Monoaminergic hypo- or hyperfunction in adolescent and adult attention-deficit hyperactivity disorder?

DAT

Comparison of DAT between ADHD patients and controls after pooling of all available in vivo investigations (n = 20; Figure 1) yielded a significant increase in STR (+6%, p = 0.043). When the comparison of DAT binding was confined to adult patients irrespective of their medication state (n = 17), no regional differences to healthy individuals were obtained. Exclusive consideration of adult, not acutely medicated patients (12 investigations) yielded a trend towards elevated DAT binding in the STR (p = 0.07), whereas separate analyses of CAUD and PUT (five investigations, each) revealed no differences to controls. Also, thalamic and mesencephalic DAT were not affected (one investigation, each). When only the medication-naïve (adult and adolescent) subjects were considered, DAT in the whole STR (four investigations; +21%, p = 0.015), as well as in CAUD (one investigation; +16%, p < 0.05) and PUT (one investigation; +19%, p < 0.05) were significantly elevated. This also held for separate consideration of adult (three investigations; +17%, p = 0.05) and adolescent participants (one investigation; +45%, p < 0.05). Although confinement to adult, acutely medicated patients yielded no significant alteration in STR (five investigations), DAT binding was significantly reduced in the only investigation, where CAUD (−16%, p < 0.05) and PUT (−25%, p < 0.05) were assessed separately. Comparison of adolescent, not acutely medicated subjects to controls yielded a significant elevation of DAT binding in the STR (three investigations; +17%, p = 0.05), but a decrease in MB (one investigation; −13%, p < 0.05). No studies on acutely medicated, adolescent patients were available.

Figure 1:

In vivo findings on DAergic neurotransmission in ADHD (DAT, D₁R, D₂R, DA synthesis [syn], DA release [rel]). The box plots give medians and interquartile ranges (25- and 50- or 75-percentiles) of the individual synaptic constituents in the investigated brain regions (CAUD, nucleus caudate; PUT, putamen; STR, whole striatum; VSTR, ventral striatum; THAL, thalamus; PFC, prefrontal cortex; FC, frontal cortex; PC, parietal cortex; CING, cingulate; AMYG, amygdala; MB, midbrain; CER, cerebellum). The circles represent the percentages of control values in the individual investigations (filled black circle, percentual difference in adult, not acutely medicated patients relative to controls significant in the original investigation with a p of at least 0.05; unfilled black circle, percentual difference in adult, not acutely medicated patients relative to controls not significant in the original investigation; filled red circle, percentual difference in adult, acutely medicated patients relative to controls significant in the original investigation with a p of at least 0.05; unfilled red circle, percentual difference in adult, acutely medicated patients relative to controls not significant in the original investigation; filled blue circle, percentual difference in adolescent, not acutely medicated patients relative to controls significant in the original investigation with a p of at least 0.05; unfilled blue circle, percentual difference in adolescent, not acutely medicated patients relative to controls not significant in the original investigation).

D₁R

On D₁R binding in ADHD patients, only one in vivo investigation was performed (Figure 1). In the assessed cohort (adult, medication-naïve subjects), a reduction was observed in CING relative to healthy individuals (−13%, p < 0.05).

D₂R

Comparison of D₂R binding between ADHD patients and controls after pooling of all available in vivo investigations (n = 6; Figure 1) yielded no significant alterations. Confinement to adults irrespective of the medication state (five investigations) yielded a significant reduction of D₂R binding in CAUD relative to controls (−4%, p = 0.015). Separate analyses of adult, acutely medicated and adult, not acutely medicated subjects showed no difference to controls in STR (two and three investigations, respectively), VSTR (one and two investigations, respectively), PC and MB (one investigation, each). This also held for separate analyses of CAUD and PUT (two investigations, each, on acutely medicated and not acutely medicated adults). Likewise, striatal D₂R was unaltered in adolescent, not acutely medicated patients (one investigation); this also held for separate analyses of CAUD and PUT (the same investigation). Studies on acutely medicated, adolescent patients were not available. This also held for studies on medication-naïve adult or adolescent patients.

