of male and female adolescent twins and measures identified in the literature as promising individual endophenotypes for EXT. The P3 ERP, its associated TF components, and beta frequency resting state EEG have each shown promise as endophenotypes for various disinhibitory disorders in the EXT spectrum and/or a general EXT factor. Using composite measures and fitting multivariate biometric Cholesky models, we specifically addressed the following aims: 1) assess the magnitude of the phenotypic and genetic associations between each endophenotype and EXT, 2) determine if there is significant shared genetic variance among the multiple endophenotypes and EXT, 3) determine if any of the endophenotypes share unique genetic variance with EXT (i.e. variance which is not also shared with the other endophenotypes), and 4) explore potential gender differences.

Results indicated that there was substantial genetic influence on each of the endophenotypes and that each was highly heritable – to a similar degree in males and females. These results are consistent with previous reports on heritability estimates of and genetic influences on the individual ERP and EEG measures and EXT (e.g. Begleiter & Porjesz, 2006; Hicks et al., 2007; van Beijsterveldt & Boomsma, 1994; van Beijsterveldt et al., 1996). Extending these findings, we also demonstrated that these three endophenotypes 1) show significant negative phenotypic and genetic correlations with a general vulnerability to EXT disorders, 2) show modest to strong phenotypic and genetic correlations with each other, and 3) are sensitive to genetic effects that may differ as a function of gender. These relationships suggest that these three endophenotypes are likely tapping into neurophysiological processes and genes that are both common across them and unique to each – all of which are relevant to a biological vulnerability to EXT psychopathology.

In order to empirically explore the extent to which the EXT-related genetic influences were shared or unique among the three candidate endophenotypes, we performed multivariate biometric models using a Cholesky parameterization. Biometric models revealed that a substantial portion of the variance in each of these candidate endophenotypes was due to additive genes. Further, there was a significant (although modest) amount of genetic variance that was shared between EXT and each endophenotype, as well as between the endophenotypes themselves. Genetic correlations were strong within the ERP domain (i.e. between P3 and TF-PC1) for both genders, and were moderate across domains (i.e. between ERP measures and beta), being about twice as strong for males as for females. The only exception to these results was the non-significant P3-EXT correlation in females.

The Cholesky model allowed us to further assess the genetic variance shared among various combinations of the candidate endophenotypes and EXT, in order to determine if any endophenotypes shared variance with EXT independent of that shared with the other endophenotypes. Based on the literature, P3 has shown the most robust genetic associations with EXT (e.g. Hicks et al., 2007; Hill et al., 1998; Porjesz & Rangaswamy, 2007), thus we entered P3 into the model first, followed by the P3-associated TF component, the beta power measure, and EXT. For both genders, there was significant genetic variance collectively shared among all the candidate endophenotypes, and for males this genetic variance was also shared with EXT. In males, this shared genetic variance was carried by the relationship between P3 and EXT, as there was no significant variance shared among the remaining endophenotypes and EXT independent of that shared with P3. In females, however, there was no significant genetic variance shared between P3 and EXT, whereas the TF component and beta power did share significant variance with EXT independent of their genetic overlap with P3. Further, the TF component and beta power each had significant independent genetic covariance with EXT in females.

Taken together, results suggest that these three endophenotypes are tapping into different neurophysiological and genetic substrates of EXT psychopathology. Results further suggest, however, that some of this genetic variance is common across the endophenotypes. These endophenotypes, then, are making independent, yet overlapping, contributions to the general genetic vulnerability to EXT (particularly in the case of females). This genetic variance that is shared among these measures is the key to their potential utility as a ‘multivariate endophenotype’ for EXT.

In the search for psychopathology-related genes, having an appropriate endophenotype is important for understanding the relationship between genes and behavior. Robust findings in genetic studies searching for gene-phenotype associations are generally hampered by a lack of statistical power due to the fact that a) most subject samples are typically heterogeneous with regard to specific phenotypic characteristics, and b) genes affecting any measured phenotype usually only account for a small proportion of variance of that phenotype (Plomin, 2005; Plomin, Owen, & McGuffin, 1994; Viding & Blakemore, 2007). Also, the route from genes to behavior is typically complex, with multiple biological and cognitive pathways converging on the same behavioral phenotype (Viding & Blakemore, 2007).

