The Role of ADHD in Predicting the Development of Violent Behavior Among Juvenile Offenders: Participation Versus Frequency

Abstract

Past research has identified attention-deficit/hyperactivity disorder (ADHD) as a risk factor for engagement in violent offending. Despite the link between the disorder and violent offending, this risk factor has yet to be examined as a predictor of heterogeneity in the development of violent offending among juvenile offenders. It is likely that the impulsivity, genetic link, and generally chronic disorder course which are characteristics of the disorder play roles in predicting violent offending, which is consistent with both self-control theory and general developmental theory related to early life deficits and life-course persistent offending. Past research has also elucidated a developmental trajectory model of violent offending, which is utilized by the present research. The present research examines ADHD as a risk factor predicting trajectory group assignment. The Pathways to Desistance data followed 1,354 juvenile offenders for 84 months following conviction for a serious offense. Using multinomial logistic regression, this study extends past research on the development of violent offending among juvenile offenders by examining ADHD as a risk factor predicting assignment to violent offending trajectory groups. Results indicate that meeting criteria for ADHD at baseline predicted membership to all trajectory groups relative to the Abstaining group when all covariates were included. This increase in risk is highest for the trajectory group characterized by the highest frequency of violent offending. This indicates the relevance of identifying and treating ADHD among juvenile offenders to best mitigate risk of violent recidivism throughout adolescence and early adulthood.

Keywords

youth violence mental health and violence criminology

Introduction

Violent offending is a social problem which research has consistently demonstrated to peak in participation and frequency of engagement in adolescence. Following adolescence, rates of engagement and participation tend to decline (Centers for Disease Control and Prevention [CDC], 2009). Despite these general trends, there remain individuals who continue to consistently engage in violent offending into adulthood. While research has explored the development of violent offending across adolescence extensively, there is yet to be an examination of the relevance of attention-deficit/hyperactivity disorder (ADHD) for predicting this development. This research seeks to fill this gap in the literature by examining ADHD symptomatology at baseline as a predictor of general developmental patterns of violent offending in adolescence and early adulthood. Past research utilizing the group-based trajectory modeling (GBTM) analytic strategy revealed that a four-group model best fit the Pathways to Desistance violent offending data (see Wojciechowski, 2020, for this model). While ADHD has been found to predict increased participation and frequency of engagement in violent offending (Araiza, 2014; Bernat, Oakes, Pettingell, & Resnick, 2012; Buitelaar, Posthumus, Scholing, & Buitelaar, 2014), this has yet to be explored solely among an indicated sample of serious juvenile offenders. There is reason to believe that the effects of this disorder among juvenile offenders may differ from that of the general population because of the greater exposure to risk factors to which juvenile offenders are exposed. This research seeks to provide greater understanding of the role that a diagnosis of ADHD has for predicting both participation in violent offending and increased frequency of violent offending.

Violent Offending Across Adolescence

While research indicates that violent offending peaks in adolescence, more advanced methodologies and higher quality data are needed to untangle differences in rates of participation and levels of frequency of engagement change across the life course. Longitudinal data are necessary to model this type of change over time and elucidate the ways that participation (zero offending vs. any offending) and frequency (rates of engagement among those who participate) change over time. As noted above, the GBTM method has been used extensively to provide greater understanding of the development of violent offending during the adolescence period of the life course using longitudinal data. Jennings and Reingle (2012) provide a comprehensive review describing the existing literature using the GBTM method to describe violent offending. The existing literature overwhelmingly identifies three- and four-group models as the best fitting models describing the development of violence in adolescence. Generally, the existing research also demonstrates that participation in violent offending as well as the frequency at which those who do participate tend to decline across adolescence. This review generally indicates that individuals tend to mature and desist engagement in violent behavior following adolescence. However, the majority of this research examines the development of violent offending among general population samples. Very little research has examined the development of violent offending in adolescence among samples of juvenile offenders utilizing the GBTM method. Wojciechowski (2020) provides one of the few examples of such research examining this development among juvenile offenders. This research found four trajectory groups in the given violent offending data. The Abstaining group was characterized by a complete lack of violent behavior perpetrated during adolescence and early adulthood. The Moderate Stable trajectory group was characterized by consistent engagement in two to three violent offenses per year during adolescence and early adulthood. The Desisting trajectory group was characterized by a high violence intercept at age 16 and a sharp deceleration in violence leading to full desistance by age 20. Finally, the High Chronic trajectory group was characterized by consistently high engagement in violent behavior across adolescence and early adulthood. The path of the High Chronic trajectory group indicated violent offending frequencies ranging between five and eight violent offenses per observation period depending on the specific interval of the life course. Further research on high-risk samples of juvenile offenders is necessary to evaluate potential differences in the development of violent behavior between this group and the general population. This study added to the relative dearth of research examining the development of violence among juvenile offenders using the GBTM method. This research also extended this elucidated model by examining post-traumatic stress disorder (PTSD) as a risk factor predicting assignment to developmental trajectory groups. The flexibility of the GBTM method allows for similar extensions of the model for examining the role that other risk factors play in predicting assignment to these developmental trajectory groups. Juvenile offenders are an interesting population to examine in this capacity because of the wide range of risk factors to which they are typically exposed.

