The relationship between attention-deficit/hyperactivity disorder (ADHD) and aggressive behaviour has been the subject of considerable research involving animals and humans (1–9). Evidence suggests that neurobiological and environmental factors contribute to different phenotypes of aggression as represented within the diagnosis of ADHD, conduct disorder (CD) and antisocial personality disorder (ASP) (10), and with oppositional defiant disorder (ODD) also being a part of the ADHD–aggression spectrum.
It is important to outline that variants of aggressive behaviour in ADHD may also be influenced by different developmental and neurobiological trajectories, in particular in view of the frequently observed co-morbidity of conduct problems with ADHD. In view of one of the most replicated findings in biological psychiatry one has to note that the serotonin deficiency hypothesis of aggression explains considerable variance (5, 6, 8, 9, 11–15). However, so far there are only few studies dealing with the serotonin deficiency hypothesis of aggression in children and adolescents, in particular with ADHD and related co-morbid conditions such as CD or ODD (16–24). With lower serotonergic responsivity assessed in childhood predicting the development of ASP in early adulthood these findings underline that this particular trajectory strongly correlated with neurobiological markers of aggression such as changed serotonergic neurotransmission, but to date there is still a lack of studies in this field as other trajectories (i.e. in terms of hormonal changes) as well as interactions between hormonal differences and altered serotonergic functioning may also be taken into consideration (25–28).
One has to note that recent research has suggested that trait-impulsivity and aggressiveness in ADHD could be potential moderators as far as serotonin-dependent aggressive behaviour in children and adolescents is concerned (1, 29–35), which also stands in relationship to the mentioned trajectory with serotonergic responsivity influencing the development from ADHD to ASP. In particular, with low impulsive patients being more susceptible to behave aggressively under a diminished central nervous serotonergic turnover one could speculate that this relationship could outline one part of a trajectory from childhood to early adulthood of serotonin-dependent aggression, and with trait-impulsivity moderating the vulnerability to develop aggressive behaviour (1) or even ASP.
In line with this, with low resting heart rate being a strong predictor of aggressive and conduct problems in children and adolescents, low impulsive children and adolescents with ADHD were shown to have a lowered heart rate under a diminished serotonergic neurotransmission (30, 36). Interestingly, low aggressive boys with ADHD also showed increased physiological reactivity along with increasing provocation of aggression (37). However, treatment with methylphenidate increased the heart rate but did not interact with the aggressive response in this study (37). Additional research indicated that boys with co-morbid ADHD/CD/ODD had higher levels of behavioural aggression along with a heightened heart rate acceleration, and that these boys were rated as angrier than all other boys compared with boys with only ADHD or ODD/CD (38). Finally, trait-aggression in terms of hostility also influenced serotonin-dependent behavioural inhibition in male children and adolescents with ADHD (34). However, a major limitation is that data on aggression in females with ADHD are scarce, underlining the need for ongoing research involving subjects of both genders and different age groups. Taken together the discussed data suggest that trait-impulsive and -aggressive characteristics along with physiological parameters (i.e. such as heart rate) could be of particular value in view of serotonin-dependent aggression from a developmental viewpoint, and could also map the pathway for research aiming to distinguish between potential endophenotypes as far as the relationship between ADHD, aggression and changed serotonergic neurotransmission is concerned.
The article by Dowson and Blackwell in the present issue clearly adds an important part of the puzzle to the ADHD–aggression relationship, in particular as such data investigating co-variations between ADHD and aggression, a scenario frequently observed in clinical settings, are currently missing (39). Moreover, DSM-IV domains of ADHD were shown to predict co-morbid impulsive aggression, with a significant overlap in view of specific features of borderline personality disorder (BPD) (39). Noteworthy is the fact that data in adults as well as in female subjects on this matter are currently lacking. In view of epidemiological differences in prevalence rates of BPD in males and females future research should take this bias into careful consideration. However, despite these limitations taken together these findings by Dowson and Blackwell have major implications for the selection of future research samples aiming to investigate aggressive behaviour in adults with ADHD (39).
In summary I believe that investigations with larger sample sizes and both genders are needed in order to further tap into the relationship between ADHD and co-varying aggressive behaviour, in particular with respect to changed serotonergic neurotransmission. This relationship could be of particular importance when it comes to potential endophenotypes in ADHD and patients with co-varying conduct problems, and also if one aims to further disentangle the frequently observed diagnostic overlap between ADHD and aggressive behaviour, in particular in view of impulsivity as a potential moderator variable. Furthermore we strongly need to focus on particular types of aggressive behaviour, i.e. such as proactive vs. reactive aggression and how these data are obtained (i.e. within a laboratory setting or within real-life interactions). Finally, as mentioned before, research aiming to involve both genders is highly mandatory as gender differences, i.e. as represented within rather indirect forms of aggression in females, are very likely to have a major impact on the data aiming to assess aggressive behaviour. These methodological issues are crucial and should be addressed in future investigations.