Abstract
The Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5; American Psychiatric Association [APA], 2013) characterizes ADHD as chronic, developmentally inappropriate inattention and/or hyperactivity-impulsivity that causes impairment. The prevalence rates of ADHD in children is approximately 5% (APA, 2013; Faraone, Sergeant, Gillberg, & Biederman, 2003; Polanczyk, de Lima, Horta, Biederman, & Rohde, 2007; Willcutt, 2012). ADHD appears to continue into adulthood for up to two thirds of those diagnosed as children (Barkley, 2006; Wender, Wolf, & Wasserstein, 2006; Weyandt & DuPaul, 2006). It is estimated that prevalence rates for ADHD in adults falls between 2.5% and 5% (Garnier-Dykstra, Pinchevsky, Caldeira, Vincent, & Arria, 2010; Simon, Czobor, Bálint, Mészáros, & Bitter, 2009; Willcutt, 2012). “Outgrowing” a disorder is typical for some childhood disorders such as Tourette’s syndrome (APA, 2013). The fact that children can outgrow some disorders may have contributed to the assumption that it is possible to outgrow ADHD; however, most researchers no longer support this notion (Barkley, Murphy, & Fischer, 2008; M. D. Weiss & Weiss, 2004). In fact, ADHD in adults and emerging adults has been gaining more and more research attention. Emerging adults are young people between the ages of 18 and 25 years who are going through a unique developmental period (Arnett, 2000). Emerging adults are considered to be significantly different from both adolescents and adults because of the ambiguity of their social role and the exploration of varying avenues of life (Arnett, 2000). Until the age of 18, children and adolescents experience stability in many domains of life, but within approximately the next 10 years, there are significant changes in things such as place of residence, marital status, and education for the majority of emerging adults (Arnett, 2000). This unique developmental period may be linked to specific issues in a formative period of one’s life in terms of social, romantic, and educational functioning (Arnett, 2000).
ADHD is marked by deficits in the executive function (EF) system. EF processes are thought to cause emotional impulsivity in individuals with ADHD (Barkley & Fischer, 2010). More specifically, a person who is high in emotional impulsivity is more likely impatient, quick to anger, easily frustrated, prone to overreact emotionally, easily excited, likely to lose temper, and easily annoyed (Barkley & Fischer, 2010). Therefore, individuals with ADHD may behave more reactively to their environment (Barkley, 1997). In addition, this EF deficit may impair the ability of individuals with ADHD to regulate and inhibit problematic behaviors (Barkley et al., 2008), and this could lead to impairment in multiple life domains.
ADHD impairs a variety of important domains of life, such as education, occupation, economic stability, health and well-being, sexual behavior, psychological functioning, driving ability, and substance use (Barkley, Fischer, Smallish, & Fletcher, 2006; Barkley et al., 2008; Biederman et al., 2006; Flory, Molina, Pelham, Gnagy, & Smith, 2006; Ramsay, 2010; Thompson, Molina, Pelham, & Gnagy, 2007). In addition, ADHD-related impairment in social functioning seems to continue from childhood into emerging adulthood (Nijmeijer et al., 2008; G. Weiss & Hechtman, 1993). Difficulties in adult social relationships can be a product of difficulty with impulse control, which can be manifested as talking too much, interrupting others, and making discourteous comments (M. D. Weiss & Weiss, 2004). Adult romantic relationships among individuals with ADHD may also suffer. Prior studies confirm this in that both male and female college students with elevated ADHD symptoms endorsed higher stress and poorer coping strategies in their romantic relationships (e.g., withdrawal from and denial of problems, using alcohol/drugs; Overbey, Snell, & Callis, 2011). Furthermore, females with ADHD are simply in fewer romantic relationships than females without ADHD (Babinski et al., 2011). Another study revealed that only a small portion of young adults with hyperactive symptoms were dating, engaged, or married (Milman, 1979). When adults with ADHD participate in therapy, a common complaint is of problems with their spouses (M. Weiss, Hechtman, & Weiss, 1999). In addition, adults with ADHD tend to report more marital dissatisfaction and romantic relationship maladjustment than adults without ADHD (Canu & Carlson, 2003; Eakin et al., 2004; Murphy & Barkley, 1996), and they also have higher rates of divorce and more numerous marriages than adults without ADHD (Biederman, Faraone, Spencer, & Wilens, 1993; Murphy & Barkley, 1996).