Juvenile offenders often possess a number of risk factors, which have a measurable impact on raising their likelihood of engagement in criminal offending or leading to increased frequency of offending. As Nagin and Paternoster (2000) note, one of the best predictors of criminal offending is past criminal offending. Despite this oft-observed continuity, there are likely to be some individuals at higher risk for recidivism even among populations of individuals that all have already been identified as delinquents by the criminal justice system. While a great deal of research has delineated the risk factors that set nonviolent adolescents apart from their violent peers, there is less research which attempts to identify characteristics that delineate violent offenders from those who do not engage in violence among a juvenile population whose members are already at high risk for such offending. Mental health issues have proven to be problematic for a great many adolescents in the juvenile justice system (Underwood & Washington, 2016), and mental health has also demonstrated ability for predicting violent behavior (Connolly & Beaver, 2015; Hay & Evans, 2006; Scheuerman, 2013). It may be that mental disorder is a characteristic, which delineates adolescents who have the highest potential for violence from nonviolent or less violent peers among populations of youth who have already been indicated by the criminal justice system as deviant. One Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5; American Psychiatric Association [APA], 2013) classified mental disorder which has been demonstrated to increase individuals’ potential for violence is ADHD (Coker, Smith, Westphal, Zonana, & McKee, 2014; Elkington et al., 2015; Román-Ithier, González, Vélez-Pastrana, González-Tejera, & Albizu-García, 2017).

ADHD and the Development of Violent Offending

ADHD diagnosis has been found to be a robust risk factor for engagement in violent behavior in adolescence and beyond (Bernat et al., 2012; Ralic & Sifner, 2014; Shelton & Pearson, 2016). Impulsivity is a major defining feature of ADHD and would appear to be the main characteristic of the disorder, which contributes to increased propensity for violent offending (APA, 2013; Krakowski & Czobor, 2014; Moeller, Barratt, Dougherty, Schmitz, & Swann, 2001). When faced with provocative situations, the impulsivity associated with ADHD predisposes sufferers to reactive aggressive behavior when instigated (Babcock, Tharp, Sharp, Heppner, & Stanford, 2014; Chester et al., 2015). Beyond this effect on violent offending, ADHD may also function to increase risk for violence because of the impact that the high impulsivity characteristic of the disorder has on other deviant behavior. ADHD sufferers also have a high risk for suffering from comorbid substance use disorders (APA, 2013). Substance use, especially problematic levels of use, has been consistently identified as a risk factor for engagement in violence (Lundholm, Haggard, Moller, Hallqvist, & Thiblin, 2013; Temple, Shorey, Fite, Stuart, & Le, 2013). The use of substances, like alcohol, can further lower inhibition and self-control, which leads to increased risk for violent behavior in provocative situations (Zerhouni et al., 2013). This is especially problematic among individuals who already have notably high levels of impulsivity, like individuals afflicted with ADHD.

Research has demonstrated a great deal of stability in ADHD symptom presentation (Law, Sideridis, Prock, & Sheridan, 2014). Individuals suffering from ADHD will likely possess the risk for violence associated with the disorder across the life course. This stability is due, in large part, to the genetic nature of ADHD (Sharp, McQuillin, & Gurling, 2009). Many ADHD sufferers then are distinguishable from their non-ADHD peers from birth, a difference that will persist throughout life. Risk factors like this which delineate those at risk for offending from those not at risk based on stable individual differences provide evidence for the population heterogeneity argument of offending risk. Risk factors (such as ADHD, psychopathy, and gender) often found to have formative roots in early life experience or even genetics are evidence of heterogeneity of risk possession related to behaviors of interest among members of a population. Possessing a time-stable risk factor does not condemn individuals to a life of crime. This individual risk is still a function of individuals’ environments. In the case of ADHD, risk may be further mitigated as the development of pharmacological remedies has led to significant impacts on symptom presentation when properly administered (Faraone, 2013; Vaughan & Kratochvil, 2012). Such risk factors representative of population heterogeneity are ideal for predicting the risk for following certain developmental pathways of behavior across the life course as an extension of the GBTM method. In the case of the present research, ADHD diagnosis is examined as a time-stable risk factor for predicting development of violent behavior through adolescence and into early adulthood. Beyond the utility inherent to ADHD for predicting development, characteristics of the disorder link ADHD to more abstract theoretical frameworks in criminology.

