Delinquency,Androgens,and the Family

Abstract

There is little doubt that family factors can influence involvement in delinquency, although the full nature and extent of their influences remain unclear. In recent decades, testosterone has been increasingly implicated as a contributor to adolescent offending. The present study sought to determine whether two important types of familial factors—parental socioeconomic status and amicable parent–child relationships—are interacting with testosterone (and possibly other androgens) to affect delinquency. A large sample of North American college students self-reported their involvement in eight categories of delinquency along with self-ratings of various androgen-promoted traits (e.g., muscularity and low-deep voice), parental social status, and the quality of the relationships they had with parents. In both sexes, parent–child relationships and androgens were significantly associated with delinquency but parental social status was not. Factor analysis revealed that the authors’ measures of all four categories of variables exhibited strong loadings onto their respective factors. Androgens and amicable parent–child relationships were associated with delinquency but parental social status was not. About one third of the influence of parent–child relationships on delinquency appeared to be attributable to androgens. Findings are discussed from the perspective of the evolutionary neuroandrogenic theory of delinquent and criminal behavior.

Keywords

delinquency social status parent–child relationships androgen-promoted traits sex differences testosterone evolutionary neuroandrogenic theory

Having good relationships with ones parents has been shown to be inversely associated with delinquency involvement (Canter, 1982; Cernkovich & Giordano, 1987; Haapasalo & Tremblay, 1994; Laub & Sampson, 1988; Patterson & Dishion, 1985; Stattin & Kerr, 2000; Wissink, Dekovic, & Meijer, 2006). One interpretation of these studies is that “[p]erhaps the closest thing to a general law of parenting is that warm, supportive, nurturant, and accepting behavior by mothers and fathers [are] associated with the development of social competence by adolescents” (Peterson, 2005, p. 40).

Others have suggested that the connection between favorable parent–child relationships and delinquency could actually be driven by genetics (Cleveland, Wiebe, Van Den Oord, & Rowe, 2000). In other words, genes contributing to good relationships between parents and children could also help to inhibit delinquency among the offspring (Rowe & Osgood, 1984). If so, relationships between parents and offspring could have little causal impact on offspring delinquency (Rowe, 1995).

Regarding other family variables, such as parental social status, being a contributor to delinquency, the findings have been mixed, at least regarding self-reported delinquency (Ellis & McDonald, 2001). Whereas some studies have found significant inverse correlations between delinquency and parental social status (e.g., Farrington & Painter, 2002; Graham & Bowling, 1995; Piquero & Tibbetts, 1999), others have concluded that the correlations are not statistically significant (Empey & Erickson, 1966; Lynam, Moffitt, & Stouthamer-Loeber, 1993; Miech et al., 1999; Riley & Shaw, 1985; Wiebe, 2006; see review by Ellis, Beaver, & Wright, 2009).

Over the past few decades, considerable research has pointed toward biological factors as contributing to delinquent and criminal behavior (Beaver, Wright, & DeLisi, 2007; Caspi, Lynam, Moffitt, & Silva, 1993; Walsh & Beaver, 2008). Among the most specific and frequently implicated biological factors are the so-called male sex hormones (collectively known as androgens), particularly testosterone (Booth & Osgood, 1993; Dabbs, 1990; Dabbs & Morris, 1990). A large number of studies have suggested that exposing the brain to high levels of testosterone increases offending rates, especially for violent crimes (Ellis, Das, & Buker, 2007; Rowe, Maughan, Worthman, Costello, & Angold, 2004; Sjöberg et al., 2007; van Bokhoven et al., 2006), although some failures to replicate have also been published (reviewed by Ellis et al., 2009).

Theoretical Background

In recent decades, the accumulation of evidence that both, biological and social, factors are interwoven into the tapestry of delinquency and crime causation has given rise to the biosocial perspective in criminology (Beaver et al., 2007; Beaver & Wright, 2005; Ellis & Walsh, 1997; Gabrielli & Mednick, 1984; Mednick, 1977; Nachshon, 1982; Raine, 2002; Ratchford & Beaver, 2009; Rowe, 1986). Within this perspective, a variety of specific theories have been proposed (e.g., Raine, 1993; Rowe, 1996; Walsh, 2008; Wilson & Herrnstein, 1985).

The most explicit and broad ranging biosocial theory so far is the evolutionary neuroandrogenic (ENA) theory (Ellis, 2003, 2005). A major impetus behind this theory was to explain why sex remains the single best predictor of criminality ever identified. The theory’s breadth was recently expanded by using it to account for not only criminal and antisocial behavior but also a number of traits predictive of criminality—such as risk taking, conduct disorders, learning difficulties, attention deficit hyperactivity disorder, and childhood aggression—which also all happen to exhibit substantial sex differences (Ellis, 2011).

Central to ENA theory is the evolutionary assumption that over countless generations females have enhanced their reproduction by biasing their mate choices toward loyal and capable provisioners of resources. Theoretically, males have devised various responses to this female bias. Some male responses essentially comply with female preferences, even if doing so involves skirting the law (such as engaging in shady business practices). Other male responses involve clearly violating criminal statutes to (a) acquire resources and/or (b) obtain mating opportunities with little or no resource investment. As a result, when the behavior of males and females are compared, males are not only more criminal, they are also more aggressive, competitive, prone to taking risks, and likely to exhibit a variety of psychopathic traits (Ellis, 2011). In other words, males have been favored by female mate choices to be more competitive and victimizing in their behavior, and criminality is a fairly common accompaniment of such general male behavioral tendencies (Ellis, 2005).

In addition to offering an evolutionary explanation for delinquent and criminal behavior, ENA theory identifies specific biological mechanisms by which competitive-victimizing behavior is produced. It asserts that all of these behavioral tendencies are enhanced by exposing the brain to high (male typical) levels of androgens, especially testosterone and its metabolites.

Neurological exposure to androgens occurs in two stages: a perinatal stage and a postpubertal stage. If brain exposure to androgens during the perinatal and postnatal stages is high (i.e., male typical), the brain will be highly masculinized/defeminized (Berenbaum, 1999; Hines, Brook, & Conway, 2004). Otherwise, the brain will remain in its “default” female-typical mode. As a result of high androgen exposure, male brains tend to be much more highly inclined to provision for resources than is true for female brains, even at the risk of retaliation by others seeking to do likewise.