Research suggests that anger may be influentially associated to the relational difficulties experienced by emerging adults with ADHD, as they report significantly higher levels of state and trait anger than emerging adults without ADHD (Barkley et al., 2008; Ramirez et al., 1997). The distinction between state and trait anger allows for separate analysis of the relatively brief emotional state of anger (state anger), assumed to be a fluid state; and anger as a personality trait (trait anger), which is assumed to be a more stable personality characteristic (Spielberger & Sydeman, 1994). More specifically, state anger can be defined as “a psychobiological state or condition consisting of subjective feelings of anger that vary in intensity, from mild irritation or annoyance to intense fury and rage, with concomitant activation or arousal of the autonomic nervous system” (Spielberger & Sydeman, 1994, p. 302); conversely, trait anger is defined as “individual differences in the frequency that S[tate]-anger was experienced over time,” the higher degree of trait anger an individual has, the more he or she will experience state anger (Spielberger & Sydeman, 1994, p. 302). Although the co-occurrence of ADHD and anger does not guarantee negative outcomes, there is reason for concern. For example, adults with ADHD self-reported more maladaptive anger expression strategies (i.e., verbal altercations, breaking objects, and negative body language) than adults without ADHD (Ramirez et al., 1997). Similarly, externalized aggressive behaviors were predicted by ADHD-related impulsivity and hyperactivity, whereas self-harm was predicted by ADHD-related inattention (Dowson & Blackwell, 2009). These aggressive behaviors that are sometimes present in individuals with ADHD may result in relationship strain (Ramirez et al., 1997) and/or other problems.
A possible extension of these issues to romantic relationships is the occurrence of intimate partner violence (IPV). IPV is defined as “physical, sexual, or psychological harm by a current or former partner or spouse” (Centers for Disease Control and Prevention [CDC], 2010, para. 1). Research has demonstrated EF impairments among IPV perpetrators. Specifically, males with a history of domestic violence demonstrated poorer EF than controls (Cohen, Rosenbaum, Kane, Warnken, & Benjamin, 1999; Stanford, Conklin, Helfritz, & Kockler, 2007). Schafer and Fals-Stewart (1997) found that poor performance on specific neuropsychological tests of EF was correlated with higher levels of husband-initiated partner violence. Moreover, if ADHD is defined by EF deficits, it would follow that ADHD may be related to IPV. Research conducted in a court-mandated domestic violence treatment program found that 94% of men tested met criteria for an ADHD diagnosis (Mandell, 1999). Similarly, a relation between ADHD and relationship aggression was discovered in emerging adults (Theriault & Holmberg, 2001). Another study of adult men with ADHD revealed that symptoms of ADHD predicted IPV, and a diagnosis of ADHD predicted IPV with injury (Fang, Massetti, Ouyang, Grosse, & Mercy, 2010). A more recent study examined adult males with and without ADHD, and found that males with ADHD reported more verbal aggression and violence against their romantic partners than did males without ADHD (Wymbs et al., 2012). Results revealed that the number and severity of ADHD symptoms were positively related to the reported level of violence (Wymbs et al., 2012).
Current Study
Due to the relatively limited research on impairment in emerging adults with ADHD, there is room for more exploration of social impairment, anger experience, romantic relationship satisfaction, and rates of IPV in young adults with this disorder. Thus, the current study explored whether ADHD symptomology was related to social impairment, state and trait anger, romantic relationship satisfaction, and IPV in a college-student population, a subset of emerging adults. College students were of particular interest to the authors as they represent a relatively high-achieving group subject to higher education-related stressors, while experiencing equal or greater impairment—but less treatment-seeking—than non-college-attending individuals with ADHD (Blanco et al., 2008; Hunt & Eisenberg, 2010). It was hypothesized that higher levels of ADHD would predict higher levels of social impairment and state and trait anger, lower levels of romantic relationship satisfaction, and more occurrences of IPV.