This potential complexity is evidenced in the present report by the significant genetic associations between brain electrophysiological endophenotypes from different domains (stimulus-evoked P3 and TF components, and intrinsic, resting state beta activity) and EXT psychopathology. Functionally, a genetic predisposition to disinhibitory disorders has been associated with central nervous system hyperexcitability (Begleiter & Porjesz, 1999). Reduced amplitudes in the brain measures used in the present report theoretically reflect different manifestations of a dysfunction in the capacity for neuronal inhibition and in inhibitory control processes (Begleiter & Porjesz, 1999; Iacono et al., 2000). Additionally, each endophenotype has been linked to different genes, all of which have also been associated with EXT-spectrum disorders (e.g. Chen et al., 2009; Dick et al., 2008; Edenberg et al., 2004; Jones et al., 2004; Williams et al., 1999). Thus, there is theoretical and empirical support for the measures in the present report with regard to their potential collective utility as a multivariate endophenotype for EXT.

Using these endophenotypes collectively, then, has the potential to overcome the problem of heterogeneous subject samples by increasing the ability to identify those in the population with the endophenotypes who are at especially high risk for EXT disorders (and, thus, more likely to be true gene carriers; e.g. Iacono et al., 2000). The small to moderate effect sizes we found when comparing EXT between those scoring high and low on the endophenotypes provides initial evidence supporting this hypothesis. A multivariate endophenotype could also be advantageous given the polygenic influences on the general vulnerability to EXT. Endophenotypes are putatively less genetically complex correlates of the disorder that, because they tap the neurobiological basis of gene effects, are closer to gene action. The extent to which the genetic influences overlap between the endophenotypes and EXT, then, may provide increased power to detect "hits" for genetic linkage and association studies that may account for exceedingly small proportions of variance in the EXT phenotype (Gottesman & Gould, 2003). These are necessary avenues of future research in evaluating the utility of this multivariate approach to defining endophenotypes for EXT.

An important, novel finding in the present report was differential patterns of shared genetic variance as a function of gender. In males, the TF and beta endophenotypes accounted for an additional, yet non-significant, increment in genetic variance in EXT above that accounted for by P3; whereas in females, the TF and beta endophenotypes did account for significant variance above and beyond that shared with P3. Based on these results, the question arises as to the best way to utilize these endophenotypes collectively in genetic studies. Two possibilities emerge: 1) use a multivariate endophenotype that is specific to each gender, thus requiring two “passes” in the search for EXT-related genes, each of which may be ideally suited for males or females, or 2) use the same multivariate endophenotype for both genders, which requires only one pass at the data, but may include a measure that is not associated with significant variance in EXT. Given the depth of information contained in the genetic code, we submit that the latter approach may be the more fruitful one 1) in terms of efficiency, and 2) because the genetic etiology of the individual measures themselves is also of interest. Including all the measures regardless of their gender-specific association with EXT may, nonetheless, yield interesting results. Further research is necessary to address these gender-specific issues.

Given the exploratory nature of the present report, it necessarily has some notable limitations. The brain electrophysiological measures were recorded from a limited number of electrodes: 3 parietal sites for P3 and its related TF components, and Cz and bilateral parietal sites for beta power. The implications are that we have limited topographic information and that we may not be capturing all measures that have been associated with EXT in the literature. For instance, a frontal theta TF component has been associated with alcoholism (Jones et al., 2006). It will be important to examine a denser array of electrodes in followup studies. Also, the sample consisted solely of twins in late adolescence and was not racially or ethnically diverse. This age (~17-years-old), however, is that at which the most is known regarding the endophenotypic relationship between the measures in the current study and EXT (e.g. Gilmore et al., in press; Iacono et al., 2002; Rangaswamy et al., 2007). Thus, we initially focused our efforts on this age. Further research needs to be done, however, regarding the generalizability of the findings to other age, racial, and ethnic groups.