Another stipulation of ENA theory is that for provisioning to be effective, considerable learning is required. Thus, one of the ways that family factors could help to reduce offspring delinquency is by facilitating the learning of legally acceptable provisioning skills. The other familial factor would involve genetic contributions to basic learning abilities. Thus, for environmental and genetic reasons, provisioning capabilities (including the extent to which they are antisocial in nature) should run in families.

Theoretically, especially following the adolescent surge in androgens, males should be more highly motivated than females to acquire resources. However, due to limited learning opportunities, many early competitive efforts will irritate others, reflected in the fact that these efforts have been declared unlawful in virtually all societies. The more capable adolescents are at learning, the sooner they should transition from “crude” to “sophisticated” forms of provisioning (the latter being far less likely to violate criminal statutes).

An assumption that is central to ENA theory is that competitive-victimizing behavior is genetically influenced. This assumption derives in part from evidence that (a) such behavior is androgen promoted (Ellis, 2011) and (b) androgen levels are genetically influenced (Harris, Vernon, & Boomsma, 1998; Hoekstra, Bartels, & Boomsma, 2006; Sluyter et al., 2000).

ENA theory leads one to expect such family variables as parental social status and harmonious interactions between offspring and parents will be associated with offspring delinquency. The reason is that intergenerational correlations in these traits are reflecting the influence of familial genes as well as social observation and imitation of parents by offspring. A recently published study by Wright, Beaver, Delisi, and Vaughn (2008) provided support for this line of reasoning by reporting that delinquency was only negligibly influenced by parenting behavior once genetic influences on offspring behavior were statistically controlled (also see Rowe, Rodgers, & Bushey, 1992). In other words, much of linkage between family variables and offspring delinquency is the result of shared genes rather than simply being due to the lack of affection, training, and guidance by parents.

Returning once again to the issue of genetics, twin studies of androgen levels have revealed that androgens are to a substantial degree regulated by genes (Harris et al., 1998; Hoekstra et al., 2006). If so, one can deduce that some of the same genes that help to promote amicable parent–child relationships and inhibit offspring delinquency are those that help to reduce/moderate testosterone levels.

Theoretically Derived Hypotheses

The present study was undertaken to partially test ENA theory. If the theory is correct, androgens should have a substantial impact on delinquency, even within each sex. Androgens should even affect how well parents and offspring relate to one another. Furthermore, there should be some measurable relationship between offspring delinquency, on one hand, and parental social status and the quality of parent–child relationships, on the other hand. However, if ENA theory is correct, relationships between parenting skills and offspring delinquency should diminish once the influence of androgens on delinquency is controlled. Therefore, the following hypotheses were tested:

Hypothesis 1 (H1): Within both genders, high androgen exposure should be positively correlated with involvement in delinquency, especially for violent offenses.

Hypothesis 2 (H2): Inverse correlations should exist between offspring involvement in delinquency and amicable parent–child relationships.

Hypothesis 3 (H3): Parental social status should be inversely correlated with offspring delinquency.

Hypothesis 4 (H4): Within both genders, statistically controlling for androgen exposure should substantially diminish the relationship between offspring delinquency and amicable parent–child relationships and parental social status.

Method

Sample

A questionnaire was completed by 3,782 male and 7,688 female college students at 20 U.S. universities and 2 Canadian universities as part of a larger research investigation. These students were recruited primarily from introductory psychology and sociology courses.

The exact procedures used in recruitment varied from one university to another, but usually involved giving a small amount of extra credit to students willing to complete a questionnaire. Any student who objected (which virtually none did) was allowed to read a course-relevant paper and write a brief report based on what he or she read for the same credit. Because the questionnaires were anonymous, extra credit was given to each student when he or she submitted a questionnaire contained within a blank sealed envelope. Overall, roughly 90% of those in courses in which the opportunity to participate in the study was offered completed a questionnaire.

Participants ranged in age from 18 to 56, with a mean of 22. Racially/ethnically, the participants were 85% White, 4% Black, 2% Native American, 2% Asian/Pacific Islander, 1% Hispanic, and the remaining 6% provided no answer.

Variable Measurement

Four factors were sought in the present study, three of which were treated as independent variables. These were androgens, amicable parent–child relationships, and parental social status. The dependent factor was involvement in delinquent behavior.

To measure delinquency, participants were presented with a list of eight offense categories, and asking them to report the number of times they had engaged in such behavior within the following three age ranges: 10 to 15, 16 to 18, and 19 to their present age. Because the participants’ ages varied considerably, but all were at least 18 years of age, the present analysis was limited to reported involvement in the eight offense categories during the first two age ranges (i.e., ages 10-18). The offense categories were as follows: serious violence (requiring medical treatment of the victim), less serious violence, vehicle theft, theft more generally, vandalism, illegal entry, illegal drug use, and illegal commerce (mainly sale of drugs).

Bodily exposure to androgens was assessed by asking participants to rate themselves on eight traits (nine in the case of males), each based on a 100-point scale. The traits were masculine mannerisms, masculine body appearance, physical strength, strength of sex drive, low/deep voice, upper body strength, lower body strength, amount of body hair, and penis size (males only).

Numerous studies have shown all these traits to be enhanced by exposing the body to testosterone or other androgens perinatally and postpubertally (e.g., Baskin et al., 1997; Bhasin, 2003; Boas et al., 2006; Goldstat, Briganti, Tran, Wolfe, & Davis, 2003; Perry, Miller, Patrick, & Morley, 2000; Persky et al., 1982; Schroeder et al., 2005; Seidman & Klein, 2004; Traish, 2009; Wang, Swerdloff, & Iranmanesh, 2000). Even though the measures of androgens used in the present study were indirect, they are likely to be more reflective of the long-term combined influence of perinatal and postpubertal androgen exposure than measures based on blood or saliva samples (because blood and saliva samples only assay testosterone at one point in time; see Dabbs, 1990).