DA synthesis

After pooling of all available in vivo studies on DA synthesis (n = 5; Figure 1), a reduction was observed in FC (−2%, p = 0.05). No investigations on acutely medicated, adult patients were performed. However, comparison of DA synthesis between adult, not acutely medicated patients and controls (three investigations) also revealed significant decreases in PUT (−6%, p = 0.05), whole STR (−2%, p = 0.05) and MB (−3%, p = 0.05). Moreover, in both investigations available on DA synthesis in the AMYG of adult, not acutely medicated patients, significant reductions (−17% and −20%, p < 0.05) were observed relative to controls. This also held for the only investigation of DA synthesis in PFC and FC (−46%, p < 0.05). When only medication-naïve, adult subjects were considered (one investigation), significant decreases were observed in PUT (−6%, p < 0.05) and AMYG (−16%, p < 0.05), while no alterations were found in whole STR and MB. In adolescent, not acutely medicated patients, DA synthesis was not different from controls in STR (two investigations), VSTR (one investigation), PFC (one investigation), FC (two investigations) and MB (two investigations). No studies on medication-naïve or acutely medicated, adolescent patients were available.

DA release

On DA release in ADHD patients, only one in vivo investigation was performed (Figure 1). In the assessed cohort (adult, acutely challenged with AMPH), a significant increase (+7%, p < 0.05) was observed in the STR relative to healthy individuals.

NET

NET binding was only assessed in adult, not acutely medicated patients (two investigations; Figure 2). No differences relative to healthy controls were observed in CAUD, PUT, whole STR, GP, THAL, FC, PC, CING, MB and LC.

Figure 2:

In vivo findings on 5-HTergic (SERT) and NEergic neurotransmission (NET) in ADHD. The box plots give medians and interquartile ranges (25- and 50- or 75-percentiles) of the individual synaptic constituents in the investigated brain regions (STR, whole striatum; GP, globus pallidus; VSTR, ventral striatum; THAL, thalamus; PFC, prefrontal cortex; FC, frontal cortex; PC, parietal cortex; CING, cingulate; HIPP, hippocampus; AMYG, amygdala; INS, insula; MP, midbrain/pons; LC, locus coeruleus; CER, cerebellum). The circles represent the percentages of control values in the individual investigations (filled black circle, percentual difference in adult, not acutely medicated patients relative to controls significant in the original investigation with a p of at least 0.05; unfilled black circle, percentual difference in adult, not acutely medicated patients relative to controls not significant in the original investigation).

SERT

SERT binding was only assessed in adult, not acutely medicated patients (three investigations; Figure 2). Significant decreases relative to controls were observed in STR (−8%, p = 0.05) and THAL (−6%, p = 0.05). No alterations were found in VSTR (one investigation), PFC (two investigations), FC (two investigations), PC (one investigation), CING (two investigations), HIPP (one investigation), AMYG (one investigation), INS (one investigation) and MP (three investigations). When only the medication-naïve subjects were considered (one investigation), reductions were found in STR (−25%, p < 0.05) and MP (−16%, p < 0.05), whereas no alterations were observed in THAL, PFC and FC.

Effects of age and state of medication

DAT binding was tested in acutely medicated adult and in not acutely medicated adult and adolescent patients. The performance of two-way ANOVAs (factors: “age” with the levels “adult” and “adolescent” and “medication state” with the levels “acutely medicated” and “not acutely medicated”) was feasible for CAUD, PUT and whole STR. For neither region, the differences in DAT binding among the levels of “age” and “medication state” were great enough to exclude the possibility of mere random sampling variability (0.164 ≤ p ≤ 0.727).

Also, D₂R binding was tested in acutely medicated adult and in not acutely medicated adult and adolescent patients. Two-way ANOVAs were performed for CAUD, PUT and whole STR. For CAUD, we obtained a significant effect of “age” (p = 0.034) but no effect of “medications state” (p = 0.370), with the percentage to controls in adolescents significantly exceeding the percentage to controls in adults (p [Holm–Sidak test] = 0.034). For the other regions, no effects of either “age” or “medication state” were observed (0.298 ≤ p ≤ 0.859).

DA synthesis was assessed in not acutely medicated adults and adolescents. Hence, one-way ANOVAs (factor: “age” with the levels “adult” and “adolescent”) were conducted for CAUD, PUT, STR and MB. No effects of “age” were observed (0.199 ≤ p ≤ 0.914).