Method
Participants
Participants included a sample of 176 college students (66.5% female), a large portion of whom identified as Caucasian/White (89.8%). A majority of participants were 18 or 19 years of age (79.0%), and a majority reported freshman or sophomore academic standing (83.0%). A total of 31 participants reported an ADHD/ADD (attention deficit disorder) diagnosis (17.6%), and the remaining 145 participants (82.4%) did not. Of the participants who reported an ADHD diagnosis, more than half endorsed currently taking medication for ADHD (61.3%) and a comorbid disorder (54.8%). Approximately 84% of participants were recruited through the psychology department participant pool, and the remainder were recruited because of their likelihood of having ADHD (see “Procedure” section for details). Exclusion criteria included individuals younger than 18 and older than 25 (8 participants). In addition, participants who completed the study in an unrealistic time (i.e., less than 10 min) were excluded (14 participants). These 22 excluded participants are not included in the total sample of 176.
Procedure
The Institutional Review Board of the primary university (in the Midwestern United States) approved the study, and letters of cooperation were obtained from three other universities (one in the Mountain West, one in the Upper Midwest, and one in the Mid-Atlantic regions of the United States). To recruit individuals with high levels of ADHD symptomology, researchers emailed constituents of the student disability services at two of the aforementioned universities, as well as individuals who had completed previous ADHD studies at three of the aforementioned universities (totaling 16% of the sample). In addition, a description of the study was posted on the introductory psychology participant pool website at the primary university. Participants consented and completed the survey on Qualtrics©. At the conclusion of the study, participants were presented with a list of resources for counseling services and IPV-related resources. As compensation for participating in the study, students recruited from previous studies or disability services received a US$10 Amazon.com gift card, and students recruited from the introductory psychology participant pool received research credit.
Measures
Demographics form
Created by the authors for the current study, 17 items measured age, race, gender, academic standing, self-reported ADHD status, other diagnoses, and treatment history.
Barkley Adult ADHD Rating Scale–IV (BAARS-IV)
Eighteen items measured current symptoms of ADHD according to Diagnostic and Statistical Manual of Mental Disorders (4th ed., text rev.; DSM-IV-TR; APA, 2000) criteria (inattention, hyperactivity, impulsivity) in adults (Barkley, 2011). Individuals indicate how true an item is (e.g., “Don’t listen when spoken to directly”) on a 4-point scale ranging from 1 = never or rarely to 4 = very often. The reliability of this measure was excellent in past research (α = .91) and the current study (α = .95).
Weiss Functional Impairment Rating Scale, Social Subscale (WFIRS)
Nine items measured social functioning impairment (M. D. Weiss, 2000). Participants indicated the occurrence of impairment (i.e., “getting into arguments,” “trouble cooperating,” “trouble getting along with people,” “problems having fun with other people,” “problems participating in hobbies,” “problems making friends,” “problems keeping friends,” “saying inappropriate things,” and “complaints from neighbors”) on a 4-point scale, ranging from 0 = never or not at all to 3 = very often or very much. The reliability of this measure from the current study was good (α = .79).
State-Trait Anger Expression Inventory–2 (STAXI-2)
The experience, expression, and control of anger were measured on three dimensions (i.e., state anger, trait anger, and anger expression) using 57 items (Spielberger, 1999) . Individuals rated how much they identify with a statement on a 4-point scale (e.g., “I feel like yelling at somebody”) from 1 = not at all to 4 = almost always. The reliability of this scale and its subscales is good to excellent, with a median α = .87. The reliability from the current study was excellent for State Anger (α = .90), good for Trait Anger (α = .82), and good for the Anger Expression (α = .77) subscales.
Relationship Assessment Scale (RAS)
Seven items measured romantic relationship satisfaction (e.g., “How well does your partner meet your needs?”) on a 5-point scale from 1 = not satisfied to 5 = very satisfied (Hendrick, Dicke, & Hendrick, 1998). Concurrent validity was established between the RAS and the Dyadic Adjustment Satisfaction subscale (r = .80; Dinkel & Balck, 2005). The reliability of this measure has been good (α = .86), but in the current study, it was poor (α = .60).
Revised Conflict Tactics Scale Short Form (CTS2S)
A total of 20 items measured couple conflicts and partner violence (e.g., “I insulted or swore or shouted or yelled at my partner,” “I punched or kicked or beat-up my partner”) on an 8-point scale ranging from “once in the past year” to “more than 20 times in the past year,” and including “not in the past year, but it did happen before” and “this has never happened” (Straus & Douglas, 2004). This measure examines Perpetration (α = .77-.89) and Victimization (α = .65-.94). The Perpetration variable had poor reliability (α = .59) and the Victimization variable had acceptable reliability (α = .67) in the current study.