Seven parental social status measures were obtained. Participants were asked to report the years of education for both parents, and their occupation if gainfully employed (which was coded on a 6-point scale ranging from high to low). They were also asked to estimate their family income from “extremely high” to “extremely low” on a 100-point scale. The last two status measures involved asking each participant to rate his or her family’s overall social status during their childhood and during their adolescent years on a 100-point scale.

Assessing the quality of parent–offspring relationships involved asking participants to make six ratings, three regarding each parent. Regarding their father, participants rated how well they and their father got along during childhood and during adolescence and how well they enjoyed being in their father’s company. The same three questions were asked with reference to the mother as well, again on 100-point scales.

Factor Analysis

To assess the degree to which our variables conformed to the four factors of interest, a factor analysis was performed on all 30 variables using principal components analysis. As shown in Table 1, four factors were extracted after being subjected to varimax rotation. The first factor clearly encompasses all eight types of delinquency. The only loading that was less than .50 (and even then only slightly below) was for illegal drug use.

Table 1.

Factor Loadings for the Three Groups of Independent Variables and One Group of Dependent Variables.

	Factors
Variable groupings	1	2	3	4
Delinquency measures
Serious violence	.716	.086	.112	−.027
Minor violence	.618	.086	.064	.005
Vehicle theft	.822	−.023	.064	−.015
Thefts in general	.768	.001	−.033	.007
Vandalism	.802	.030	−.073	.040
Illegal entry	.807	.025	.010	−.009
Illegal drugs use	.487	.041	−.262	.043
Illegal commerce	.786	.006	−.152	.067
Androgen-promoted trait measures
Masculine mannerisms	.047	.796	.073	.037
Masculine body appearance	.003	.827	.012	−.007
Overall physical strength	.056	.669	.046	.149
Strength of sex drive	.025	.498	.039	.148
Lower body strength	−.005	.591	−.061	.043
Upper body strength	.041	.675	.088	.084
Low-deep voice	.037	.512	.098	.035
Body hair	−.006	.424	.076	−.034
Penis size (males only)	.029	.681	.028	.022
Amicability of parent–child relationship measures
Relationship with father during childhood	−.021	.083	.736	.092
Relationship with father during adolescence	−.018	.095	.714	.103
Relationship with mother during childhood	−.013	−.016	.714	.044
Relationship with mother during adolescence	1.037	−.010	.694	.098
Enjoyed father’s company	−.019	.183	.617	.119
Enjoyed mother’s company	−.020	.001	.597	.089
Parental social status measures
Family income	−.010	.078	.103	.737
Father’s years of education	.040	.012	.057	.749
Father’s occupation	.012	.064	.123	.790
Mother’s years of education	.047	.061	.075	.687
Mother’s occupation	.015	.088	.123	.676
Social status during childhood	−.011	.112	.451	.401
Social status during adolescence	−.029	.120	.436	.513

Note: The highest loadings for each variable are bolded.

Table 1 shows that the second factor to emerge had to do with all eight androgen-influenced measures (nine in the case of males). Only two of these measures had loadings of less than .50; they were strength of sex drive and amount of body hair. In all likelihood, the weak loading of body hair reflects the fact that such hair is primarily regulated by the androgen, androstenediol, whereas all of the other traits are mainly influenced directly by testosterone (Root, 1973). When we performed factor analysis separately by sex, we found that the loadings were strong for males but weak for females. This can be attributed to the fact that whereas testosterone is largely responsible for the sex drive in males, in females, other hormones such as estradiol also appear to be involved (Davis & Tran, 2001; DeCherney, 2004).

The strongest loadings on the second factor shown in Table 1 involved masculine body build and masculine mannerisms. Factor analyses run separately for males and females revealed that the same two traits retained strong loadings.

Regarding the variables used to measure the study’s social factors, Table 1 shows that two remaining factors clearly emerged. Six variables had to do with parent–child amicability (the third factor) and five clearly involved parental social status. However, the last two social variables, which were included in our analysis to measure social status, actually shared as much variance with the third factor (i.e., amicability) as with the fourth factor (status). We are unable to account for why this was so, but decided not to remove them from subsequent analyses as they were still indicative of potentially pertinent family factors.

Statistical Analysis

Beyond factor analysis, the analytic procedures of this study concentrated on finding evidence linking the social and biological variables to offspring delinquency. Then a structural equation model (SEM) was used to assess the separate and combined influence of the independent variables on the dependent variables. To ensure that gender was not a confounding variable when applying the SEM to the relationships under study (especially the androgen-promoted traits), the model was applied separately to each gender.

Results

The first hypothesis (H1) asserted that androgen-promoted traits will be positively correlated with involvement in delinquency, especially for violent offenses. Table 2 presents the Pearson correlations between the eight androgen-promoted traits (nine in the case of males) and the eight delinquency measures for both genders. One can see that the vast majority of these relationships are statistically significant and positive. In addition, as anticipated, the strongest correlations involved violent offenses for males.

Table 2.

Correlations Between Androgen-Promoted Physiological Traits and Self-Reported Delinquency Among Males (in Bold) and Among Females (in Italics).

Measures of androgen-promoted traits	Sex	Serious violence	Less serious violence	Vehicle theft	Theft	Vandalism	Illegal entry	Illegal drugs	Illegal commerce
Masculine mannerisms	Males	.064 **	.061 **	.025	.004	.052 **	.023	−.029	−.003
Masculine mannerisms	Females	.027*	.062**	.046**	.059**	.043**	.040**	.069**	.064**
Masculine body build	Males	.040*	.038*	−.002	−.017	.029 **	−.004	−.001	−.015
Masculine body build	Females	.063**	.055**	.043**	.037**	.042**	.026*	.026*	.045**
Physical strength	Males	.104 **	.084 **	.030	.038 *	.025	.033 *	.009	.020
	Females	.045**	.059**	.028*	.050**	.019	.038**	.024	.016
Strength of sex drive	Males	057 **	.061 **	.022	.028	.043 **	.026	.023	.018
Strength of sex drive	Females	.021	.028*	.007	.032**	.023*	.016	.042**	.016
Low-deep voice	Males	.059 **	.064 **	.027	.026	.020	.015	.031	.004
	Females	.012	.038**	.051**	.063**	.055**	.039**	.113**	.090**
Upper body strength	Males	.098 **	.106 **	.030	.045*	.077 **	.054 **	.033	.015
Upper body strength	Females	.026	.043**	.023	.012	.017	.024	.036	.056**
Lower body strength	Males	.047*	.072 **	.012	.014	.047*	.029	−.071	−.012
Lower body strength	Females	.004	.033*	.033*	.019	.030*	.034*	.043**	.048**
Amount of body hair	Males	.066 **	.048*	.022	−.026	.012	.014	−.029	−.006
Amount of body hair	Females	.035**	.033*	.031*	.031*	.053**	.027*	.002	.008
Penis size	Males	.074 **	.066 **	.016	.003	.042*	.007	.006	−.010

p < .05. **p < .01.

