Child Abuse and Neglect, MAOA, and Mental Health Outcomes: A Prospective Examination

Abstract

Background

Studies have examined the interaction of MAOA genotype with childhood maltreatment in relation to depressive symptomatology and alcohol abuse with conflicting findings. Both high and low activity allele combinations have been shown to be protective for maltreated children with direction of findings varying by study methodology and participant’s sex.

Methods

Participants in a prospective cohort design study involving court substantiated cases of child abuse and neglect and a matched comparison group were followed up into adulthood and interviewed (N = 802). Eighty-two percent consented to provide blood, 631 gave permission for DNA extraction and analyses, and 575 were included in the final sample. This sample included male, female, White, and Non-White (primarily Black) participants. Symptoms of dysthymia, major depression and alcohol abuse were assessed using the NIMH Diagnostic Interview Schedule-III-R.

Results

Significant three-way interactions, MAOA genotype by abuse by sex, predicted dysthymic symptoms. Low-activity MAOA genotype buffered against symptoms of dysthymia in physically abused and multiply maltreated women. Significant three-way interactions, MAOA genotype by sexual abuse by race, predicted all outcomes. Low-activity MAOA genotype buffered against symptoms of dysthymia, major depressive disorder and alcohol abuse for sexually abused White participants. The high-activity genotype was protective in the Non-White sexually abused group.

Conclusions

This prospective study provides evidence that MAOA interacts with child maltreatment to predict mental health outcomes. Reasons for sex differences and race findings are discussed.

Research has shown that adults who were abused and/or neglected as children are at increased risk for a variety of negative consequences, including depression, alcohol abuse, and violent and criminal behavior [1–19]. At the same time, several scholars have called attention to a group of children who appear resilient despite histories of childhood adversities including maltreatment [20–27].

Rresearchers have begun to examine gene by environment interactions to understand differences in outcomes for children exposed to early adversities. Studies have focused on monoamine oxidase A (MAOA) genotype as a moderator between the environmental stressor of child maltreatment and subsequent outcomes. MAOA is a gene located on the X chromosome (Xp11.23-11.4) that encodes the MAOA enzyme which breaks down neurotransmitters serotonin (5-HT), norepinephrine (NE) and dopamine (DA) [28]. Variability in MAOA genotype is due to a polymorphism of the MAOA gene with a variable number of tandem repeats (VNTR). The variability has been shown to correlate with transcription with shorter, 3-repeat allele, being less active than the longer, 4-repeat allele [29–31]. MAOA has been implicated in affective functioning in knock-out mice suggesting that normal MAOA activity is important for anger expression [32]. Neurotransmitters 5-HT, NE and DA that MAOA metabolizes have been linked to depressive and substance abuse symptomatology [33] which are also outcomes predicted by child maltreatment [2, 3, 10, 34–36].

Caspi et al [37] found that monoamine oxidase A (MAOA) genotype moderated the relationship between child maltreatment and antisocial and violent offences in males, with high-activity MAOA protecting maltreated children. This research has since been replicated [38–40], replicated in part [41–43], or not replicated [44, 45]. A recent meta-analysis showed a significant effect of the interaction of MAOA genotype and child maltreatment on antisocial behaviors [46]. Most these studies have been conducted with White males. However, some research [47, 48] with females reported that low-activity allele was protective, opposite to Caspi et al [37] and one study that examined these relationships separately in Blacks and Whites found different results [41].

Although the MAOA enzyme metabolizes neurotransmitters implicated in depressive symptomatology and alcohol abuse, few studies have examined whether MAOA interacts with child maltreatment to predict these outcomes. Cicchetti et al. [49] reported a significant interaction of MAOA genotype and level of maltreatment in the prediction of depressive symptomatology using a sample of low-income adolescents. Children who were more severely maltreated and had low MAOA activity displayed more symptoms of depression than those with high-activity MAOA and severe levels of maltreatment. A study of White adults [50], using retrospective self-reports of physical and sexual abuse, found that MAOA genotype interacted with abuse to predict Major Depressive Disorder symptoms in men and women, but in the opposite direction to Cicchetti et al [49]. Abuse and high activity MAOA were associated with more depression symptoms, compared to those with low activity MAOA and abuse histories [50].