In the present report, we explored a multivariate approach to identifying electrophysiological endophenotypes related to EXT psychopathology. The P3 ERP, its associated TF components, and beta frequency resting state EEG were shown to be heritable and to collectively share genetic variance in EXT, suggesting that these measures may be tapping into common neurophysiological and genetic substrates associated with EXT psychopathology. This shared genetic variance, however, differed as a function of gender. P3 seemed to be more closely related to EXT in males, whereas beta and the TF component were more related in females. Regardless of the gender difference, however, current findings of significant relationships among these electrophysiological endophenotypes and EXT would suggest that a multivariate endophenotype including these measures constitutes a worthy target, in addition to consideration of any of the individual measures, for studies investigating the genetic etiology of externalizing spectrum disorders.


This research was supported by grants DA 05147, AA 09367, DA 024417, and K01 AA 015621 (Malone) from the National Institutes of Health.


  • Agrawal A, Edenberg HJ, Foroud T, Bierut LJ, Dunne G, Hinrichs AL, et al. Association of GABRA2 with drug dependence in the collaborative study of the genetics of alcoholism sample. Behav Genet. 2006;36(5):640–650. [PubMed]
  • Akaike H. A new look at the statistical model identification. IEEE Transactions on Automatic Control. 1974;19(6):716–723.
  • American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 3rd ed., rev. Washington, D.C: Author; 1987.
  • Amin N, Aulchenko YS, Dekker MC, Ferdinand RF, van Spreeken A, Temmink AH, et al. Suggestive linkage of ADHD to chromosome 18q22 in a young genetically isolated Dutch population. Eur J Hum Genet. 2009 [PMC free article] [PubMed]
  • Arya R, Duggirala R, Williams JT, Almasy L, Blangero J. Power to localize the major gene for disease liability is increased after accounting for the effects of related quantitative phenotypes. Genet Epidemiol. 2001;21 Suppl 1:S774–S778. [PubMed]
  • Basar-Eroglu C, Basar E, Demiralp T, Schurmann M. P300-response: possible psychophysiological correlates in delta and theta frequency channels. A review. Int J Psychophysiol. 1992;13(2):161–179. [PubMed]
  • Bauer LO. Electroencephalographic and autonomic predictors of relapse in alcohol-dependent patients. Alcohol Clin Exp Res. 1994;18(3):755–760. [PubMed]
  • Bauer LO. Predicting relapse to alcohol and drug abuse via quantitative electroencephalography. Neuropsychopharmacology. 2001;25(3):332–340. [PubMed]
  • Begleiter H, Porjesz B. What is inherited in the predisposition toward alcoholism? A proposed model. Alcohol Clin Exp Res. 1999;23(7):1125–1135. [PubMed]
  • Begleiter H, Porjesz B. Genetics of human brain oscillations. Int J Psychophysiol. 2006;60(2):162–171. [PubMed]
  • Begleiter H, Porjesz B, Bihari B, Kissin B. Event-related brain potentials in boys at risk for alcoholism. Science. 1984;225(4669):1493–1496. [PubMed]
  • Berman SM, Whipple SC, Fitch RJ, Noble EP. P3 in young boys as a predictor of adolescent substance use. Alcohol. 1993;10(1):69–76. [PubMed]
  • Bernat EM, Malone SM, Williams WJ, Patrick CJ, Iacono WG. Decomposing delta, theta, and alpha time-frequency ERP activity from a visual oddball task using PCA. Int J Psychophysiol. 2007;64(1):62–74. [PMC free article] [PubMed]
  • Bernat EM, Williams WJ, Gehring WJ. Decomposing ERP time-frequency energy using PCA. Clin Neurophysiol. 2005;116(6):1314–1334. [PubMed]
  • Bice P, Valdar W, Zhang L, Liu L, Lai D, Grahame N, et al. Genomewide SNP Screen to Detect Quantitative Trait Loci for Alcohol Preference in the High Alcohol Preferring and Low Alcohol Preferring Mice. Alcohol Clin Exp Res. 2008 [PMC free article] [PubMed]
  • Burt SA, Mikolajewski AJ. Preliminary evidence that specific candidate genes are associated with adolescent-onset antisocial behavior. Aggress Behav. 2008;34(4):437–445. [PubMed]
  • Buzsáki G. Rhythms of the brain. New York: Oxford University Press; 2006.