Personality Assessment Inventory–Antisocial Features (PAI-ANT)
This 24-item scale measures antisocial personality (ASP) traits (Morey, 1991). Responses are rated on a 4-point scale ranging from 1 = false, not at all true to 4 = very true (e.g., “I’ve done some things that weren’t exactly legal”). The internal consistency reliability for the PAI-ANT is good ranging from α = .84 (census sample) to α = .86 (college and clinical sample). The reliability of this measure from the current study was acceptable (α = .67).
Data Analysis
Power analyses suggested that 107 participants would be needed to detect effects (G*Power 3.1.3; Faul, Erdfelder, Lang, & Buchner, 2007). Less than 10% of the data were missing; therefore, missing data were imputed using linear interpolation methods. Several tests of homogeneity of variance were conducted. A violation of normality was observed (i.e., substantial skewness and kurtosis) for measures of IPV perpetration, IPV victimization, and state anger. To correct for this, logarithmic data transformation was conducted for IPV perpetration, IPV victimization, and state anger variables. This type of data transformation involves a logarithmic algorithm that yields more normally distributed data (Osborne, 2002); however, nontransformed data are presented in the tables. Therefore, data analysis was conducted under the assumption of normality.
Antisocial behaviors have been found to co-occur with ADHD (Black, Gunter, Loveless, Allen, & Sieleni, 2010; Gudjonsson, Sigurdsson, Adalsteinsson, & Young, 2013; Murphy, Barkley, & Bush, 2002). Thus, to examine the relation between ADHD and our dependent variables without the influence of ASP, we controlled for the effects of ASP. Specifically, ASP scores were found to be correlated with ADHD symptomology (r = .32, p < .001), and therefore ASP features were included in many of the following analyses as a covariate.
As for ADHD symptoms, data analyses were conducted in two ways: (a) summary scores where all ADHD items were added together (scores of 1-4 for each symptom), and (b) symptom counts where participants with six or more often or very often symptoms were considered “elevated ADHD,” and participants with three or fewer often or very often symptoms were considered “low ADHD.” To optimize statistical power, summary scores were used in a majority of analyses because the authors only slightly oversampled for individuals with ADHD diagnoses, and therefore grouping the participants into high and low ADHD groups would have resulted in highly discrepant ns. However, we did compare elevated and low ADHD groups in some analyses to explore the data categorically.
Results
A correlation matrix was created to examine the relations among predictor and outcome variables investigated in the current study (see Table 1). Several significant correlations were revealed.
Correlation Matrix.
Note. ASP = antisocial personality; IPV = intimate partner violence.
p < .05. **p < .01. ***p < .001.
Six multiple regression analyses were conducted to test whether ADHD symptomology predicted the outcome variables of interest, that is, (a) social impairment, (b) state anger, (c) trait anger, (d) romantic relationship satisfaction, (e) IPV victimization, and (f) IPV perpetration. ASP features were entered on the first step to exclude the effects of ASP. ADHD symptomology was entered into the equation at the second step.
The results of the six multiple regressions are as follows. (a) The hypothesis that ADHD symptomology would predict social impairment was supported. The first step of this analysis was significant, F(1, 174) = 7.38, p = .007, R2 = .04. The overall model was significant when ADHD symptomology was entered into the equation at the second step, F(2, 173) = 39.73, p < .001, R2 = .32. ADHD symptomology significantly predicted social impairment (see Table 2). (b) The hypothesis that ADHD symptomology would predict state anger was supported. The first step was not significant, F(1, 174) = 2.73, p = .101, R2 = .02. The overall model was significant when ADHD symptomology was entered into the equation at the second step, F(2, 173) = 8.03, p < .001, R2 = .09. ADHD symptomology significantly predicted state anger (see Table 3). (c) The hypothesis that ADHD symptomology would predict trait anger was supported. The first step was significant, F(1, 174) = 28.12, p < .001, R2 = .14. The overall model was significant when ADHD symptomology was entered into the equation at the second step, F(2, 173) = 34.41, p < .001, R2 = .29. ADHD symptomology significantly predicted trait anger (see Table 4). (d) The hypothesis that ADHD symptomology would predict romantic relationship satisfaction was not supported. The first step was significant, F(1, 174) = 4.18, p = .042, R2 = .02. The overall model was significant when ADHD symptomology was entered into the equation at the second step, F(2, 173) = 2.10, p = .125, R2 = .02. However, ADHD symptomology did not significantly predict romantic relationship satisfaction. (e) With respect to IPV victimization, the first step was significant, F(1, 94) = 6.56, p = .012, R2 = .07. Furthermore, the overall model was significant when ADHD symptomology was entered into the equation at the second step, F(2, 93) = 3.33, p = .04, R2 = .07. However, ADHD symptomology did not significantly predict IPV victimization; instead, the significant effect was driven by ASP traits. (f) With respect to IPV perpetration, the first step was not significant, F(1, 94) = 1.71, p = .19, R2 = .02. Similarly, the overall model was not significant when ADHD symptomology was entered into the equation at the second step, F(2, 93) = 0.85, p = .43, R2 = .02. ADHD symptomology did not significantly predict IPV perpetration.