Research on alcohol abuse to date has also produced conflicting results. In a sample of Native American adult women, Ducci et al [51] found that MAOA genotype interacted with self-reported childhood sexual abuse to predict alcohol abuse: individuals with low levels of MAOA and abuse histories reported a higher number of symptoms, compared to those with high levels of MAOA and abuse histories. Studies with adolescents in Sweden [52], [53] found a clear difference between boys and girls in the interaction of family adversity and genotype predicting alcohol-related problem behaviors and risky alcohol consumption. Boys with the low activity and girls with high activity MAOA genotype were at increased risk, if they were also exposed to family adversity [54].

In summary, several studies have reported interactions of the MAOA genotype and child maltreatment to predict a variety of outcomes. However, the direction of these relationships has varied considerably, with differences in the operationalization of maltreatment, study design (cross-sectional versus longitudinal), sex, race/ethnicity, and outcome assessed. Some studies have used documented cases of childhood abuse and neglect, whereas others have relied on retrospective self-reports. Thus, it is difficult to draw firm conclusions about these relations.

The current research builds upon previous studies by examining the interaction of child maltreatment and MAOA genotype using a prospective cohort design with court documented cases of childhood physical abuse, sexual abuse, and neglect and matched controls who were followed up into adulthood and assessed. This study offers several advantages for testing the gene (MAOA) x environment (child abuse and neglect) interaction. First, this study involves a clear operationalization and lack of ambiguity about the child maltreatment in our sample by using documented cases of child abuse and neglect. Second, there is a comparison group matched on the basis of age, sex, race/ethnicity and approximate social class background [55]. Third, we use a structured diagnostic interview to assess psychiatric disorders. Fourth, we have a diverse sample with males, females, Whites, and non-Whites. Finally, we have followed our participants into adulthood so that we can compare risks beyond adolescence and into adulthood.

This paper examines the interaction of child maltreatment and the MAOA genotype in relation to three adult mental health outcomes -- major depressive disorder (MDD), dysthymia (DD), which has been shown to have a common genetic underpinning to MDD [56], and alcohol abuse (AA). We ask two basic questions: (1) Does child maltreatment interact with MAOA genotype to predict MDD, DD, and AA? (2) Does the interaction between childhood maltreatment and MAOA genotype in predicting outcomes of MDD, DD, and AA differ by race, sex or type of maltreatment experienced?

Method

Participants

This research is based on a specialized cohort design study in which abused and neglected children were matched with non-abused and non-neglected children and followed prospectively into young adulthood [18]. Only court substantiated cases of child abuse and neglect were included. Cases were drawn from the records of county juvenile and adult criminal courts in a metropolitan area in the Midwest during the years 1967 through 1971. To avoid potential problems with ambiguity in the direction of causality, and to ensure that temporal sequence was clear (that is, child neglect or abuse led to subsequent outcomes), abuse and neglect cases were restricted to those in which children were 11 years of age or less at the time of the abuse or neglect incident.

The abuse and neglect group includes three types: physical abuse, sexual abuse, and neglect. Physical abuse cases included injuries such as bruises, welts, burns, abrasions, lacerations, wounds, cuts, bone and skull fractures, and other evidence of physical injury. Sexual abuse charges varied from relatively non-specific charges of “assault and battery with intent to gratify sexual desires” to more specific charges of “fondling or touching in an obscene manner,” sodomy, incest, rape, and so forth. Neglect cases reflected a judgment that the parents’ deficiencies in child care were beyond those found acceptable by community and professional standards at the time. These cases represented extreme failure to provide adequate food, clothing, shelter, and medical attention to children.

A critical element of the design was the establishment of a comparison group, matched to the maltreatment group as closely as possible on the basis of sex, age, race, and approximate family socio-economic status during the time period under study (1967 through 1971). To accomplish this matching, the sample of abused and neglected cases was first divided into two groups on the basis of their age at the time of the abuse or neglect incident. Children who were under school age at the time of the abuse or neglect were matched with children of the same sex, race, date of birth (+/− 1 week), and hospital of birth through the use of county birth record information. For children of school age, records of more than 100 elementary schools for the same time period were used to find matches with children of the same sex, race, date of birth (+/− 6 months), and same class in same elementary school during the years 1967 through 1971. Overall, there were matches for 74% of the abused and neglected children [55].