The second hypothesis (H2) was that inverse relationships should exist between offspring self-reported involvement in delinquency and amicable parent–child relation relationships. Pertinent results appear in Table 3. As one can see, the hypothesis was largely supported. Thus, as most prior literature has shown, offspring who had amicable relationships with their parents while growing up were less involved in delinquency than offspring with less amicable relationships. The importance of amicable parent–child relationships generally held for both genders for all forms of delinquency but especially in the case of illegal drug use and illegal commerce (usually drug sale).

Table 3.

Correlations Between Parent–Child Amicability Measures and Self-Reported Delinquency for Males (in Bold) and Females (in Italics).

Parent–child relations measures	Sex	Serious violence	Less serious violence	Vehicle theft	Theft	Vandalism	Illegal entry	Illegal drugs	Illegal commerce
Relationship with father during child	Males	−.052*	−.040*	−.050*	−.091 **	−.059 **	−.069 **	−.125 **	−.114 **
Relationship with father during child	Females	−.036*	−.021	−.001	−.043**	−.022	−.006	−.061**	−.112**
Relationship with father during adolescence	Males	−.042*	−.032	−.015	−.069 **	−.042*	−.051*	−.112 **	−.153 **
Relationship with father during adolescence	Females	−.034*	−.028*	.000	−.054**	−.048**	−.013	−.094**	−.162
Relationship with mother during child	Males	−.029	−.015	−.062 **	−.053*	−.054 **	−.019	−.109 **	−.118 **
Relationship with mother during child	Females	−.013	−.020	−.019	−.063**	−.037**	−.019	−.085*	−.137**
Relationship with mother during adolescence	Males	.052 *	−.052 *	−.027	−.088 **	−.059	−.042*	−.148 **	−.195 **
Relationship with mother during adolescence	Females	.006	−.023*	−.019	−.090**	−.042**	−.007	−.121**	−.211**
Enjoy being with father	Males	.003	−.008	−.013	−.046	−.039*	−.025	−.064 **	−.076 **
Enjoy being with father	Females	−.033**	−.002	.004	.001	−.010	−.004	−.030*	−.040**
Enjoy being with mother	Males	.016	.012	−.013	−.005	−.031	.009	−.028	−.095 **
Enjoy being with mother	Females	.008	.007	−.011	−.047**	−.019	−.019	−.059**	−.092**

p < .05. **p < .01.

According to the third hypothesis (H3), there should be an inverse correlation between parental social status and offspring involvement in delinquency. By examining Table 4, one can see that very few significant negative correlations exist. The main exception involved “overall status during childhood” and “overall status during adolescence” being inversely correlated with illegal drug use and illegal commerce. As noted earlier in discussing Table 1, these latter two parental status variables actually loaded as much on amicable parent–child relationships as on parental social status. Overall, contrary to H3, there appears to be no significant relationship between parental social status and offspring delinquency.

Table 4.

Correlations Between Parental Social Status and Self-Reported Delinquency for Males (in Bold) and Females (in Italics).

Parental social status measures	Sex	Serious violence	Less serious violence	Vehicle theft	Theft	Vandalism	Illegal entry	Illegal drugs	Illegal commerce
Father’s years of education	Males	.004	.036	.030	.036	.059*	.006	.043	.026
	Females	.018	−.015	.017	.009	−.007	.002	−.021	−.014
Mother’s years of education	Males	.006	.046*	.057*	.026	.037	.015	.022	.010
	Females	.017	.001	.024	.015	.002	.000	−.010	.008
Father’s occupation	Males	.005	.025	.025	.030	.047*	−.014	.034	.020
	Females	.027*	−.004	.018	.005	−.014	−.009	−.011	−.006
Mother’s occupation	Males	.014	.035	.036	.029	.032	−.003	.016	−.002
	Females	.033*	.008	.001	.009	−.017	−.007	−.009	.000
Family income	Males	.017	.023	.019	−.018	.019	−.019	.010	−.025
	Females	−.006	−.028*	.014	−.002	−.009	−.007	−.027	−.003
Overall status during child	Males	.031	.024	.028	−.022	−.031	−.027	−.103 **	−.065*
	Females	.020	.010	.010	−.014	.024	.016	−.058**	−.033*
Overall status during adolescence	MalesFemales	.033.007	.014 −.005	.028 −.019	−.012−.049**	.023.010	−.018.009	−.079 −.076	−.039−.074**

p < .05. **p < .01.

The fourth hypothesis (H4) derived from Ellis’s (2005) ENA theory asserts that brain exposure to androgens not only elevates the probability of offspring delinquency but also has detrimental effects on parent–child relationships. To test this hypothesis, we consolidated our four categories of variables into the following four factors: the eight delinquency measures, the six parent–child amicability measures, the seven parental social status measures, and the eight androgen-influence measures (nine in the case of males). Then we performed a bivariate correlation between the consolidated delinquency factor and the remaining three factors. This was following by our performing a partial correlation between delinquency and parent–child amicability and parental social status with the androgen factor controlled.

As one can see in Table 5, for both genders, bivariate analysis revealed that the overall androgen factor was positively correlated with delinquency, and parent–child relations were negatively correlated with delinquency. As in our prior analysis, parental social status was not significantly related to delinquency. When one examines the partial correlations, one sees that controlling for androgen exposure reduced the correlations between delinquency and parent–child relationships for males and females by roughly one fourth. Such a reduction supports H4 by indicating that androgen exposure is not only contributing positively to delinquency but also negatively to amicable parent–child relationships.