Multiple Regression of ADHD Symptoms Predicting Social Impairment.
Note. CI = confidence interval; ASP = antisocial personality.
Multiple Regression of ADHD Symptoms Predicting State Anger.
Note. CI = confidence interval; ASP = antisocial personality.
Multiple Regression of ADHD Symptoms Predicting Trait Anger.
Note. CI = confidence interval; ASP = antisocial personality.
Next, a multiple regression analysis was conducted to determine whether social impairment, romantic relationship satisfaction, state anger, trait anger, and IPV significantly predict symptoms of ADHD. ASP traits were entered on the first step; this first step was significant, F(1, 87) = 8.19, p = .005, R2 = .08. The overall model was significant when social impairment, romantic relationship satisfaction, state and trait anger, and IPV were entered into the equation at the second step, F(7, 81) = 7.21, p < .001, R2 = .38. Social impairment and trait anger significantly predicted ADHD symptomology (see Table 5).
Multiple Regression of All Variables Predicting ADHD.
Note. CI = confidence interval; ASP = antisocial personality; IPV = intimate partner violence.
Elevated and low ADHD symptom categories were computed to examine the differences between groups with relatively high and relatively low ADHD symptomology (i.e., symptom counts of six or more vs. symptom counts of three or less). A total of 29 participants were grouped into the elevated ADHD symptomology group, and 122 participants were grouped into the low ADHD symptomology group. No significant differences emerged between the groups with respect to gender (t = 0.65, p = .522). However, significant differences between the elevated and low ADHD groups were found for comorbid diagnoses (t = 3.25, p = .003) and age (t = 3.38, p < .001). Thus, comorbidity and age were included as covariates in the categorical analyses. Seven one-way ANCOVAs were conducted between the elevated and low ADHD categories. There was a significant main effect between these two groups on trait anger (see Table 6). Given the small cell size of the elevated ADHD symptomology group (n = 29), it is notable that a significant difference was detected between the two groups. This finding suggests that the relation between ADHD symptomology and trait anger is robust.
ANCOVAs for Elevated and Low ADHD Symptom Categories and Outcome Variables.
Note. Possible scores of Social Impairment ranged from 0 to 22; possible scores of State Anger ranged from 15 to 60; possible scores of Trait Anger ranged from 10 to 40; possible scores of ASP Traits ranged from 0 to 72; possible scores of IPV victimization ranged from 0 to 56; possible scores of IPV perpetration ranged from 0 to 56. CI = confidence interval; ASP = antisocial personality; IPV = intimate partner violence.
The means, standard deviations, and η2 reflect nontransformed data, and the p values reflect data analysis conducted with transformed data.
There was a significant sex difference in IPV Victimization (p = .022).
Because of the strong relation between ADHD symptomology and trait anger, the ADHD symptom clusters and trait anger were explored further. The symptom clusters include inattention and hyperactivity-impulsivity. A multiple regression analysis was conducted to determine which symptom cluster was the strongest predictor of trait anger. ASP traits were entered on the first step; this first step was significant, F(1, 160) = 25.16, p < .001, R2 = .14. The overall model was significant when the two symptom clusters were entered into the equation at the second step, F(3, 158) = 18.99, p < .001, R2 = .27. The inattention symptom cluster was a significant predictor of trait anger (see Table 7).
Multiple Regression of ADHD Symptom Clusters Predicting Trait Anger.
Note. CI = confidence interval; ASP = antisocial personality.