The impulsivity which is characteristic of ADHD and the association that it has to violent offending are analogous to the postulations of Gottfredson and Hirschi (1990) in their self-control theory. Self-control theory posits that low self-control is the only major important predictor of criminal offending as, they assert, it is the root cause of crime and delinquency through which other mechanisms may be derived from. Indeed, tests of the theory have demonstrated the explanatory power of the perspective (DeLisi, Tostlebe, Burgason, Heirigs, & Vaughn, 2018; Hay, 2001; Vazsonyi, Jiskrova, Ksinan, & Blatný, 2016). High impulsivity is characteristic of ADHD and, for all intents and purposes, would seem to be analogous to low self-control. In this way, it may be that ADHD diagnosis provides a medicalized terminology for this loose concept of low self-control, which has been examined extensively in the field of criminology. Further rooting ADHD in the criminological canon is Moffitt’s (1993) taxonomy of life-course persistent and adolescent limited offenders. Moffitt indicates that early life deficits are major risk factors for early onset of offending and life-course persistent offending. ADHD has a highly genetic link, and symptoms may become apparent early in life (APA, 2013). ADHD then may function as one of these early life deficits whose symptoms may contribute to the problematic caregiver relationships described by Moffitt, as well as deficits associated with high impulsivity inherent to the disorder, which may contribute to high and chronic violent offending across the life course.

Despite the general relevance of associated concepts for predicting offending, there has yet to be an investigation of how ADHD predicts the development of violent offending across adolescence and early adulthood among juvenile offenders. While past research has examined ADHD as a risk factor for the development of general offending among juvenile offenders (Baglivio, Wolff, Piquero, Greenwald, & Epps, 2016), it is necessary for researchers to examine different types of offending separately to examine the possibility of differing risk factors associated with these different types of offending. Furthermore, prior research has yet to examine these relationships past the age of 18. Terminus of investigation at this age falls within the general adolescent time frame, which Moffitt (1993) describes as being a period of time in which some level of engagement in deviance is normative. Investigations of violent offending in early adulthood are necessary as continued engagement at that time may be indicative of the more chronic courses of offending and should be made the priority of criminal justice officials. The present research seeks to fill this gap by examining ADHD at baseline as a risk factor for predicting assignment to the violent offending trajectory groups described by Wojciechowski (2020).

Method

Data

The Pathways to Desistance data follows a purposive sample of 1,354 juvenile offenders over the course of 84 months, resulting in 11 possible data points for all participants. Enrollment in the study took place between 2000 and 2003. All participants had recently been convicted of a serious crime at baseline. Serious crimes resulting in convictions mostly consisted of felonies, with some misdemeanor convictions acting as qualifying offenses. Enrollment of male drug offenders in the study was capped at 15% to maintain sample heterogeneity. Inclusion criteria also included that the serious offense had to have been committed when participants were between 14 and 17 years of age. All participants were between ages 14 and 18 at baseline (one participant was 19 years of age at baseline). Recruitment of participants took place in Maricopa County (Arizona) and Philadelphia (Pennsylvania); 20% of potential participants approached for recruitment declined participation.

Computer-assisted interview (CAI) technology was utilized to obtain data from participants. Participants were read prompts by a member of the research team, and the CAI technology was used by participants to report answers. Interviews were conducted at the convenience of participants in participants’ homes, public places, or criminal justice facilities if participants were currently under supervision. Collateral interview with peers and guardians of participants, as well as official records, provided supplemental data regarding information which it would be difficult for participants to provide themselves (e.g., ADHD symptomatology).

Measures

Violent offending

Violent offending was measured via a count of the number of violent offenses, which each participant self-reported engagement in during each of the 11 observation periods. This count measure was a combination of 10 individual violent offense types added together to form a single index of violent offending.¹

ADHD diagnosis

ADHD was captured as a measure of whether or not participants had currently met criteria for a diagnosis for ADHD at baseline. This is a binary measure delineating those who did (1) or did not (0) meet criteria for diagnosis at baseline. Research has demonstrated that ADHD is a disorder with a relatively stable course with a genetic component (Law et al., 2014; Sharp et al., 2009). This means that participants demonstrating diagnosable symptomatology at baseline are likely to continue to demonstrate this same symptomatology across time if no treatment is provided. This provides relatively strong rationale for utilizing this baseline measure as a predictor of violent offending at later time points in the absence of measures of ADHD symptomatology beyond baseline.

Control variables

Several variables were included in analyses as controls to best reduce bias. The first of these variables was gender. Research has demonstrated that males engage in violent offending at much higher rates than do females (Steffensmeier, Schwartz, Zhong, & Ackerman, 2005). This variable was captured at baseline as a binary indicator of each participant’s self-reported gender identity (0 = male, 1 = female).