  • Carlson SR, Iacono WG, McGue M. P300 amplitude in non-alcoholic adolescent twin pairs who become discordant for alcoholism as adults. sychophysiology. 2004;41:841–844. [PubMed]
  • Chen AC, Tang Y, Rangaswamy M, Wang JC, Almasy L, Foroud T, et al. Association of single nucleotide polymorphisms in a glutamate receptor gene (GRM8) with theta power of event-related oscillations and alcohol dependence. Am J Med Genet B Neuropsychiatr Genet. 2009;150B(3):359–368. [PMC free article] [PubMed]
  • Clarke AR, Barry RJ, McCarthy R, Selikowitz M. Age and sex effects in the EEG: differences in two subtypes of attention-deficit/hyperactivity disorder. Clin Neurophysiol. 2001;112(5):815–826. [PubMed]
  • Clarke AR, Barry RJ, McCarthy R, Selikowitz M. Children with attention-deficit/hyperactivity disorder and comorbid oppositional defiant disorder: an EEG analysis. Psychiatry Res. 2002;111(2–3):181–190. [PubMed]
  • Clarke AR, Barry RJ, McCarthy R, Selikowitz M, Clarke DC, Croft RJ. EEG activity in girls with attention-deficit/hyperactivity disorder. Clin Neurophysiol. 2003;114(2):319–328. [PubMed]
  • Cook RD, Weisberg S. Residuals and Influence in Regression. New York: Chapman and Hall; 1982.
  • Costa L, Bauer L. Quantitative electroencephalographic differences associated with alcohol, cocaine, heroin and dual-substance dependence. Drug Alcohol Depend. 1997;46(1–2):87–93. [PubMed]
  • Covault J, Gelernter J, Jensen K, Anton R, Kranzler HR. Markers in the 5′-region of GABRG1 associate to alcohol dependence and are in linkage disequilibrium with markers in the adjacent GABRA2 gene. Neuropsychopharmacology. 2008;33(4):837–848. [PMC free article] [PubMed]
  • Cowan WM, Kopnisky KL, Hyman SE. The human genome project and its impact on psychiatry. Annu Rev Neurosci. 2002;25:1–50. [PubMed]
  • Dick DM, Aliev F, Wang JC, Grucza RA, Schuckit M, Kuperman S, et al. Using dimensional models of externalizing psychopathology to aid in gene identification. Arch Gen Psychiatry. 2008;65(3):310–318. [PubMed]
  • Dick DM, Plunkett J, Wetherill LF, Xuei X, Goate A, Hesselbrock V, et al. Association between GABRA1 and drinking behaviors in the collaborative study on the genetics of alcoholism sample. Alcohol Clin Exp Res. 2006;30(7):1101–1110. [PubMed]
  • Dien J, Spencer KM, Donchin E. Localization of the event-related potential novelty response as defined by principal components analysis. Brain Res Cogn Brain Res. 2003;17(3):637–650. [PubMed]
  • Donchin E, Coles M. Is the P300 component a manifestation of context updating? Behavioral & Brain Sciences. 1988;11:357–374.
  • Draper D. Assessment and Propagation of Model Uncertainty. Journal of the Royal Statistical Society. Series B (Methodological) 1995;57(1):45–97.