Summary of findings

DAT binding was elevated in the STR of adolescent and adult medication-naïve and not acutely medicated patients relative to healthy subjects. Concurrently, in adult, medication-naïve patients, separate analyses of CAUD and PUT also showed augmented DAT compared to healthy individuals. In acutely medicated, adult patients, DAT binding in the whole STR was no longer different from controls, while separate analyses of CAUD and PUT even showed reductions relative to healthy individuals. Mesencephalic DAT was unaltered in adult, but decreased in adolescent, not acutely medicated subjects.

In contrast to DAT, striatal D₂R was unaltered in both adolescent (not acutely medicated) and adult (acutely medicated as well as not acutely medicated) patients.

In adolescent, not acutely medicated patients, DA synthesis was not different from controls in CAUD, PUT, whole STR, VST, PFC, FC and MB, whereas in adult, not acutely medicated patients, DA synthesis was reduced in PUT, whole STR, PFC, FC, AMYG and MB compared to healthy individuals. However, in medication-naïve adults, decreases were only observed in PUT and AMYG, while DA synthesis in STR and MB was normal.

In adult patients, striatal DA release was elevated relative to controls after acute challenge with AMPH.

In adult, not acutely medicated patients, D₁R was diminished in CING.

SERT was decreased in STR and THAL, but unaltered in neocortical and limbic regions as well as MP, whereas in medication-naïve subjects SERT binding was diminished in STR and MP, but unaltered in THAL and neocortex.

No changes of NET were detected in ADHD patients.

ADHD in adulthood – DA function

In the present survey, entirely medication-naïve adults presented with unaltered DA synthesis in whole STR and MB, but decreased DA synthesis in PUT and AMYG, whereas DAT binding was elevated in CAUD, PUT and whole STR. In contrast, in the not acutely medicated collective (comprising cohorts of patients who were either medication-naïve or free of medication for at least 3 days prior to in vivo imaging), DA synthesis was no longer “normal”, but diminished in whole STR and MB, as well as, additionally, in PFC and FC. In analogy to entirely medication-naïve adults, however, also in the not acutely medicated subjects, striatal DAT was elevated. In the latter group, thalamic and mesencephalic DAT were not different from controls.

An increase or decrease of neurotransmitter availability is believed to trigger compensatory sensitization and desensitization, respectively, of presynaptic transporter molecules (for review, see Best et al. 2010). Conversely, increases or decreases of synaptic neurotransmitter concentrations are assumed to elicit a desensitization and sensitization, respective, of postsynaptic receptor binding sites. In medication-naïve ADHD subjects, DA synthesis in the mesencephalic origins of DAergic fibers as well as in the target region of ascending nigral projections was not different from controls, implying “normal” DA concentrations in either site. According to the above presumption, “normal” DA levels in striatal regions should not have incurred any regional alterations of DAT binding, while decreases of DA availability should have elicited a desensitization (but not the observed sensitization) of DAT function. Unfortunately, no in vivo imaging studies are available on mesencephalic DAT in medication-naïve adults. Strikingly, however, in a cohort comprising patients, who were either medication-naïve or free of psychotropics for at least 3 months (Table 1), mesencephalic DAT was not different from controls. Together with the “normal” DA synthesis in the MB of medication-naïve adults, this implies “normal” DA function at the sites of origin of DAergic fibers. It can be hypothesized, however, that an overexpression of DAT in the STR constitutes the initial deficit in ADHD, which leads to an abnormally high reuptake of DA into the presynaptic terminal and subsequent DA shortage in the synaptic cleft, ultimately resulting in a disruption of signal transduction via postsynaptic receptor binding sites. This conclusion is in line with the mutations of DAT genes observed in ADHD (for review, see Faraone and Larsson 2019). The assumption of DA shortage in the synaptic cleft, moreover, agrees with the observation of elevated striatal D₂R binding by Ilgin et al. (2001), whose study, however, was not enclosed in the present survey on account of the fact that the used reference values in healthy controls had been acquired by a different scientific group (Ichise et al. 1998) and in a different experimental setting.

Results on DAT in medication-naïve and medication-free subjects are contrasted by the findings on ADHD patients under current medication with MPH, who displayed reductions of DAT binding in CAUD, PUT and whole STR. The decrease of DAT binding under acute MPH can be accounted for, firstly, by the competition between radioligand and MPH molecules and, secondly, by the competition between radioligand and DA molecules, who are increasingly available due to the MPH-induced inhibition of DA reuptake. Since DA release is stimulated by 5-HT_1BR (De Groote et al. 2003), and MPH has 5-HT_1BR agonistic properties (Markowitz et al. 2006), in addition, it can be surmised, that the striatal DA shortage is balanced by MPH-induced DA efflux.