This research began in 1986 as an archival study focused on the “cycle of violence” [18]. A second phase located and interviewed 1,196 individuals between 1989 and 1995 and 896 of the study participants were re-interviewed in 2001–2002. Data for the present study were collected during 2003–2005 in the context of a medical status examination (including blood collection through venipuncture) and interview of 806 participants. Of those interviewed, 638 (82%) consented to provide blood and 631 gave permission for DNA extraction and analyses. IRB regulations that define prisoners as a vulnerable population requiring additional protections prevented the collection of blood from 31 incarcerated participants.

Although there was attrition associated with death, refusals, and our inability to locate individuals over the various waves of the study, the composition of the samples at the four time points has remained about the same. The abuse and neglect group represented 56–58% at each time period; Whites were 62–66%; and males were 48–51% of the samples. There were no significant differences across the samples on these variables or in mean age across the four phases of the study.

The average age of participants is 41 years old (SD = 3.85; range = 31–51). Approximately half the sample is female (48.7%) and 60.8% is White, non-Hispanic. Using participant-reported race/ethnicity, the non-white/minority group includes African Americans (35.1%) and Hispanics (4.1%). The average highest grade of school completed for the sample was 11.47 (SD = 2.19) and the median occupational level [57] was semi-skilled workers (only 11.3% were in the professions). Thus, the overall sample is skewed toward the lower end of the socio-economic spectrum. Maltreatment was divided into types: physical abuse (7%), sexual abuse (9%), neglect (72%) and multiple maltreatment (12%).

The interviewers and participants were blind to the purpose of the study, to the inclusion of an abused and/or neglected group, and to the participants’ group membership. Participants were told that they had been selected as part of a large group of individuals who grew up in that area during the late 1960s and early 1970s. Institutional Review Board approval was obtained for the procedures and participants signed a consent form acknowledging that they understood the conditions of their participation and were participating voluntarily.

Assessment of Dysthymia, Major Depressive Disorder, and Alcohol Abuse Symptoms

Assessments of lifetime DSM-III-R symptoms of Dysthymia (DD), Major Depressive Disorder (MDD), and Alcohol Abuse (AA) were based on information from the NIMH Diagnostic Interview Schedule (DIS-III-R) [58]. The DIS-III-R is a highly structured interview designed for use by lay interviewers. Field interviewers received a week of study-specific training and successfully completed practice interviews before beginning the interviews. Field interviewer supervisors recontacted a random 10% of the respondents for quality control. Frequent contacts between field interviewers and supervisors were held to prevent interview drift, to monitor quality, and to provide continuous feedback. Adequate reliability for the DIS has been reported [59]. The DIS has been used in a variety of NIMH sponsored research projects that require psychiatric assessments of large numbers of subjects [60]. Number of symptoms reported by participants were used as outcomes and ranged from 0 to 9 for MDD and AA and 0 to 7 for DD: depression (M = 3.85, SD = 2.76), dysthymia (log transformed to correct high skew, M = 1.12, SD = .66) and alcohol abuse (log transformed to correct high skew, M = 1.00, SD = .81).

Assessment of MAOA

At mean age 41, DNA was obtained from usable blood samples from 617 study members of 631 who submitted DNA samples using the PureGene (Gentra Systems Inc) system according to the manufacturer’s instructions. The MAOA promoter polymorphism was genotyped by PCR amplification using primers MAOA-F2 5′-(TGCTCCAGAAACATGAGCAC) -3′ and MAOA-R2 5′-(GGACAGGCTGTAGGAGGTGTC) -3′. PCR reactions contained 80 ng of template DNA, 1.0 U AmpliTaq Gold polymerase (Applied Biosystems), 1.0 μM of each primer, 0.2 mM dNTP, 2.0 mM MgCl₂, and 2μl of GeneAmp 10x buffer II (Applied Biosystems), in a 20 μl volume. After 12 min at 96°C, 40 cycles were done at 96°C for 15 s, at 67.7°C for 20 s, and at 72°C for 30 s, followed by a final extension step at 72°C for 10 min. Products were resolved by Higher Resolution Microplate Array Diagonal Gel Electrophoresis [61], using an 8% polyacrylamide gel run at 150 V for 1hour and 15 minutes.