Race was another variable included in analyses as a control variable. Research has demonstrated differing propensity for engagement in violent offending based on race (Hawkins, Laub, Lauritsen, & Cothern, 2000). Race was included in analyses as a series of dummy variables, which delineated Black, Hispanic, and Other Race participants from White participants. White was the reference category for each of the race dummy variables.

The final variable included as a control captured the socioeconomic status (SES) of participants’ parents at baseline. Research has demonstrated SES to be a predictor of violent offending (Piotrowska, Stride, Croft, & Rowe, 2015). The measures of SES captured by the Pathways to Desistance study include separate mother and father measures of SES. This measure was a weighted composition of occupational prestige and educational attainment, resulting in a continuous approximation of SES score. In addition to the individual measures, a measure of total parental SES was also available. This measure was simply the average of the sum of the individual parent measures. This combined measure of SES was chosen as the measure utilized in analyses by this study.

Analytic Strategy

The present research utilized a GBTM model describing general developmental pathways of violence across adolescence and early adulthood to examine the role of ADHD diagnosis for predicting membership to elucidated trajectory groups in the model (Wojciechowski, 2020). The GBTM method begins with the identification of the number of groups and the shapes of these groups, which provides the best overall fit to the data. Participants are “assigned” to a specific trajectory group based on their violent offending history. Participants were assigned to the group to which they had the highest probability of assignment. The model which provided the best overall fit to the data based on several fit criteria was selected as the best fitting model to represent violent offending among juvenile offenders. This was the four-group model described previously in this article. The second phase of the analytic process utilized a series of multinomial logistic regression models to better understand the determinants of assignment to the elucidated trajectory groups. The baseline measure of ADHD was included in the first model as a means of determining the effect that this disorder has on affecting participants’ risk of assignment to particular trajectory groups. In the case of multinomial logistic regression, risk of assignment to a given trajectory group is relative to a reference group omitted from analysis. The second model included race, gender, and SES as controls to determine the least biased effect of ADHD on the risk of assignment to trajectory groups.

Hypotheses

Hypothesis 1: Meeting criteria for ADHD at baseline will predict assignment to the Moderate Stable, Desisting, and High Chronic trajectory groups relative to the Abstaining trajectory group.

With research indicating that ADHD is a risk factor for engagement in violent behavior (Araiza, 2014; Bernat et al., 2012; Buitelaar et al., 2014), it would be expected that meeting criteria for the disorder at baseline will predict assignment to trajectory groups characterized by violence. Given that the all trajectory groups in the elucidated model demonstrate some participation in violent behavior at some point in adolescence and early adulthood other than the Abstaining group, ADHD should predict assignment to the three other groups. Effects demonstrating an increase in the risk of assignment to these groups relative to the Abstaining trajectory would be demonstrative of the effect of ADHD increasing propensity to participate in violent behavior.

Hypothesis 2: Meeting criteria for ADHD at baseline will result in the greatest increase in relative risk of assignment to the High Chronic trajectory group.

Research indicates that ADHD does increase propensity for engagement in violent offending (Araiza, 2014; Bernat et al., 2012; Buitelaar et al., 2014). Given that the High Chronic trajectory is characterized by consistently high engagement in violent behavior across adolescence and early adulthood, it is predicted that meeting clinical criteria for ADHD at baseline will predict assignment to this trajectory group. Because the High Chronic trajectory group demonstrates greater engagement in violent offending at all time points, it can be inferred that increased risk of assignment to this group greater than increased risk of assignment to other groups is demonstrative of ADHD’s effect on increasing the frequency of engagement in violent offending.

Results

As noted above, the trajectory model examined consisted of four trajectory groups with a proportion of the study sample assigned to each (Abstaining = 47.9%, Moderate Stable = 16.5%, Desisting = 25.4%, High Chronic = 10.1%). This model met all standards for adequacy of model fit described by Nagin (2005). This includes posterior probabilities of assignment exceeding .7, average odds of correct classification exceeding 5 for all groups, adequate comparative model fit based on Bayesian Information Criteria (BIC) statistics, and visual confirmation of nuance provided by all groups and relatively tight confidence intervals surrounding each trajectory group. All of these criteria provide confidence in the chosen trajectory model for best describing the heterogeneity in the development of violence in the Pathways to Desistance data (see Wojciechowski, 2020, for a full discussion of the fit criteria assessed, a full description of the analytic process involved in the elucidation of the four-group model, and a visual presentation of the four-group model).

The analytic process examined the explanatory power that ADHD had for predicting trajectory group membership utilizing a series of multinomial logistic regression models. Table 1 provides correlations of all variables utilized in the multinomial logistic regression models. Model 1 estimated the effect of ADHD alone on the relative risk of group membership, while Model 2 examines these effects together with gender, race, and SES included in the model as covariates. Table 2 provides coefficient estimates for Model 1, whereas Table 3 provides these estimates for Model 2. All coefficients represent the risk of assignment to a trajectory group relative to the reference category (Abstaining). Model 1 estimates demonstrated that meeting clinical criteria for ADHD at baseline does indeed predict increased propensity for engagement in violent offending. Meeting criteria for ADHD diagnosis more than tripled the risk of assignment to the Moderate Stable trajectory group (3.284) and nearly tripled the risk of assignment to the High Chronic group (2.993) relative to the reference group.