Table 5.

Bivariate and Partial Correlations Between the Total of All Eight Delinquent Acts and the Totals of All Three Independent Variables (Androgen-Promoted Traits, Amicable Parent–Child Relations, and Parental Social Status).

		Independent factors
Offspring delinquency	Gender	Androgen influences	Parent–child relations	Parental SES
Bivariate correlations	Males	.077**	−.112**	.024
Bivariate correlations	Females	.133**	−.099**	−.015
Partial correlations with androgen influences removed	Males	Controlled	−.086**	.017
Partial correlations with androgen influences removed	Females	Controlled	−.074**	−.028

Note: SES = socioeconomic status.

p < .05. **p < .01.

Conclusions

This study sheds light on how social and biological factors might both contribute to delinquency. Although our measurements of both factors, along with delinquency itself, were all derived from self-reports, our analysis suggests that exposure to androgens increases and amicable parent–child relationships decreases delinquent behavior. Such a conclusion supports a recently proposed theory that brain exposure to testosterone and other “male sex hormones” may be important in explaining antisocial behavior (Ellis, 2005).

According to the theory, androgens modify the brain (perinatally and postpubertally) so as to enhance resource-procuring tendencies. Theoretically, resource procurement is promoted primarily through increased aggression, competitiveness, and risk taking (collectively known as competitive-victimizing behavior). As male brains are exposed to higher levels of androgens than the brains of females, males should exhibit more competitive-victimizing behavior, especially following the onset of puberty. ENA theory asserts that delinquency (especially of a victimizing nature) constitutes a crude expression of resource-procuring tendencies. As individuals mature, especially if they have a high ability to learn, they will quickly transition away from these crude resource-procuring tendencies toward forms that are usually within legal boundaries.

According to ENA theory, brain exposure to androgens should not only enhance delinquent behavior but also have adverse effects on parent–child relationships. This latter deduction follows from noting that high levels of aggression, competitiveness, and risk-taking work against offspring obedience toward parents. Along these lines, Booth and Osgood (1993) presented evidence that high androgen exposure makes offspring harder to manage and effectively discipline. Overall, ENA theory leads one to expect that (a) delinquency and amicable parent–child relationships should be inversely correlated and (b) if androgen exposure were to be statistically controlled, the inverse correlation should be reduced.

It is worth noting that the amicability of parent–child relationships could actually be reflecting genetic influences. Such a possibility comes from evidence that personality traits such as patience, friendliness, obedience, and agreeableness are all substantially heritable (Bergeman et al., 1993; Goldsmith, 1983). If so, genes for these personality traits could be largely responsible for why harmonious parent–child relationships vary a great deal. Such a line of reasoning could explain why Wright et al. (2008) found that most of the influences parents seem to have on offspring delinquency could actually be mediated by intergenerational genetic influences on personality traits. If genes are involved and ENA theory is correct, some of the contributing genes could be ones controlling androgen production throughout development. The reasonableness of this line of reasoning is bolstered by reiterating a point made in the introduction that genes help regulate testosterone levels throughout life (Harris et al., 1998; Hoekstra et al., 2006; Sluyter et al., 2000). Such reasoning could also help to explain why the onset of puberty (or biological age) appears to be a more important contributor to delinquency than chronological age (Piquero & Brezina, 2001).

In the case of correlations between parental social status and delinquency, ENA theory suggests that the relationship should be minimal, especially among college students, even though a negative relationship between social status and adult criminality should be substantial. The reason is that the vast majority of college students emanate from middle to upper status households. Theoretically, most of these individuals will exhibit some delinquency, especially if they are males. However, it is mainly noncollege bound males of lower status backgrounds that ENA theory predicts will be most likely to offend at high rates.

Another noteworthy element of ENA theory is that it applies to both sexes. In other words, even though the theory predicts that delinquency—as well as competitive-victimizing behavior in general—will be greater in males than in females, the correlations between androgen exposure and these behavior traits should be the same for both sexes. To determine whether this latter prediction was accurate, all of our derived hypotheses were tested for males and females separately, and, as expected, the androgen-offending correlations were more or less equally strong for both sexes. This means that it is unnecessary to consider any gender-related social influences on delinquency to account for male–female differences in such behavior.

The theoretical predictions were investigated using questionnaire responses provided by a large sample of North American research participants. The participants reported their involvement in eight categories of delinquency. The independent variables consisted of seven measures of parental social status, six indicators of how well they and their parents got along, and eight (or nine in the case of males) measures of long-term androgen exposure. Although it would have been preferably to have had independent verification of all of these traits, especially those having to do with androgen exposure, we believe that the self-reported data still provide a minimum estimate of the true effects of each of these traits.

Results were largely consistent with theoretical predictions: Nearly all of the measures of androgen exposure were positively correlated with all eight forms of delinquency for males and for females to significant degrees. In addition, as expected, amicable parent–child relationships negatively correlated with delinquency involvement. Furthermore, when we statistically controlled the presumed effects of androgens on delinquency, the remaining negative correlation between delinquency and amicable parent–child relations were reduced, although not eliminated, in both sexes.

Regarding the issue of parental social status, the present study confirmed numerous prior studies that parental social status and self-reported delinquency are not significantly correlated (Lynam et al., 1993; Riley & Shaw, 1985; Thornberry, & Farnworth, 1982; reviewed by Ellis et al., 2009). If ENA theory is correct, this lack of a correlation reflects the fact that within each gender, juvenile offending tendencies should exhibit few social status variations. It is only among adults that significant negative correlations should exist between offending and social status (whether of one’s parents or oneself). Another factor worth mentioning here is that persons who commit the fewest adolescent offenses tend to be more conscientious in reporting each one of them than persons who violate the law the most (Empey & Erickson, 1966; Farrington, Loeber, Loeber, Kammen, & Schmidt, 1996).

During the review process, several questions were raised about the validity of inferring high exposure to androgens using self-reported information about muscularity, voice timber, and physical strength (as well as penis size in the case of males). We recognize that it is a novel approach to androgen measurement and concede that validation research is warranted. However, we hypothesize that the measurement of physical traits such as musculature and voice timber will actually surpass measuring testosterone in blood or saliva samples (e.g., Dabbs, 1990; Halpern, Udry, Campbell, & Suchindran, 1993) as being valid long-term indicators of testosterone exposure. Of course, as noted earlier, we recognize that actual physical measurements of these androgen-influenced traits will be preferable to the simple self-report measures herein utilized.