The regional reductions of DA synthesis (PUT and AMYG in medication-naïve and whole STR, PFC, FC and MB in medication-free subjects), at first glance, are consistent with the assumption of DA shortage in ADHD. Interestingly, however, in a not acutely medicated cohort (comprising patients who were either medication-naïve or free of MPH for 2 y), DA release was augmented in whole STR relative to controls. Hence, the observed reductions of DA synthesis in MB, STR and the limbic and neocortical target regions of DAergic projections can also be conceived to reflect adaptations to elevated DA concentrations in the MB, which are meant to reduce the elevated DAergic input to STR, limbic system and neocortex. Moreover, the increase of striatal DA efflux is in agreement with the observed reduction of D₂R binding sites in the CAUD of acutely medicated and not acutely medicated adults, which reflects receptor desensitization in response to an increased availability of synaptic DA. Likewise, it is in line with the reduction of D₁R binding sites in the CING, which is known to receive mesocortical DAergic projections (Berger et al. 1991). Consequently, also an enhancement of striatal DA efflux cannot be discarded as the source of ADHD symptoms. A surplus of striatal DA, as a matter of fact, also would be consistent with the observed increase of DAT binding sites in striatal regions, with the increased availability of striatal DA eliciting a compensatory sensitization of DAT. The question remains to be solved, however, as to whether DA abundance is the root cause or merely an intermittent condition, counteracting antecedent acute DA shortage.

Strikingly, in entirely medication-naïve adults, DA synthesis in the whole STR was “normal”, whereas in the not acutely medicated collective (comprising cohorts of patients who were either medication-naïve or free of medication for at least 3 days prior to in vivo imaging), deficient synthesis extended from the sites of origin of DAergic neurons in SN and VTA to the STR and to the limbic and neocortical target regions of DAergic projections. It must be borne in mind that the results obtained in the individual studies on not acutely medicated patients may have been biased by the pooling of medication-naïve and medication-free subjects, whose respective portions, additionally, differed between studies. Besides, in the participants, who were currently medication-free, not only the prescribed pharmaceuticals, but also the durations of intake and withdrawal varied or were not even specified (Table 1). To render meaningful conclusions even more difficult, investigations of DA synthesis in acutely medicated patients, so far, are completely lacking. With the available data, we can only speculate on the possible effects of pharmaceutical treatments. It cannot be excluded, for example, that the recurring blockade of DAT due to long-term intake of MPH basically influenced DA regulation, with the recurring boost of synaptic DA triggering a down-regulation of DA synthesis or, even, a compensatory upregulation of DAT beyond the level of the untreated stage. Such an induction of overage DAT density might also explain the efficacy of MPH-treatment in case of striatal DA hyperfunction, since it accounts for some removal of excess DA out of the synaptic cleft.

5-HT function

In the only investigation performed on medication-naïve adults, SERT was reduced in STR and MP but “normal” in THAL as well as in PFC and FC. In two investigations on not acutely medicated subjects, however, SERT was additionally lowered in THAL, but “normal” in MP, limbic regions and neocortex. Since both studies were conducted on patients who were either medication-naïve or drug-free for at least 30 days after precedent medication with MPH, selective 5-HT reuptake inhibitors (SSRIs), the selective NE reuptake inhibitor atomoxetine or not further specified antidepressants, it is difficult to draw conclusions on SERT regulation in ADHD. The decrease of SERT in STR and MP of medication-naïve adults, however, infers adaptations of SERT densities to diminished 5-HT levels in the sites of origin as well as in the target region of 5-HTergic projections. The decline of 5-HT reuptake in response to synaptic 5-HT shortage is in agreement with the assumption of a 5-HT deficit as underlying cause of ADHD (for review, see Banerjee and Nandagopal 2015).