Representative genotypes were identified and sequenced using a Beckman-Coulter CEQ8000 semi-automated fluorescent sequencing system to confirm the sizes and number of repeats present in the observed alleles. These samples were included as size standards on subsequent gels. DNA was visualized by staining with ethidium bromide and gel images were captured and analyzed using Kodak 1D image analysis software. Any ambiguous genotypes that could not be resolved by repeat PCR and electrophoresis were determined by direct DNA sequencing. One male was identified as heterozygous for MAOA. Presence of a Y-chromosome was verified by PCR amplification of the marker DYS392. As it is unknown if this person has two copies of the MAOA gene (either through an XXY karyotype or a segmental duplication) or is mosaic for MAOA genotype, he was excluded from further analysis.

Overall, the frequencies of the MAOA alleles were 0.023 for 2 repeats, 0.407 for 3 repeats, 0.014 for 3.5 repeats, 0.550 for 4 repeats, and 0.006 for 5 repeats. Chi-square analysis revealed no evidence for deviation from Hardy-Weinberg Equilibrium (χ² = 7.9, df = 14, p=.89). There is good agreement about the levels of expression associated with the two most common allelic variants (3 and 4 repeats) of the functional promoter polymorphism in MAOA [29–31]. As these comprised over 95% of the alleles observed, we limited further analyses to these alleles

Individuals with genotypes other than 3- and 4-repeat allele combinations (n=41) were excluded from analyses because levels of expression associated with these are ambiguous [29–31]. Males with one or females with two copies of the 3-repeat allele were designated as low activity, males with one or females with two copies of the 4-repeat allele were designated as high activity[29–31]. Heterozygous females were also included as a separate group. The final sample consisted of 575 participants: 238 (41%) controls and 337 (59%) maltreated.

Data Analysis

To determine whether child maltreatment interacts with MAOA genotype to predict MDD, DD, and AA symptoms, we first conducted Ordinary Least Squares (OLS) regressions. The genotype variables were dummy coded (DC) into two variables [DC1 = low-activity genotype (1), others (0), and DC2 = heterozygous women (1), others (0)] and entered as main effects. Interaction terms were created following criteria set in Cohen et al. [62] by multiplying types of maltreatment by the dummy codes for genotype and entering the interaction terms separately.

To assess whether there were further differences in the child maltreatment by genotype interactions, two regressions assessed three-way interactions (MAOA genotype by type of maltreatment by sex and MAOA genotype by type of maltreatment by race), for each outcome (DD, MDD and AA).

Preliminary Analyses

To assess the representativeness of our sample, we compared the original sample to that analyzed here (n = 575) and found no differences in age, sex, race/ethnicity, and abuse/neglect status. We found a significant difference in the distribution of allele frequencies for Whites compared to Non-Whites (χ² = 19.66, df = 4, p = .001), although there was no difference between the groups (abuse/neglect versus control) (χ² = 4.56, df = 4, p = .33). Table S1 in the Supplement presents descriptive statistics. Analyses were run using SPSS, Version 19 [63].

Results

Study Findings

Table 1 shows that none of the maltreatment types interacted with MAOA genotype to predict DD, MDD or AA. Table 2 shows the results of regressions conducted to examine three way interactions between genotype, type of maltreatment, and sex predicting AA, DD, and MDD symptoms. For females only, childhood physical abuse and multiple maltreatment interacted with MAOA genotype to predict dysthymia symptoms, such that the homozygous 3, 3, low-activity allele, but not the homozygous 4, 4, high-activity allele, genpotype was protective for both types of maltreated females (see Figure 1). The heterozygous group tended to fall between the other two groups of females. MAOA genotype did not interact with any of the maltreatment types or in males.¹

Table 1.