  • Duncan-Johnson CC, Donchin E. On quantifying surprise: the variation of event-related potentials with subjective probability. Psychophysiology. 1977;14(5):456–467. [PubMed]
  • Edenberg HJ, Dick DM, Xuei X, Tian H, Almasy L, Bauer LO, et al. Variations in GABRA2, encoding the alpha 2 subunit of the GABA(A) receptor, are associated with alcohol dependence and with brain oscillations. American Journal of Human Genetics. 2004;74:705–714. [PMC free article] [PubMed]
  • Enoch MA. The role of GABA(A) receptors in the development of alcoholism. Pharmacol Biochem Behav. 2008;90(1):95–104. [PMC free article] [PubMed]
  • Fox J. Companion to Applied Regression. Hamilton, Ontario: 2009. (Version 1.2–16) [Package for R]
  • Frederick JA, Iacono WG. Beyond the DSM: defining endophenotypes for genetic studies of substance abuse. Curr Psychiatry Rep. 2006;8(2):144–150. [PubMed]
  • Gilmore CS, Malone SM, Bernat EM, Iacono WG. Relationship between the P3 event-related potential, its associated time-frequency components, and externalizing psychopathology. Psychophysiology. (in press) [PMC free article] [PubMed]
  • Goldman D, Oroszi G, Ducci F. The genetics of addictions: uncovering the genes. Nat Rev Genet. 2005;6(7):521–532. [PubMed]
  • Gottesman II, Gould TD. The endophenotype concept in psychiatry: etymology and strategic intentions. Am J Psychiatry. 2003;160(4):636–645. [PubMed]
  • Gratton G, Coles MGH, Donchin E. A new method for off-line removal of ocular artifact. Electroencephalography and Clinical Neurophsyiology. 1983;55:468–484. [PubMed]
  • Hicks BM, Bernat E, Malone SM, Iacono WG, Patrick CJ, Krueger RF, et al. Genes mediate the association between P3 amplitude and externalizing disorders. Psychophysiology. 2007;44(1):98–105. [PMC free article] [PubMed]
  • Hicks BM, Krueger RF, Iacono WG, McGue M, Patrick CJ. Family transmission and heritability of externalizing disorders: a twin-family study. Arch Gen Psychiatry. 2004;61(9):922–928. [PubMed]
  • Hill SY. Trajectories of alcohol use and electrophysiological and morphological indices of brain development: distinguishing causes from consequences. Ann N Y Acad Sci. 2004;1021:245–259. [PubMed]
  • Hill SY, Locke J, Steinhauer SR. Absence of visual and auditory P300 reduction in nondepressed male and female alcoholics. Biol Psychiatry. 1999;46(7):982–989. [PubMed]
  • Hill SY, Locke J, Zezza N, Kaplan B, Neiswanger K, Steinhauer SR, et al. Genetic association between reduced P300 amplitude and the DRD2 dopamine receptor A1 allele in children at high risk for alcoholism. Biol Psychiatry. 1998;43(1):40–51. [PubMed]
  • Hill SY, Muka D, Steinhauer S, Locke J. P300 amplitude decrements in children from families of alcoholic female probands. Biol Psychiatry. 1995;38(9):622–632. [PubMed]
  • Hill SY, Steinhauer SR. Event-related potentials in women at risk for alcoholism. Alcohol. 1993;10(5):349–354. [PubMed]
  • Hill SY, Steinhauer SR, Locke-Wellman J, Ulrich R. Childhood Risk Factors for Young Adult Substance Dependence Outcome in Offspring from Multiplex Alcohol Dependence Families: A Prospective Study. Biol Psychiatry. 2009 [PMC free article] [PubMed]
  • Hobbs MJ, Clarke AR, Barry RJ, McCarthy R, Selikowitz M. EEG abnormalities in adolescent males with AD/HD. Clin Neurophysiol. 2007;118(2):363–371. [PubMed]
  • Holdcraft LC, Iacono WG, McGue MK. Antisocial Personality Disorder and depression in relation to alcoholism: a community-based sample. J Stud Alcohol. 1998;59(2):222–226. [PubMed]
  • Hopfer CJ, Lessem JM, Hartman CA, Stallings MC, Cherny SS, Corley RP, et al. A genome-wide scan for loci influencing adolescent cannabis dependence symptoms: evidence for linkage on chromosomes 3 and 9. Drug Alcohol Depend. 2007;89(1):34–41. [PMC free article] [PubMed]