Discussion
Similar to previous literature with children and adolescents, this study revealed ADHD-related impairment in important life domains, including social functioning (Frankel & Feinberg, 2002; Hoza et al., 2005; Nijmeijer et al., 2008; G. Weiss & Hechtman, 1993) and anger (Kitchens, Rosén, & Braaten, 1999; Strine et al., 2006) within a sample of emerging adults. We found that ADHD symptomology predicted social impairment and state and trait anger, and that the elevated ADHD group had significantly higher levels of trait anger than the low ADHD group. However, contrary to previous literature (Canu & Carlson, 2003; Eakin et al., 2004; Murphy & Barkley, 1996), ADHD symptomology was not related to romantic relationship satisfaction. Moreover, we did not find a relation between ADHD symptomology and IPV. The literature concerning ADHD and IPV is mixed: some studies have found significant relations between ADHD and IPV (Fang et al., 2010; Wymbs et al., 2012); however, another study did not find significance (Crane, Hawes, Devine, & Easton, 2014). Thus, ADHD in emerging adults certainly leads to serious impairment in social functioning, but the current data do not extend to romantic relationship satisfaction or IPV. It could be that the participants in our sample were so young as to not have had the opportunity for relationship dissatisfaction. That is, it is possible that 18- and 19-year-olds are still in the “honeymoon” phase of a serious romantic relationship, and have not yet encountered major relationship problems. In fact, our sample had a mean age of 18.94; whereas Murphy and Barkley (1996) had a mean age of 32, Eakin and colleagues (2004) had participants with a mean age of 38.65 and an average of 11 years of marriage, and Canu and Carlson (2003) had a mean age of 19.2 for ADHD–Combined (ADHD/C) and 20.1 for ADHD–Inattentive (ADHD/IA) subgroups. Alternatively, we controlled for ASP, and other previous research groups (Canu & Carlson, 2003; Eakin et al., 2004; Murphy & Barkley, 1996) did not, which might explain the divergent results. Finally, the measures used to assess relationship satisfaction and IPV showed low reliability in the current study (αs = .60 and .67), which may also explain our null results.
Controlling for antisocial features allowed us to examine the sole contribution of ADHD symptoms on impairment in the current article. However, examining the link between ADHD and ASP in terms of social functioning, romantic relationships, and IPV is also of interest. ASP has been linked to IPV (Fals-Stewart, Leonard, & Birchler, 2005), and given that ADHD and ASP can co-occur (Black et al., 2010; Gudjonsson et al., 2013; Murphy et al., 2002), controlling for it might be considered overly conservative. There is certainly opportunity for future research in this area.
One surprising and interesting finding was that inattention was more predictive of trait anger than hyperactivity-impulsivity. EF deficits related to impulsivity might be most obviously associated with impulsive anger expression, but perhaps inattention leads to daily frustration (e.g., often losing needed items, frequently forgetting things, being consistently disorganized) that causes sustained anger. Thus, although individuals who are impulsive may be more likely to act on anger without considering the consequences, it is possible that inattention leads to a more sustained, internal feeling of anger.
Clinical Implications
Overall, the results of the current study suggest that there is significant impairment associated with ADHD symptomology in college students, especially in terms of social functioning and anger. It may be inferred from the findings of the current study that psychosocial treatments for ADHD should target social skills and anger management. Some researchers have incorporated these components into their psychosocial treatments. Specifically, Novotni and Petersen (1999) presented social skills training as an option in the treatment of ADHD in adults, and Safren et al. (2005) included optional modules related to anger management and communication skills. Furthermore, Stevenson, Whitmont, Bornholt, Livesey, and Stevenson (2002) incorporated both social skills training (e.g., listening skills) and anger management components to a cognitive remediation program for adults with ADHD. Results from the study revealed a reduction in symptoms of ADHD and both state and trait anger, as well as improvements in self-esteem and organization at posttreatment. A review of psychological treatment for adults with ADHD indicated that a wide variety of psychosocial treatments (i.e., Cognitive Behavioral Therapy, Dialectical Behavior Therapy, Meta-Cognitive Therapy, and Cognitive Remediation Therapy) affected anger modestly (M. Weiss et al., 2008). If anger management is targeted more specifically in psychosocial treatment for adults with ADHD, it is likely that more significant gains will be made.