Two-Way Interactions of MAOA Genotype with Maltreatment Types Predicting Dysthymia, Major Depressive Disorder and Alcohol Abuse Symptoms in Adulthood

	Dysthymiaa (N=571)	Major Depressive Disorder (N=573)	Alcohol Abusea (N=571)
Predictors	Beta	Beta	Beta
Age	.04	.02	.03
Sex (Female=1)	.19***	.19***	−.30***
Race (White=1)	.02	.05	.15***
Genotype DC 1 (3,3=1, others =0)	.11	.07	−.07
Genotype DC 2 (heterozygous =1, others=0)	.02	.02	−.09
Physical Abuse	.05	−.001	−.01
Sexual Abuse	.10	.05	.05
Neglect	.23**	.16*	.07
Multiple Maltreatment	.12†	.09	−.10
Physical Abuse by DC1	−.05	−.03	.02
Physical Abuse by DC2	.05	.09†	.04
Sexual Abuse by DC1	−.04	−.01	.04
Sexual Abuse by DC2	−.01	.02	.05
Neglect by DC1	−.10	−.04	−.03
Neglect by DC2	−.11	−.08	.11
Multiple Maltreatment by DC1	−.04	−.02	.06
Multiple Maltreatmetn by DC2	−.03	−.04	.09†
Adjusted R2	.03***	.06**	.13***

^aOutcomes are log transformed

^***p < .001;

^**p < .01;

^*p < .05;

^†p < .10

DC1= Dummy code of genotype (3,3=1, others=0); DC2=Dummy code of genotype (heterozygous=1; others =0)

Table 2.

Three-Way Interactions of MAOA Genotype, Type of Child Maltreatment, and Sex Predicting Dysthymia, Major Depressive Disorder, and Alcohol Abuse Symptoms

	Dysthymia Symptomsa (N=573)	Major Depressive Disorder Symptoms (N=573)	Alcohol Abuse Symptomsa (N=571)
Age	.05	.03	.03
Sex (Female=1)	.05	.14	−.34**
Race (White=1)	.01	.05	.15***
Physical Abuse	−.04	−.08	−.03
Sexual Abuse	.02	.03	−.06
Neglect	.14†	.11	.02
Multiple Maltreatment	.07	.06	−.08
DC1 (3,3=1, others =0)	.08	.09	−.05
DC2 (heterozygous =1, others=0)	.10	.06	−.07
Physical abuse by DC1	.05	.03	.05
Physical abuse by DC2	−.05	.004	.03
Sexual abuse by DC1	−.02	−.04	.02
Sexual abuse by DC2	−.05	.01	.03
Neglect by DC1	−.05	−.03	−.03
Neglect by DC2	−.23*	−.15	.05
Multiple maltreatment by DC1	.04	.02	.07
Multiple maltreatment by DC2	−.07	−.06	.09
Physical abuse by Sex	.21*	.18*	.03
Sexual abuse by Sex	.12	.13	.11
Neglect by Sex	.22t	.03	.13
Multiple Maltreatment by Sex	.09	.05	−.02
Sex by DC1	.07	−.03	−.03
Physical Abuse by DC1 by Sex	−.18*	−.12†	−.06
Sexual Abuse by DC1 by Sex	−.02	.06	.05
Neglect by DC1 by Sex	−.10	−.01	−.001
Multiple by DC1 by Sex	−.14*	−.09	−.04
Adjusted R2	.05**	.03*	.11**

Note:

^aOutcomes are log transformed

^***p < .001;

^**p < .01;

^*p < .05;

^†p < .10

DC1= Dummy code of genotype (3,3=1, others=0); DC2=Dummy code of genotype (heterozygous=1; others =0)

Figure 1.

Means of symptoms of dysthymia as a function of MAOA activity and a history of physical abuse and multiple maltreatment in females. MAOA is divided into 3,3 alleles, 4,4 alleles and 3,4 heterozygous. Sexual abuse and multiple maltreatment histories are based on documented court cases. Symptoms of dysthymia were log tranformed to correct for skew. Physically abused or multiply maltreated women with 4,4 high activity genotype were at significantly increased risk (p < .05 for both) for developing dysthymia symptoms than women with 3,3, low-activity genotype. Pattern was absent for non-maltreated females and males.

The results of regressions to test interactions between genotype, type of maltreatment, and race were significant for all three mental health outcomes (AA, DD, and MDD) (see Table 3). Childhood sexual abuse interacted with genotype and race to predict symptoms of DD, MDD and AA. The 3,3 or 3 allele genotype was protective against DD, MDD, and AA symptoms for sexually abused Whites, whereas the 4,4 or 4 allele genotype was protective for non-Whites (see Figure 2A–C).²

Table 3.