Table 1.

Correlations of Independent Variables Included in Analyses.

	Gender	Black	Hispanic	Other Race	SES
Gender (0 = male, 1 = female)	—
Race (reference = White)
Black	−.0360	—
Hispanic	−.0260	—
Other Race	.0218	—
SES	−.0510	−.1140*	.3207*	−.0365
ADHD diagnosis at baseline (0 = no, 1 = yes)	−.0053	−.0725*	−.0487	−.0267	−.0636*

Note. SES = socioeconomic status; ADHD = attention-deficit/hyperactivity disorder.

p ≤ .05.

Table 2.

Relative Risk of Assignment to Violence Trajectory Groups Based on Covariates—Model 1.

	Abstaining	Moderate Stable	Desisting	High Chronic
ADHD diagnosis at baseline (0 = no, 1 = yes)	—(reference)	3.284**	2.028	2.993*
Constant	—	0.304***	0.493***	0.178***

Note. ADHD = attention-deficit/hyperactivity disorder.

p ≤ .05. **p ≤ .01. ***p ≤ .001.

Table 3.

Relative Risk of Assignment to Violence Trajectory Groups Based on Covariates—Model 2.

	Abstaining	Moderate Stable	Desisting	High Chronic
ADHD diagnosis at baseline (0 = no, 1 = yes)	—(reference)	2.974**	2.121*	2.981*
Gender (0 = male, 1 = female)	—	0.317***	0.380***	0.102***
Race (reference = White)
Black	—	0.585*	0.945	0.686
Hispanic	—	0.851	1.019	1.117
Other Race	—	0.485	1.060	1.699
SES	—	0.998	1.003	0.995
Constant	—	1.657	1.303	2.919

Note. ADHD = attention-deficit/hyperactivity disorder; SES = socioeconomic status.

p ≤ .05. **p ≤ .01. ***p ≤ .001.

Model 2 examined these same effects while controlling for the effects of race, gender, and SES. Being male significantly increased the risk of assignment to the Moderate Stable, Desisting, and High Chronic groups relative to the Abstaining group. Race only significantly affected assignment to the Moderate Stable trajectory, with being Black significantly raising the relative risk of assignment to this trajectory group, relative to White participants. SES did not significantly affect the relative risk of assignment to the trajectory groups. The inclusion of these variables in the model significantly affected the effects of ADHD relative to the effects demonstrated in Model 1. Meeting clinical criteria for ADHD diagnosis at baseline significantly affected the risk of assignment to the Moderate Stable, Desisting, and High Chronic groups. Meeting criteria for ADHD at baseline more than doubled the relative risk of assignment to the Desisting trajectory group (2.121). The relative risk of assignment was attenuated somewhat for the Moderate Stable trajectory, though this risk was still nearly tripled for individuals who met criteria for ADHD diagnosis at baseline (2.974). Inclusion of these covariates resulted in an increase in the effect of ADHD on the risk of assignment to the High Chronic group relative to the Abstaining group. The relative risk of assignment to the High Chronic group was also nearly tripled (2.981) for participants who met criteria for ADHD diagnosis at baseline. This was the greatest effect on the relative risk of assignment for any covariate for any of the three trajectory groups.

BIC statistics were calculated for Model 1 and Model 2 to ascertain information regarding comparative model fit. The obtained BIC statistics indicate that the inclusion of relevant covariates in Model 2 (–5,445.691) provides better fit relative to Model 1 (–5,534.875), which only includes baseline ADHD status as a predictor.

Discussion

ADHD was examined as a risk factor related to assignment to the elucidated trajectory groups. Meeting criteria for ADHD diagnosis at baseline was demonstrated to be an important predictor assignment to trajectory groups characterized by violent offending among juvenile offenders. When all relevant covariates were added to the model, participants who had met criteria for ADHD diagnosis at baseline were at greatest risk for being assigned to the High Chronic violent offending trajectory group relative to the Abstaining trajectory group. In addition, meeting criteria for ADHD diagnosis at baseline significantly increased the risk of assignment to the Desisting and Moderate Stable trajectory groups relative to the Abstaining group. The ways in which the relative risk of assignment to each of the trajectory groups is affected provide support for both hypotheses of this research. Not only did meeting criteria for ADHD at baseline predict participation in violent offending, but there is evidence that it has an impact on increasing the frequency of offending among juvenile offenders as well. The findings of this research present several relevant implications for future research on ADHD and violent offending among juvenile offenders.