Support for the above arguments come from studies showing that exposing the brain to androgens prenatally and again following puberty enhances spatial reasoning (Hier & Crowley, 1982; Puts, McDaniel, Jordan, & Breedlove, 2008). Castration experiments with laboratory animals (Isgor & Sengelaub, 1998) have also shown that brain exposure to testosterone helps to enhance spatial reasoning in adulthood. Despite this evidence, a recent study of healthy men and women found no within-sex correlation between circulating saliva testosterone levels and spatial reasoning (Puts et al., 2010). Rather than casting doubt on the hypothesis that androgens promote spatial reasoning, this latter finding simply suggests that obtaining one-time testosterone samples is a poor indicator of the long-term effects of androgens on spatial reasoning.

Along similar lines, most studies have reported significant positive correlations between circulating testosterone levels and aggressive behavior (independent of their illegality), but some studies have found no significant relationship (reviewed by Ellis et al., 2009). A problem with these studies is that nearly all of them simply used one-time assays of circulating testosterone as their measure of testosterone exposure. We submit that these assays are not reliable proxies for the long-term effects of testosterone exposure.

Even more relevant to the present investigation, more than 30 studies have investigated testosterone’s relationship to various forms of criminal and delinquent behavior, nearly all based again on one-time saliva or blood samples (reviewed by Ellis et al., 2009). Although more than 80% of these studies have revealed significant positive correlations, quite a few have not. If one focuses only on studies of delinquency, the evidence seems to become even more contradictory. Two studies reported a significant positive correlation (Banks & Dabbs, 1996; Mattsson, Schalling, Olweus, Law, & Svensson, 1980), two found no relationship of any significance (Inoff-Germain et al., 1988; Susman et al., 1987), and one actually reported a significant negative correlation (Schaal, Tremblay, Soussignan, & Susman, 1996). We believe that because all of these latter studies were based on one-time samples of testosterone, the likelihood of their identifying the “true”—meaning long term—testosterone-offending relationship is quite dubious.

Overall, findings from this study support the view that social environmental variables are not acting alone to affect delinquent behavior; instead, biological factors are also operating (Walsh & Beaver, 2008). The biosocial theory that was focused on in this study emphasizes that sex hormones are the most crucial biological factors promoting antisocial behavior (Ellis, 2003, 2005). The theory explains delinquency as largely an expression of competitive and risk-taking behavior promoted by exposing the brain to testosterone.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Banks

Dabbs

J. M.

Jr. (1996). Salivary testosterone and cortisol in a delinquent and violent urban subculture. Journal of Social Psychology, 136, 49-56.

Baskin

Sutherland

DiSandro

Hayward

Lipschutz

Cunha

(1997). The effect of testosterone on androgen receptors and human penile growth. Journal of Urology, 158, 113-1118.

Beaver

K. M.

Wright

J. P.

(2005). Biosocial development and delinquent involvement. Youth Violence and Juvenile Justice, 3, 168-192.

Beaver

K. M.

Wright

J. P.

Delisi

(2007). Self-control as an executive function. Criminal Justice and Behavior, 34, 1345-1361.

Berenbaum

S. A.

(1999). Effects of early androgens on sex-typed activities and interests in adolescents with congenital adrenal hyperplasia. Hormones and Behavior, 35, 102-110.

Bergeman

Chlpuer

Plomin

Pedersen

McClearn

Nesselroade

McCrae

(1993). Genetic and environmental effects on openness to experience, agreeableness, and conscientiousness: An adoption/twin study. Journal of Personality, 61, 159-179.

Bhasin

(2003). Regulation of body composition by androgens. Journal of Endocrinological Investigation, 26, 814-822.

Boas

Boisen

K. A.

Virtanen

H. E.

Kaleva

Suomi

A. M.

Schmidt

I. M.

(2006). Postnatal penile length and growth rate correlate to serum testosterone levels: A longitudinal study of 1962 normal boys. European Journal of Endocrinology, 154, 125-129.

Booth

Osgood

D. W.

(1993). The influence of testosterone on deviance in adulthood: Assessing and explaining the relationship. Criminology, 31, 93-117.

10.

Canter

R. J.

(1982). Family correlates of male and female delinquency. Criminology, 20, 149-167.

11.

Caspi

Lynam

Moffitt

T. E.

Silva

P. A.

(1993). Unraveling girls’ delinquency: Biological, dispositional, and contextual contributions to adolescent misbehavior. Developmental Psychology, 29, 19-30.

12.

Cernkovich

S. A.

Giordano

P. C.

(1987). Family relationships and delinquency. Criminology, 25, 295-321.

13.

Cleveland

H. H.

Wiebe

R. P.

Van Den Oord

E. J. C. G.

Rowe

D. C.

(2000). Behavior problems among children from different family structures: The influence of genetic self-selection. Child Development, 71, 733-751.

14.

Dabbs

J. M.

Jr. (1990). Salivary testosterone measurements: Reliability across hours, days, and weeks. Physiology & Behavior, 48, 83-86.

15.

Dabbs

J. M.

Jr. Morris

(1990). Testosterone, social class, and antisocial behavior in a sample of 4,462 men. Journal of Psychological Science, 1, 209-211.

16.

Davis

S. R.

Tran

(2001). Testosterone influences libido and well-being in women. Trends in Endocrinology and Metabolism, 12, 33-37.

17.

DeCherney

A. H.

(2004). Hormone receptors and sexuality in the human female. Journal of Women’s Health & Gender-Based Medicine, 9, 9-13.

18.

Ellis

(2003). Biosocial theorizing and criminal justice policy. In Somit

Peterson

(Eds.), Human nature and public policy: An evolutionary approach (pp. 97-120). New York, NY: Palgrave.

19.

Ellis

(2005). A theory explaining biological correlates of criminality. European Journal of Criminology, 2, 287-315.

20.