  • Houston RJ, Stanford MS. Electrophysiological substrates of impulsiveness: potential effects on aggressive behavior. Prog Neuropsychopharmacol Biol Psychiatry. 2005;29(2):305–313. [PubMed]
  • Iacono WG. Identifying psychophysiological risk for psychopathology: examples from substance abuse and schizophrenia research. Psychophysiology. 1998;35(6):621–637. [PubMed]
  • Iacono WG, Carlson SR, Malone SM. Identifying a multivariate endophenotype for substance use disorders using psychophysiological measures. International Journal of Psychophysiology. 2000;38(1):81–96. [PubMed]
  • Iacono WG, Carlson SR, Malone SM, McGue M. P3 event-related potential amplitude and the risk for disinhibitory disorders in adolescent boys. Arch Gen Psychiatry. 2002;59(8):750–757. [PubMed]
  • Iacono WG, McGue M. Minnesota Twin Family Study. Twin Res. 2002;5(5):482–487. [PubMed]
  • Iacono WG, McGue M. Association between P3 event-related brain potential amplitude and adolescent problem behavior. Psychophysiology. 2006;43(5):465–469. [PubMed]
  • Jones KA, Porjesz B, Almasy L, Bierut L, Goate A, Wang JC, et al. Linkage and linkage disequilibrium of evoked EEG oscillations with CHRM2 receptor gene polymorphisms: implications for human brain dynamics and cognition. Int J Psychophysiol. 2004;53(2):75–90. [PubMed]
  • Jones KA, Porjesz B, Chorlian D, Rangaswamy M, Kamarajan C, Padmanabhapillai A, et al. S-transform time-frequency analysis of P300 reveals deficits in individuals diagnosed with alcoholism. Clin Neurophysiol. 2006;117(10):2128–2143. [PubMed]
  • Kamarajan C, Porjesz B, Jones K, Chorlian D, Padmanabhapillai A, Rangaswamy M, et al. Event-related oscillations in offspring of alcoholics: neurocognitive disinhibition as a risk for alcoholism. Biol Psychiatry. 2006;59(7):625–634. [PMC free article] [PubMed]
  • Kendler KS, Prescott CA, Myers J, Neale MC. The structure of genetic and environmental risk factors for common psychiatric and substance use disorders in men and women. Arch Gen Psychiatry. 2003;60(9):929–937. [PubMed]
  • Krueger RF, Hicks BM, Patrick CJ, Carlson SR, Iacono WG, McGue M. Etiologic connections among substance dependence, antisocial behavior, and personality: modeling the externalizing spectrum. J Abnorm Psychol. 2002;111(3):411–424. [PubMed]
  • Krueger RF, Watson D, Barlow DH. Introduction to the special section: toward a dimensionally based taxonomy of psychopathology. J Abnorm Psychol. 2005;114(4):491–493. [PMC free article] [PubMed]
  • Leckman JF, Scholomskas D, Thompson WD, Belanger A, Weisman MM. Best estimate of lifetime psychiatric diagnosis: A methodlogical study. Archives of General Psychiatry. 1982;39:879–883. [PubMed]
  • Luo X, Kranzler HR, Zuo L, Wang S, Blumberg HP, Gelernter J. CHRM2 gene predisposes to alcohol dependence, drug dependence and affective disorders: results from an extended case-control structured association study. Hum Mol Genet. 2005;14(16):2421–2434. [PubMed]
  • Mann CA, Lubar JF, Zimmerman AW, Miller CA, Muenchen RA. Quantitative analysis of EEG in boys with attention-deficit-hyperactivity disorder: controlled study with clinical implications. Pediatr Neurol. 1992;8(1):30–36. [PubMed]
  • Mantini D, Corbetta M, Perrucci MG, Romani GL, Del Gratta C. Large-scale brain networks account for sustained and transient activity during target detection. Neuroimage. 2009;44(1):265–274. [PubMed]
  • McGough JJ, Loo SK, McCracken JT, Dang J, Clark S, Nelson SF, et al. CBCL Pediatric Bipolar Disorder Profile and ADHD: Comorbidity and Quantitative Trait Loci Analysis. J Am Acad Child Adolesc Psychiatry. 2008 [PMC free article] [PubMed]
  • McGuire KA, Katsanis J, Iacono WG, McGue M. Genetic influences on the spontaneous EEG: An examination of 15-year-old and 17-year-old twins. Developmental Neuropsychology. 1998;14:7–18.