Although romantic relationship impairment was not evident in the current study, it might be reasonable to assume that the 18-year-olds with social impairment in the current study may become adults with marital impairment in the future. Thus, marital therapy for adults with ADHD may also be recommended. Indeed, Kilcarr (2002) advocated for marital therapy in adults with ADHD. Moreover, because research suggests that IPV evolves from relationship dissatisfaction (Aldarondo & Sugarman, 1996), which is prevalent among individuals with ADHD, it may be beneficial to incorporate IPV prevention and intervention strategies into ADHD treatment. The relation between ADHD symptoms and IPV was not significant in the current study, but the null findings may be due to the measures chosen, age of the participants and the limited number of individuals reporting clinical levels of ADHD. Thus, interventions targeting individuals with ADHD and anger problems who are at risk of IPV perpetration may still be warranted.
Limitations
Limitations of the current study should be considered when interpreting the results. Due to the small sample of individuals who reported perpetration and/or victimization of IPV, few inferences may be made about IPV. Several factors may be related to the limited endorsement of IPV perpetration and victimization. The actual prevalence of IPV may be relatively low among emerging adults, and college students in particular. In fact, a meta-analysis found values close to zero on measures of injury due to IPV in participant samples with an age range of 14 to 22 years (Archer, 2000). Alternatively, participants may have underreported perpetration and victimization of IPV due to their reluctance to report socially undesirable behaviors (Arias & Beach, 1987).
Relatedly, because the current study used a sample of college students to represent emerging adults, we might have had a higher functioning group of participants. That is, emerging adults in college may have less impairment than emerging adults who are not enrolled in college. However, one study that found greater levels of IPV among individuals with ADHD (Wymbs et al., 2012) also used a college-student sample. In addition, Straus (2004) conducted a study of IPV among university students in 16 different countries, including the United States, and found high rates of violence at all universities included. Rates of self-reported physical assault of a partner within the last year ranged from 17% to 45%, with a median of 29%, for the university students surveyed (Straus, 2004). This study suggests that the prevalence of IPV varies across universities, but that even the universities on the low end of the spectrum had approximately one in six students reporting violence.
In addition, the current study examined the impact of self-reported symptoms of ADHD on measures of impairment rather than ADHD diagnoses. This could limit the generalizability of the results, as individuals in our study did not necessarily have a diagnosis of ADHD. The online format of the study may also be considered as a limitation. Although the utilization of an online study allowed for recruitment of individuals in a wider geographic range, this methodology may have limited the researcher’s ability to monitor participant engagement, confirm participant characteristics, appropriately obtain consent, effectively debrief participants, and ensure completion of measures (Kraut et al., 2004). Next, the Cronbach’s alpha values obtained in the current study were sometimes lower than demonstrated previously. For example, in the current study, α = .60 for the RAS (Hendrick et al., 1998), whereas in a previous study, α = .86. This lowered reliability might explain some of our results.
Future Directions and Conclusions
The limitations of the current study may serve as a guide for future research. Specifically, to manage the possible underreporting of IPV perpetration and victimization, researchers may want to consider gathering corroborating data from the romantic partners of participants. In recruiting pairs of individuals, richer data may be gathered, and stronger conclusions made regarding the role of ADHD symptoms in IPV and relationship satisfaction. Relatedly, future researchers should strive to gain a deeper understanding of the relation between psychopathology and IPV among college students, especially because college students are less likely to seek treatment for psychological issues than their noncollege-attending peers (Blanco et al., 2008). It is important to understand the factors related to this serious societal problem. Furthermore, future researchers may find that it is more appropriate to use ADHD diagnostic groups categorically rather than ADHD symptoms continuously when conducting studies of this type. In fact, the recently updated diagnostic criteria for ADHD make the criteria more applicable to adults. Researchers should take advantage of this revision and selectively recruit individuals with an ADHD diagnosis. In sum, the current study found that ADHD symptoms are linked to impairment in social functioning and anger in a sample of college students, but further research is certainly warranted.
Footnotes
Acknowledgements
The authors would like to thank Brittany Lewno, Sundé Nesbit, Carolyn Hildebrandt, Cynthia Hartung, Will Canu, and Jeffrey Howard.
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 $300 from the Intercollegiate Academic Fund at University of Northern Iowa for participant compensation. However, the authors did not receive financial support for the authorship and/or publication of this article.