Ellis

(2011). Identifying and explaining apparent universal sex differences in cognition and behavior. Personality and Individual Differences, 51, 552-561.

21.

Ellis

Beaver

Wright

(2009). Handbook of crime correlates. Amsterdam, Netherlands: Elsevier.

22.

Ellis

Das

Buker

(2007). Androgen-promoted physiological traits and criminality: A test of the evolutionary neuroandrogenic theory. Personality and Individual Differences, 44, 701-711.

23.

Ellis

McDonald

J. N.

(2001). Crime, delinquency, and social status. Journal of Offender Rehabilitation, 32, 23-52.

24.

Ellis

Walsh

(1997). Gene based evolutionary theories in criminology. Criminology, 35, 229-276.

25.

Empey

L. T.

Erickson

M. L.

(1966). Hidden delinquency and social status. Social Forces, 44, 546-554.

26.

Farrington

D. P.

Loeber

M. S.

Kammen

W. B.

Schmidt

(1996). Self-reported delinquency and a combined delinquency seriousness scale based on boys, mothers and teachers: Concurrent and predictive validity for African Americans and Caucasians. Criminology, 34, 493-517.

27.

Farrington

D. P.

Painter

K. A.

(2002). Gender differences in offending: Implications for risk-focused prevention. London, England: Home Office Research Study.

28.

Gabrielli

W. F.

Jr. Mednick

S. A.

(1984). Urban environment, genetics, and crime. Criminology, 22, 645-652.

29.

Goldsmith

(1983). Genetic influences on personality from infancy to adulthood. Child Development, 54, 331-355.

30.

Goldstat

Briganti

Tran

Wolfe

R. R.

Davis.

S. R.

(2003). Transdermal testosterone therapy improves well-being, mood, and sexual function in premenopausal women. Journal of the North American Menopause Society, 10, 390-398.

31.

Graham

Bowling

(1995). Young people and crime. Home Office Research Study No. 145. London, England: Home Office.

32.

Haapasalo

Tremblay

R. E.

(1994). Physically aggressive boys from ages 6 to 12: Family background, parenting behavior, and prediction of delinquency. Journal of Consulting and Clinical Psychology, 62, 1044-1052.

33.

Halpern

C. T.

Udry

J. R.

Campbell

Suchindran

(1993). Relationships between aggression and pubertal increases in testosterone: A panel analysis of adolescent males. Social Biology, 40, 8-24.

34.

Harris

J. A.

Vernon

P. A.

Boomsma

D. I.

(1998). The heritability of testosterone: A study of Dutch adolescent twins and their parents. Behavior Genetics, 28, 165-171.

35.

Hier

D. B.

Crowley

W. F.

Jr. (1982). Spatial ability in androgen-deficient men. New England Journal of Medicine, 306, 1202-1205.

36.

Hines

Brook

Conway

(2004). Androgen and psychosexual development: Core gender identity, sexual orientation, and recalled childhood gender role behavior in women and men with congenital adrenal hyperplasia (CAH). Journal of Sex Research, 35, 75-81.

37.

Hoekstra

R. A.

Bartels

Boomsma

D. I.

(2006). Heritability of testosterone levels in 12-year-old twins and its relation to pubertal development. Twin Research and Human Genetics, 9, 558-565.

38.

Inoff-Germain

Arnold

G. S.

Nottelmann

E. D.

Susman

E. J.

Cutler

G. B.

Jr. Chrousos

G. P.

(1988). Relations between hormone levels and observational measures of aggressive behavior of young adolescents in family interactions. Developmental Psychology, 24, 129-139.

39.

Isgor

Sengelaub

D. R.

(1998). Prenatal gonadal steroids affect adult spatial behavior, CA1 and CA3 pyramidal cell morphology in rats. Hormones and Behavior, 34, 183-198.

40.

Laub

J. H.

Sampson

R. J.

(1988). Unraveling families and delinquency: A reanalysis of the Gluecks’ data. Criminology, 26, 355-380.

41.

Lynam

D. R.

Moffitt

T. E.

Stouthamer-Loeber

(1993). Explaining the relation between IQ and delinquency: Class, race, test motivation, school failure, or self-control? Journal of Abnormal Psychology, 102, 187-196.

42.

Mattsson

Schalling

Olweus

Law

Svensson

(1980). Plasma testosterone, aggressive behavior, and personality dimensions in young male delinquents. Journal of the American Academy of Child Psychiatry, 19, 476-490.

43.

Mednick

S. A.

(1977). A bio-social theory of the learning of law-abiding behavior. In Mednick

S. A.

(Ed.), Biosocial bases of criminal behavior (pp. 1-8). New York, NY: Taylor & Francis.

44.

Miech

R. A.

Caspi

Moffitt

T. E.

Wright

Bradley

Silva

P. A.

(1999). Low socioeconomic status and mental disorders: A longitudinal study of selection and causation during young adulthood. American Journal of Sociology, 104, 1096-1131.

45.

Nachshon

(1982). Toward biosocial approaches in criminology. Journal of Social and Biological Systems, 5, 1-9.

46.

Patterson

G. R.

Dishion

T. J.

(1985). Contributions of families and peers to delinquency. Criminology, 23, 63-79.

47.

Perry

H. M.

Miller

D. M.

Patrick

Morley

J. E.

(2000). Testosterone and leptin in older African-American men: Relationship to age, strength, function, and season. Metabolism, 49, 1085-1091.

48.

Persky

Dreisbach

Miller

W. R.

O’Brien

C. P.

Khan

M. A.

Lief

H. I.

. . . Strauss

(1982). The relation of plasma androgen levels to sexual behaviors and attitudes of women. Psychosomatic Medicine, 44, 305-319.

49.

Peterson

G. W.

(2005). Family influences on adolescent development. In Gullotta

T. P.

Adams

G. R.

(Eds.), Handbook of adolescent behavioral problems (pp. 27-55). New York, NY: Springer.

50.

Piquero

A. R.

Brezina

(2001). Testing Moffitt’s account of adolescent limited delinquency. Criminology, 39, 353-370.

51.

Piquero

A. R.

Tibbetts

(1999). The impact of pre/perinatal disturbances and disadvantaged familial environment in predicting criminal offending. Studies on Crime and Crime Prevention, 8, 52-70.