  • Merikangas KR, Risch N. Genomic priorities and public health. Science. 2003;302(5645):599–601. [PubMed]
  • Parsons OA, Sinha R, Williams HL. Relationships between neuropsychological test performance and event- related potentials in alcoholic and nonalcoholic samples. Alcohol Clin Exp Res. 1990;14(5):746–755. [PubMed]
  • Patrick CJ, Bernat EM, Malone SM, Iacono WG, Krueger RF, McGue M. P300 amplitude as an indicator of externalizing in adolescent males. Psychophysiology. 2006;43(1):84–92. [PMC free article] [PubMed]
  • Plomin R. Finding genes in child psychology and psychiatry: when are we going to be there? J Child Psychol Psychiatry. 2005;46(10):1030–1038. [PubMed]
  • Plomin R, Owen MJ, McGuffin P. The genetic basis of complex human behaviors. Science. 1994;264(5166):1733–1739. [PubMed]
  • Polich J, Pollock VE, Bloom FE. Meta-analysis of P300 amplitude from males at risk for alcoholism. Psychol Bull. 1994;115(1):55–73. [PubMed]
  • Pollock VE, Earleywine M, Gabrielli WF. Personality and EEG beta in older adults with alcoholic relatives. Alcohol Clin Exp Res. 1995;19(1):37–43. [PubMed]
  • Porjesz B, Almasy L, Edenberg HJ, Wang K, Chorlian DB, Foroud T, et al. Linkage disequilibrium between the beta frequency of the human EEG and a GABAA receptor gene locus. PG - 3729-33. Proc Natl Acad Sci U S A. 2002;99(6) [PMC free article] [PubMed]
  • Porjesz B, Begleiter H, Reich T, Van Eerdewegh P, Edenberg HJ, Foroud T, et al. Collaborative Study on the Genetics of Alcoholism. Amplitude of visual P3 event-related potential as a phenotypic marker for a predisposition to alcoholism: preliminary results from the COGA Project. Alcohol Clin Exp Res. 1998;22(6):1317–1323. [PubMed]
  • Porjesz B, Rangaswamy M. Neurophysiological endophenotypes, CNS disinhibition, and risk for alcohol dependence and related disorders. ScientificWorldJournal. 2007;7:131–141. [PubMed]
  • Porjesz B, Rangaswamy M, Kamarajan C, Jones KA, Padmanabhapillai A, Begleiter H. The utility of neurophysiological markers in the study of alcoholism. Clin Neurophysiol. 2005;116(5):993–1018. [PubMed]
  • Prabhu VR, Porjesz B, Chorlian DB, Wang K, Stimus A, Begleiter H. Visual p3 in female alcoholics. Alcohol Clin Exp Res. 2001;25(4):531–539. [PubMed]
  • Price GW, Michie PT, Johnston J, Innes-Brown H, Kent A, Clissa P, et al. A multivariate electrophysiological endophenotype, from a unitary cohort, shows greater research utility than any single feature in the Western Australian family study of schizophrenia. Biol Psychiatry. 2006;60(1):1–10. [PubMed]
  • Rangaswamy M, Jones KA, Porjesz B, Chorlian DB, Padmanabhapillai A, Kamarajan C, et al. Delta and theta oscillations as risk markers in adolescent offspring of alcoholics. Int J Psychophysiol. 2007;63(1):3–15. [PMC free article] [PubMed]
  • Rangaswamy M, Porjesz B, Chorlian DB, Wang K, Jones KA, Bauer LO, et al. Beta power in the EEG of alcoholics. Biol Psychiatry. 2002;52(8):831–842. [PubMed]
  • Rangaswamy M, Porjesz B, Chorlian DB, Wang K, Jones KA, Kuperman S, et al. Resting EEG in offspring of male alcoholics: beta frequencies. Int J Psychophysiol. 2004;51(3):239–251. [PubMed]
  • Reich W. Diagnostic interview for children and adolescents (DICA) Journal of the American Academy of Child & Adolescent Psychiatry. 2000;39:59–66. [PubMed]
  • Robins LN, Babor TF, Cottler LB. Composite International Diagnostic Interview: Expanded Substance Abuse Module. St. Louis: Authors; 1987.