ADHD has consistently been demonstrated to be a predictor of violent offending among adolescents (Araiza, 2014; Bernat et al., 2012; Buitelaar et al., 2014). This research provides additional evidence for this relationship and extends the scope of this knowledge to an indicated sample of juvenile offenders. Previous research had yet to examine the ways in which ADHD is related in increased propensity for and frequency of violent offending among juvenile offenders. Even among this subpopulation of adolescents at high risk for engagement in violent offending, ADHD still discriminated between those who demonstrated the highest risk for engagement in violence and those who did not. Furthermore, this research demonstrates that meeting criteria for diagnosis of ADHD also predicts increased levels of engagement in violence among those who participate. This demonstrates further just how important early diagnosis and treatment of ADHD may be for reducing the odds that a youth will engage in violent behavior. This highlights the need for comprehensive mental health screening for all youth who enter the juvenile justice system. Such screening may identify previously unrecognized mental health issues and allow for clinicians to provide individualized treatment which can best mitigate the risk for violent offending among adolescents while under criminal justice supervision and beyond. The vast differences in developmental tracks demonstrated by the High Chronic and Desisting trajectory groups, despite similar baseline levels of violence perpetration, should be interrogated further as well. It may be that these differences may be due to the receipt of treatment of ADHD around this period. Given that intake into the criminal justice system may result in diagnosis and the receipt of treatment, it is possible that individuals assigned to these groups may diverge in their levels of violence perpetration across the life course due to these treatment differences. Unfortunately, it was not possible for the present study to investigate these differences. Future research should continue to investigate the role that ADHD plays in predicting violent behavior across the life course and the role that treatment may play in predicting deceleration in violence perpetration. One possible route may be an examination of the roles that distinct symptom clusters or ADHD presentation types play in predicting violent offending. With research demonstrating a link between low self-control/impulsivity and violence (Cheung, Choi, & Cheung, 2014; Krakowski & Czobor, 2014; Moeller et al., 2001), it may be that the hyperactive–impulsive symptom cluster or presentation type may be more important for understanding violent offending. A more nuanced understanding of the role that distinct symptom presentations may play in predicting violent behavior will better allow clinicians to provide the optimum treatment possible for affecting violent offending among juvenile offenders with ADHD.

There are several limitations to the present research, which require acknowledgment. The first of these relates to the GBTM method itself. The elucidated trajectories should be interpreted with caution and should not be seen as concrete entities. Rather, the trajectory mapping functions as a powerful descriptive tool, which identifies relevant patterns and provides interpretable results in data that would be otherwise intractable. Another limitation of this research pertains to the self-report nature of the violent offending data. While official records and other reporting methods have limitations also, it is possible that self-reports of violent offending in this research may have led to underreporting. Because many participants were likely under criminal justice supervision of some sort during a portion of the study period, they may not have felt comfortable providing honest reporting regarding their offending due to concerns about it leading to reporting to criminal justice officials. The research team provided guarantees of anonymity and confidentiality in responses, but there is still no guarantee that all participants felt comfortable honestly reporting their criminal behavior. This, of course, is a concern for all research using self-report measures, but because all participants were under criminal justice supervision for some portion of this study, this concern is somewhat amplified. A final limitation of the present study relates to treatment that participants may have received for ADHD symptoms during the study period. Given that such treatment may affect violence perpetration due to the alleviation of ADHD symptoms, receipt of treatment may moderate group assignment. Unfortunately, the Pathways to Desistance data do not allow for the inclusion of this construct in analyses. While there are numerous measures related to medication and service receipt, a measure specifically related to ADHD treatment is not available. Because of this limitation of the data, further interrogation of this issue is beyond the scope of this study.

This research extends the existing research on violent offending in adolescence by providing a greater understanding of the role that ADHD plays in predicting participation and frequency of violent offending among an indicated sample of juvenile offenders. While a great deal of research has focused on developmental pathways of violence and ADHD as a risk factor, this research has mainly been conducted among general population samples of adolescents. While there are undoubtedly adolescents in these samples who qualify as juvenile offenders, the Pathways to Desistance sample consists completely of adolescents recently convicted of a serious crime at baseline. This sample, therefore, has a high degree of risk for engagement in subsequent serious offenses, like violent offending. Because of this, it is necessary to better understand the ways that violence may develop differently among this subpopulation of adolescents and the ways that ADHD may differently affect the risk for engagement in violent offending for an already high-risk group. Further research into the role that other risk factors for engagement in violence among juvenile offenders is necessary to better understand how assignment to trajectory groups may be predicted. In addition, time-varying events or risk factors should be examined in the future to determine the possibility some experience may lead to divergence from these trajectory paths.

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.

Notes

Author Biography

Thomas W. Wojciechowski is currently a PhD candidate at the University of Florida in the Department of Sociology and Criminology & Law. His reseaarch focuses on the longitudinal development of substance use and violence outcomes, prevention science, and mental health.