Three-Way Interactions of MAOA Genotype, Type of Child Maltreatment, and Race Predicting Major Depressive Disorder, Dysthymia, and Alcohol Abuse Symptoms

	Dysthymia Symptomsa (N=573)	Major Depressive Disorder Symptoms (N=573)	Alcohol Abuse Symptomsa (N=571)
Age	.02	.01	.00
Sex (Female=1)	.18**	.18**	−.30***
Race (White=1)	−.14	−.12	.14
Physical Abuse	.03	−.13	.20
Sexual Abuse	−.03	−.05	−.05
Neglect	.05	−.03	.16
Multiple Maltreatment	.13*	.11†	−.10
DC 1 (3,3=1, others =0)	.13	.04	−.09
DC 2 (heterozygous =1, others=0)	−.08	−.08	−.16
Physical abuse by DC1	−.03	−.04	.05
Physical abuse by DC2	.09	.18*	−.03
Sexual abuse by DC1	.11	.17*	.25**
Sexual abuse by DC2	.11	.13	.13
Neglect by DC1	−.09	.04	−.04
Neglect by DC2	−.07	−.06	.01
Multiple maltreatment by DC1	−.10	−.06	.001
Multiple maltreatment by DC2	−.07	−.06	.20*
Physical abuse by race	.04	.16	−.24†
Sexual abuse by race	.16t	.13	.14
Neglect by race	.25*	.27*	−.14
Multiple Maltreatment by race	.03	.02	−.14†
Race by DC1	−.04	.02	.03
Race by DC2	.12	.13	.11
Physical Abuse by DC1 by Race	N/A	N/A	N/A
Physical Abuse by DC2 by Race	−.06	−.11	.06
Sexual Abuse by DC1 by Race	−.19*	−.22*	−.27**
Sexual Abuse by DC2 by Race	−.15†	−.15†	−.11
Neglect by DC1 by Race	−.01	−.09	.01
Neglect by DC2 by Race	−.04	−.02	.13
Multiple by DC1 by Race	.06	.04	.09
Multiple by DC2 by Race	N/A	N/A	N/A
Adjusted R2	.05**	.05**	.13**

Notes:

^aOutcomes are log transformed

DC1= Dummy code of genotype (3,3 or 3 =1, others=0); DC2=Dummy code of genotype (3,4 =1; others =0)

^***p < .001;

^**p < .01;

^*p < .05;

^†p < .10

Figure 2.

Means of symptoms of dysthymia, major depressive disorder and alcohol abuse as a function of MAOA activity and a history of sexual abuse in Whites and Non Whites. MAOA is divided into 3,3or 3 alleles, 4,4 or 4 alleles and 3,4 heterozygous females. Sexual abuse histories are based on documented court cases. Symptoms of dysthymia and alcohol abuse were log tranformed to correct for skew. Sexually abused Whites with 4,4 or 4, high activity, genotype were at significantly increased risk (p < .05 for all) for developing dysthymia, major depressive and alcohol symptoms than sexually abused Whites with 3,3 or 3, low-activity, genotype. Sexually abused Non Whites with 3,3 or 3, low activity, genotype were at significantly increased risk (p < .05 for all) for developing dysthymia, major depressive and alcohol symptoms than sexually abused Non-Whites with 4,4 or 4, high-activity genotype.

Discussion

We did not find that childhood abuse and neglect interacted with MAOA genotype to predict symptoms of Dythymia, Major Depressive Disorder or Alcohol Abuse. However, participant sex and race moderated these relationships and we observed significant interactions between types of maltreatment and genotype for females, but not males, and reverse patterns of vulnerability for Whites and non-Whites.

For women, there were significant interactions in the prediction of Dysthymia, where the high-activity MAOA allele acted as a risk or vulnerability factor for physically abused and multiply maltreated females, consistent with other findings [50]. Thus, high levels of MAOA were associated with increased vulnerability for certain types of maltreatment (physical abuse and multiple) for dysthymia symptomatology as opposed to low levels that have been suggested to act as vulnerability for aggression [41]. It is interesting that depressive disorders have been treated by MAO inhibitors for decades, suggesting that reducing MAOA activity reduces depressive symptomatology [64]. Thus, one might expect that high levels of MAOA, not low levels, should be a risk factor for symptoms of depressive disorders. Research on MAOA activity outside the field of childhood maltreatment has shown some evidence suggestng that the high activity MAOA allele is a vulnerability factor. For example, depressed bereaved participants had more complicated grief symptoms with more MAOA activity than with less [65]. Meyer et al. [66] demonstrated increased MAOA density among depressed individuals compared to healthy controls, suggesting that higher MAOA activity is related to depressive symptomatology among clinically depressed individuals.