  • Rommelse NN, Arias-Vasquez A, Altink ME, Buschgens CJ, Fliers E, Asherson P, et al. Neuropsychological endophenotype approach to genome-wide linkage analysis identifies susceptibility loci for ADHD on 2q21.1 and 13q12.11. Am J Hum Genet. 2008;83(1):99–105. [PMC free article] [PubMed]
  • Schork NJ, Greenwood TA, Braff DL. Statistical genetics concepts and approaches in schizophrenia and related neuropsychiatric research. Schizophr Bull. 2007;33(1):95–104. [PMC free article] [PubMed]
  • Schuckit MA. An overview of genetic influences in alcoholism. J Subst Abuse Treat. 2009;36(1):S5–S14. [PubMed]
  • Soltani M, Knight RT. Neural origins of the P300. Crit Rev Neurobiol. 2000;14(3–4):199–224. [PubMed]
  • Soyka M, Preuss UW, Hesselbrock V, Zill P, Koller G, Bondy B. GABA-A2 receptor subunit gene (GABRA2) polymorphisms and risk for alcohol dependence. J Psychiatr Res. 2008;42(3):184–191. [PubMed]
  • Stallings MC, Corley RP, Hewitt JK, Krauter KS, Lessem JM, Mikulich SK, et al. A genome-wide search for quantitative trait loci influencing substance dependence vulnerability in adolescence. Drug Alcohol Depend. 2003;70(3):295–307. [PubMed]
  • van Beijsterveldt CE, Boomsma DI. Genetics of the human electroencephalogram (EEG) and event-related brain potentials (ERPs): a review. Hum Genet. 1994;94(4):319–330. [PubMed]
  • van Beijsterveldt CE, Molenaar PC, de Geus EJ, Boomsma DI. Heritability of human brain functioning as assessed by electroencephalography. Am J Hum Genet. 1996;58(3):562–573. [PMC free article] [PubMed]
  • van Beijsterveldt CE, van Baal GC. Twin and family studies of the human electroencephalogram: a review and a meta-analysis. Biol Psychol. 2002;61(1–2):111–138. [PubMed]
  • Van Der Stelt O. European Society for Biomedical Research on Alcoholism. ESBRA-Nordmann 1998 Award Lecture: Visual P3 as a potential vulnerability marker of alcoholism: evidence from the Amsterdam study of children of alcoholics. Alcohol Alcohol. 1999;34(3):267–282. [PubMed]
  • Viding E, Blakemore SJ. Endophenotype approach to developmental psychopathology: implications for autism research. Behav Genet. 2007;37(1):51–60. [PubMed]
  • Wall TL, Shea SH, Luczak SE, Cook TA, Carr LG. Genetic associations of alcohol dehydrogenase with alcohol use disorders and endophenotypes in white college students. J Abnorm Psychol. 2005;114(3):456–465. [PubMed]
  • Wang JC, Hinrichs AL, Stock H, Budde J, Allen R, Bertelsen S, et al. Evidence of common and specific genetic effects: association of the muscarinic acetylcholine receptor M2 (CHRM2) gene with alcohol dependence and major depressive syndrome. Hum Mol Genet. 2004;13(17):1903–1911. [PubMed]
  • Williams JT, Begleiter H, Porjesz B, Edenberg HJ, Foroud T, Reich T, et al. Joint multipoint linkage analysis of multivariate qualitative and quantitative traits. II. Alcoholism and event-related potentials. Am J Hum Genet. 1999;65(4):1148–1160. [PMC free article] [PubMed]
  • Xuei X, Flury-Wetherill L, Dick D, Goate A, Tischfield J, Nurnberger J, Jr, et al. GABRR1 and GABRR2, encoding the GABA-A receptor subunits rho1 and rho2, are associated with alcohol dependence. Am J Med Genet B Neuropsychiatr Genet. 2009 [PMC free article] [PubMed]
  • Yoon HH, Iacono WG, Malone SM, McGue M. Using the brain P300 response to identify novel phenotypes reflecting genetic vulnerability for adolescent substance misuse. Addict Behav. 2006;31(6):1067–1087. [PubMed]
  • Yordanova J, Devrim M, Kolev V, Ademoglu A, Demiralp T. Multiple time-frequency components account for the complex functional reactivity of P300. Neuroreport. 2000;11(5):1097–1103. [PubMed]
  • Young SE, Stallings MC, Corley RP, Krauter KS, Hewitt JK. Genetic and environmental influences on behavioral disinhibition. Am J Med Genet. 2000;96(5):684–695. [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...