References

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.

Araiza

J. F. R.

(2014). Attention deficit disorder with and without hyperactivity (ADHD) and violence: Literature review. Salud Mental, 37:, 75-82.

Babcock

J. C.

Tharp

A. L. T.

Sharp

Heppner

Stanford

M. S.

(2014). Similarities and differences in impulsive/premeditated and reactive/proactive bimodal classifications of aggression. Aggression and Violent Behavior, 19, 251-262. doi:10.1016/j.avb.2014.04.002

Baglivio

M. T.

Wolff

K. T.

Piquero

A. R.

Greenwald

M. A.

Epps

(2016). Racial/ethnic disproportionality in psychiatric diagnoses and treatment in a sample of serious juvenile offenders. Journal of Youth and Adolescence, 1-28. doi: 10.1007/s10964-016-0573-4

Bernat

D. H.

Oakes

J. M.

Pettingell

S. L.

Resnick

(2012). Risk and direct protective factors for youth violence results from the national longitudinal study of adolescent health. American Journal of Preventive Medicine, 43, S41-S50. doi:10.1016/j.amepre.2012.04.023

Buitelaar

N. J. L.

Posthumus

J. A.

Scholing

Buitelaar

J. K.

(2014). Impact of treatment of ADHD on intimate partner violence (ITAP), a study protocol. BMC Psychiatry, 14(1), 336. doi:10.1186/s12888-014-0336-2

Centers for Disease Control and Prevention. (2009). Youth violence: Fact sheet. Retrieved from http://www.cdc.gov/ncipc/factsheets/yvfacts.htm

Chester

D. S.

DeWall

C. N.

Derefinko

K. J.

Estus

Peters

J. R.

Lynam

D. R.

Jiang

(2015). Monoamine oxidase A (MAOA) genotype predicts greater aggression through impulsive reactivity to negative affect. Behavioural Brain Research, 283, 97-101. doi:10.1016/j.bbr.2015.01.034

Cheung

Y. W.

Choi

S. Y. P.

Cheung

A. K. L.

(2014). Strain, self-control, and spousal violence: A study of husband-to-wife violence in Hong Kong. Violence and Victims, 29, 280-299. doi:10.1891/0886-6708.vv-d-12-00071r2

10.

Coker

K. L.

Smith

P. H.

Westphal

Zonana

H. V.

McKee

S. A.

(2014). Crime and psychiatric disorders among youth in the US population: An analysis of the National Comorbidity Survey–Adolescent Supplement. Journal of the American Academy of Child and Adolescent Psychiatry, 53, 888-898. doi:10.1016/j.jaac.2014.05.007

11.

Connolly

E. J.

Beaver

K. M.

(2015). Assessing the salience of gene–environment interplay in the development of anger, family conflict, and physical violence: A biosocial test of General Strain Theory. Journal of Criminal Justice, 43, 487-497. doi:10.1016/j.jcrimjus.2015.11.001

12.

DeLisi

Tostlebe

Burgason

Heirigs

Vaughn

(2018). Self-control versus psychopathy: A head-to-head test of general theories of antisociality. Youth Violence and Juvenile Justice, 16(1), 53-76.

13.

Elkington

K. S.

Teplin

L. A.

Abram

K. M.

Jakubowski

J. A.

Dulcan

M. K.

Welty

L. J.

(2015). Psychiatric disorders and violence: A study of delinquent youth after detention. Journal of the American Academy of Child and Adolescent Psychiatry, 54, 302-312. doi:10.1016/j.jaac.2015.01.002

14.

Faraone

S. V.

(2013). Crime and therapy: Good news about pharmacotherapy for ADHD. Psychiatric Times, 30, 1-2.

15.

Gottfredson

M. R.

Hirschi

(1990). A general theory of crime. Stanford, CA: Stanford University Press.

16.

Hawkins

D. F.

Laub

J. H.

Lauritsen

J. L.

Cothern

(2000). Race, ethnicity, and serious and violent juvenile offending. Washington, DC: U.S. Department of Justice, Office of Justice Programs, Office of Juvenile Justice and Delinquency Prevention.

17.

Hay

(2001). Parenting, self-control, and delinquency: A test of self-control theory. Criminology, 39, 707-736.

18.

Hay

Evans

M. M.

(2006). Violent victimization and involvement in delinquency: Examining predictions from general strain theory. Journal of Criminal Justice, 34, 261-274. doi:10.1016/j.jcrimjus.2006.03.005

19.

Jennings

W. G.

Reingle

J. M.

(2012). On the number and shape of developmental/life-course violence, aggression, and delinquency trajectories: A state-of-the-art review. Journal of Criminal Justice, 40, 472-489. doi:10.1016/j.jcrimjus.2012.07.001

20.