Our finding that MAOA moderates the relationship between childhood maltreatment and dysthymia only in females requires further comment. There are a number of possible reasons for the sex differences in the current study, including hormonal and gene interactions. The MAOA promoter interacts with both corticosteroids and androgens [67, 68] which show sex differences during perinatal development [69]. Additionally, there is evidence that during adolescence, MAOA may interact with testosterone which is differentially active for males and females [47]. Lastly, it is not clear whether females use both alleles of the MAOA genotype, located on the X-chromosome [70], and activation of both alleles may lead to sex differences in MAOA dosage. In the current study, heterozygous females’ symptoms tended to fall between the people with low-activity and high-activity MAOA genotype.

Thus, there are at least three possible ways to interpret the findings that the high activity allele combination may act as a vulnerability only for maltreated females and not males: (1) the effects of MAOA genotype on antisocial behavior and mood are different [50], (2) the direction of expression of MAOA genotype is different for males and females, and (3) these are chance findings. These explanations have different implications regarding the transcription of MAOA, suggesting either sex-based and/or disorder based variation or no variation at all. Based on the findings from this study, none of these explanations can be disqualified. Sex effects on the direction of MAOA genotype vulnerability for maltreated participants have been observed in relation to other mental health outcomes [48, 52, 71, 72]. Thus, it is likely that both sex-based and disorder-based differences exist.

We did not find sex differences in the way genotype moderates the relationship between child maltreatment and alcohol symptoms. Other studies have reported different sex effects in the prediction of alcoholism. Studies by Nillsson and colleagues [52–54] found that the interaction of MAOA and self-reported family functioning predicted adolescent alcohol related problem behaviors and excessive alcohol consumption. The direction of vulnerabilites differed for girls and boys and was stronger for girls. Nillsson and colleagues used retrospective self-reports of poor family functioning as a proxy for abuse and self-report of sexual abuse, clearly different from the operationalizations of abuse and neglect used here which involved official records of physical and sexual abuse and neglect. Thus, it is possible that the discrepancy between the current findings and past research is due to differences in design, participant age, criteria for maltreatment and analyses. It is also possible that these relationships may pertain only to certain types of alcoholism [73]. The Nillsson et al. study focused on the more antisocial type of alcoholism, whereas here we include both types.

We also found an interaction between participant’s race, childhood sexual abuse, and MAOA genotype in the prediction of dysthymia, major depressive disorder and alcohol abuse symptoms. The low-activity genotype acted as a protective factor for Whites, whereas the high-activity genotype acted as a protective factor for non-Whites. Interestingly, Widom and Brzustowicz (2006) examined race differences in the MAOA genotype by maltreatment interaction in the prediction of antisocial behavior using the same sample. They also observed race differences, but found a different pattern – significant interactions for Whites, but not for non-Whites. To our knowledge, no other studies have examined race differences.

Despite a number of strengths, certain limitations of this study should be noted. These findings are based on genetically-inferred (not verified) ethnicities, and this might have led to mistakes in ethnicity. The cases for the current study were drawn from court records in the Midwest of the United States and likely represent the more extreme cases of abuse and neglect. Results may not generalize to cases from other regions and maltreatment based on self-reports. Although the patterns of these results was consistent, because multiple statistical analyses were performed, the findings need to be replicated. In the context of the debate about the validity of findings of studies of gene by environment interactions, the quality of our sample, based on court-substantiated cases of abuse and neglect and matched controls, and the prospective follow-up, makes the possibility of spurious findings less likely. It is also possible that our failure to find significant interactions for males may have been due to the small sample size of certain types, i.e. sexual abuse, and limited power. Finally, gene-environment correlation, the notion that children with particular types of genotypes are more likely to be maltreated, cannot be ruled out here as a potential confound.

Nonetheless, these new findings provide evidence for the role of MAOA genotype as a moderator of the relationship between childhood maltreatment and mental health outcomes. Future research would benefit from replication and examination of other genotypes to further understand how genes and adverse environments play a role in these relationships.