52.

Puts

D. A.

McDaniel

M. A.

Jordan

C. L.

Breedlove

S. M.

(2008). Spatial ability and prenatal androgens: Meta-analyses of congenital adrenal hyperplasia and digit ratio (2D:4D) studies. Archives of Sexual Behavior, 37, 100-111.

53.

Puts

D. A.

Cárdenas

R. A.

Bailey

D. H.

Burriss

R. P.

Jordan

C. L.

Breedlove

S. M.

(2010). Salivary testosterone does not predict mental rotation performance in men or women. Hormones and Behavior, 58, 282-289.

54.

Raine

(1993). The psychopathology of crime: Criminal behavior as a clinical disorder. San Diego, CA: Academic Press.

55.

Raine

(2002). Biosocial studies of antisocial and violent behavior in children and adults: A review. Journal of Abnormal Child Psychology, 30, 311-326.

56.

Ratchford

Beaver

K. M.

(2009). Neuropsychological deficits, low self-control, and delinquent involvement. Criminal Justice and Behavior, 36, 147-162.

57.

Riley

Shaw

(1985). Parental supervision and juvenile delinquency. London, England: Her Majesty’s Stationery Office.

58.

Root

(1973). Endocrinology of puberty: I. Normal sexual maturation. Journal of Pediatrics, 83, 1-19.

59.

Rowe

D. C.

(1986). Genetic and environmental components of antisocial behavior: A study of 265 twin pairs. Criminology, 24, 513-532.

60.

Rowe

D. C.

(1995). The limits of family influence: Genes, experience, and behavior. New York, NY: Guilford.

61.

Rowe

D. C.

(1996). An adaptive strategy theory of crime and delinquency. In Hawkins

J. D.

(Ed.), Delinquency and crime, current theories (pp. 268-314). Cambridge, UK: Cambridge University Press.

62.

Rowe

D. C.

Osgood

D. W.

(1984). Heredity and sociological theories of delinquency: A reconsideration. American Sociological Review, 49, 526-540.

63.

Rowe

D. C.

Rodgers

J. L.

Meseck Bushey

(1992). Sibling delinquency and the family environment: Shared and unshared influences. Child Development, 63, 59-67.

64.

Rowe

Maughan

Worthman

C. M.

Costello

E. J.

Angold

(2004). Testosterone, antisocial behavior, and social dominance in boys: Pubertal development and biosocial interaction. Biological Psychiatry, 55, 546-552.

65.

Schaal

Tremblay

R. E.

Soussignan

Susman

E. J.

(1996). Male testosterone linked to high social dominance but low physical aggression in early adolescence. Journal of the American Academy of Child and Adolescent Psychiatry, 34, 1322-1330.

66.

Schroeder

E. T.

Vallejo

A. F.

Zheng

Stewart

Flores

Nakao

. . . Sattler

F. R.

(2005). Six-week improvements in muscle mass and strength during androgen therapy in older men. Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 60, 1586-1592.

67.

Seidman

S. N.

Klein

D. F.

(2004). AA2500 testosterone gel normalizes androgen levels in aging males with improvements in body composition and sexual function. Journal of Clinical Endocrinology & Metabolism, 89, 6358-6359.

68.

Sjöberg

R. L.

Ducci

Barr

C. S.

Newman

T. K.

Dell’Osso

Virkkunen

Goldman

(2007). A non-additive interaction of a functional MAO-A VNTR and testosterone predicts antisocial behavior. Neuropsychopharmacology, 33, 425-430.

69.

Sluyter

Keijser

J. N.

Boomsma

D. I.

Van Doornen

L. J. P.

Van den Oord

Snieder

(2000). Genetics of testosterone and the aggression-hostility-anger (AHA) syndrome: A study of middle-aged male twins. Twin Research and Human Genetics, 3, 266-276.

70.

Stattin

Kerr

(2000). Parental monitoring: A reinterpretation. Child Development, 71, 1072-1085.

71.

Susman

E. J.

Inoff-Germain

Nottelmann

E. D.

Loriaux

D. L.

Cutler

G. B.

Jr. Chrousos

G. P.

(1987). Hormones, emotional dispositions, and aggressive attributes in young adolescents. Child Development, 58, 1114-1134.

72.

Thornberry

T. P.

Farnworth

(1982). Social correlates of criminal involvement: Further evidence on the relationship between social status and criminal behavior. American Sociological Review, 47, 505-518.

73.

Traish

A. M.

(2009). Androgens play a pivotal role in maintaining penile tissue architecture and erection: A review. Journal of Andrology, 30, 363-369.

74.

van Bokhoven

van Goozen

S. H. M.

van Engeland

Schaal

Arseneault

Seguin

J. R.

. . . Tremblay

R. E.

(2006). Salivary testosterone and aggression, delinquency, and social dominance in a population-based longitudinal study of adolescent males. Hormones and Behavior, 50, 118-125.

75.

Walsh

(2008). Biosocial criminology: New directions in theory and research. London, England: Routledge.

76.

Walsh

Beaver

(Eds.). (2008). Biosocial criminology: New directions in theory and research. Thousand Oaks, CA: Sage.

77.

Wang

Swerdloff

R. S.

Iranmanesh

(2000). Transdermal testosterone gel improves sexual function, mood, muscle strength, and body composition parameters in hypogonadal men. Journal of Clinical Endocrinology and Metabolism, 85, 2839-2853.

78.

Wiebe

R. P.

(2006). Using an expanded measure of self-control to predict delinquency. Psychology, Crime and Law, 12, 519-536.

79.

Wilson

J. Q.

Herrnstein

R. J.

(1985). Crime and human nature. New York, NY: Simon & Schuster.

80.

Wissink

I. B.

Dekovic

Meijer

A. M.

(2006). Parenting behavior, quality of the parent-adolescent relationship, and adolescent functioning in four ethnic groups. Journal of Early Adolescence, 26, 133-159.

81.

Wright

Beaver

Delisi

Vaughn

(2008). Evidence of negligible parenting influences on self-control, delinquent peers, and delinquency in a sample of twins. Justice Quarterly, 25, 544-569.