Krakowski

M. I.

Czobor

(2014). Depression and impulsivity as pathways to violence: Implications for antiaggressive treatment. Schizophrenia Bulletin, 40, 886-894. doi:10.1093/schbul/sbt117

21.

Law

E. C.

Sideridis

G. D.

Prock

L. A.

Sheridan

M. A.

(2014). Attention-deficit/hyperactivity disorder in young children: Predictors of diagnostic stability. Pediatrics, 133, 659-667. doi:10.1542/peds.2013-3433

22.

Lundholm

Haggard

Moller

Hallqvist

Thiblin

(2013). The triggering effect of alcohol and illicit drugs on violent crime in a remand prison population: A case crossover study. Drug and Alcohol Dependence, 129, 110-115. doi:10.1016/j.drugalcdep.2012.09.019

23.

Moeller

F. G.

Barratt

E. S.

Dougherty

D. M.

Schmitz

J. M.

Swann

A. C.

(2001). Psychiatric aspects of impulsivity. American Journal of Psychiatry, 158, 1783-1793. doi:10.1176/appi.ajp.158.11.1783

24.

Moffitt

T. E.

(1993). Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100, 674-701. doi:10.1037/0033-295x.100.4.674

25.

Nagin

(2005). Group-based modeling of development. Cambridge, MA: Harvard University Press.

26.

Nagin

Paternoster

(2000). Population heterogeneity and state dependence: State of the evidence and directions for future research. Journal of Quantitative Criminology, 16, 117-144. doi:10.1023/a:1007502804941

27.

Piotrowska

P. J.

Stride

C. B.

Croft

S. E.

Rowe

(2015). Socioeconomic status and antisocial behaviour among children and adolescents: A systematic review and meta-analysis. Clinical Psychology Review, 35, 47-55. doi:10.1016/j.cpr.2014.11.003

28.

Ralic

A. Z.

Sifner

(2014). Characteristics of peer interaction and experience of peer violence among children and adolescents with ADHD. Ljetopis Socijalnog Rada, 21, 453-484. doi:10.3935/ljsr.v21i2.24

29.

Román-Ithier

J. C.

González

R. A.

Vélez-Pastrana

M. C.

González-Tejera

G. M.

Albizu-García

C. E.

(2017). Attention deficit hyperactivity disorder symptoms, type of offending and recidivism in a prison population: The role of substance dependence. Criminal Behaviour and Mental Health, 27(5), 443-456.

30.

Scheuerman

H. L.

(2013). The relationship between injustice and crime: A general strain theory approach. Journal of Criminal Justice, 41, 375-385. doi:10.1016/j.jcrimjus.2013.06.019

31.

Sharp

S. I.

McQuillin

Gurling

H. M. D.

(2009). Genetics of attention-deficit hyperactivity disorder (ADHD). Neuropharmacology, 57, 590-600. doi:10.1016/j.neuropharm.2009.08.011

32.

Shelton

Pearson

(2016). ADHD in juvenile offenders: Treatment issues nurses need to know. Journal of Psychosocial Nursing and Mental Health Services, 43, 38-46. doi:10.3928/02793695-20050901-07

33.

Steffensmeier

Schwartz

Zhong

Ackerman

(2005). An assessment of recent trends in girls’ violence using diverse longitudinal sources: Is the gender gap closing? Criminology, 43, 355-405. doi:10.1111/j.0011-1348.2005.00011.x

34.

Temple

J. R.

Shorey

R. C.

Fite

Stuart

G. L.

V. D.

(2013). Substance use as a longitudinal predictor of the perpetration of teen dating violence. Journal of Youth and Adolescence, 42, 596-606. doi:10.1007/s10964-012-9877-1

35.

Underwood

L. A.

Washington

(2016). Mental illness and juvenile offenders. International Journal of Environmental Research and Public Health, 13(2), 228.

36.

Vaughan

Kratochvil

C. J.

(2012). Pharmacotherapy of pediatric attention-deficit/hyperactivity disorder. Child and Adolescent Psychiatric Clinics of North America, 21, 941-955. doi:10.1016/j.chc.2012.07.005

37.

Vazsonyi

A. T.

Jiskrova

G. K.

Ksinan

A. J.

Blatný

(2016). An empirical test of self-control theory in Roma adolescents. Journal of Criminal Justice, 44, 66-76.

38.

Wojciechowski

T. W.

(2020). PTSD as a risk factor for the development of violence among juvenile offenders: A group-based trajectory modeling approach. Journal of Interpersonal Violence, 35(13-14), 2511-2535.

39.

Zerhouni

Begue

Brousse

Carpentier

Dematteis

Pennel

Cherpital

(2013). Alcohol and violence in the emergency room: A review and perspectives from psychological and social sciences. International Journal of Environmental Research and Public Health, 10, 4584-4606. doi:10.3390/ijerph10104584