Motor function is controlled by the direct and indirect DAergic pathways: in the direct pathway (STR – pars reticulata of the SN/internal GP), DA disinhibits GABAergic neurons, which results in the activation of mesencephalic, diencephalic and brainstem motor centers. In turn, in the indirect pathway (STR – external GP/subthalamic nucleus – pars reticulata of the SN/internal GP), GABAergic neurons are inhibited by DA, leading to a suppression of motor activity (for review, see Grillner and Robertson 2015). Hence, decreased as well as increased availability of synaptic DA in the STR primarily incur a dysfunction of GABA, which – by differentially affecting the individual control centers of direct and indirect pathway – may result in a net enhancement of motor activity. The raphe nuclei send 5-HTergic projections to the regions of the direct and indirect pathways, but also to VTA, VSTR, limbic system and neocortex (for review, see Hornung 2003) and, thus, modulate motor behavior (for review, see, e.g., Kawashima 2018) as well as sensory (for review, see, e.g., Jacob and Nienborg 2018), emotional (for review, see, e.g., Stein and Stahl 2000; Yagishita et al. 2020) and cognitive processing (for review, see, e.g., Bacque-Cazenave et al. 2020). In adult, not acutely medicated patients, DA synthesis was reduced in PUT, whole STR, PFC, FC, AMYG and MB compared to healthy individuals. In numerous regions of the nigrostriatal and mesolimbocortical pathways, 5-HT was found to facilitate DA efflux via 5-HT_2AR action (STR: Lucas and Spampinato 2000; VSTR: Yan 2000; PFC: Bortolozzi et al. 2005; FC: Gobert and Millan 1999). Thus, an underlying 5-HT shortage in MB and STR – in conjunction with the overexpression of striatal DAT – might account for a decreased availability of DA in striatal and mesolimbocortical regions. On the other side, in rats with central 5-HT depletion, both horizontal (locomotion) and vertical activity (rearing) were reduced (Dringenberg et al. 1995). Besides, central 5-HT depletion impaired attentional performance and enhanced impulsive responding in a five-choice serial reaction time task (Harrison et al. 1997). Hence, alternatively, the reduction of synaptic 5-HT – again in conjunction with the overexpression of striatal DAT – might also be regarded as an adaptory mechanism aiming to curb DA abundance and the associated symptoms of ADHD.

ADHD in adolescence

So far, merely three in vivo imaging studies have been conducted on DAT in adolescent patients. The body of evidence (Table 1 and Figure 1) consists of one investigation on entirely medication-naïve subjects (solely of STR, yielding a significant increase), a second investigation on subjects, who were either medication naïve or free of MPH-treatment for 7 days (of STR yielding no difference to controls, and of MB yielding a significant decrease) and a third one on subjects, who were MPH-naïve and free of psychotropic treatment (psychotropics not further specified) for 6 months (solely of STR, yielding a significant increase). DA synthesis was assessed in two investigations with the first cohort consisting of patients either medication-naïve or free of stimulant-treatment (stimulants not further specified) for 14 days and the second cohort consisting of patients, who were free of MPH-treatment for 7 days, with both groups showing no alterations of striatal DA synthesis relative to controls, but the former displaying a significant augmentation and the latter a significant decline of DA synthesis in MB (in total, resulting in no significant difference; Figure 1).

Consistent with the findings on adults, the studies on not acutely medicated adolescents (either medication-naïve or, at least, MPH-naïve and free of psychotropics for 6 months) yielded significant increases of DAT in the whole STR. Thus, in analogy to adults, an overexpression of striatal DAT can be surmised, which incurs abnormally high DA reuptake, and, thus, shortage of DA in the synaptic cleft, ultimately resulting in the disruption of signal transduction on the postsynaptic side.

Striatal DAT was no longer different from controls after precedent treatment with MPH, notwithstanding a 7-day withdrawal. Again, the decline of DAT binding can be accounted for by increased competition between, firstly, radioligand and MPH molecules, and, secondly, between radioligand and DA molecules increasingly available in the synaptic cleft. It is noteworthy, however, that, in adults, “normal” DAT binding in the whole STR was only achieved under acute MPH.

Interestingly, adolescents, who were either medication naïve or free of MPH-treatment for 7 days displayed “normal” DAT in STR, but diminished DAT in MB, which was also not the case in not acutely medicated adults. Accordingly, in a different group of adolescents free of MPH for 7 days, striatal DA synthesis was “normal”, whereas mesencephalic DA synthesis was reduced. Thus, it may be inferred for adolescents, firstly, that DA concentration was lowered at the site of origin of DAergic fibers and that, secondly, the reduced availability of DA was compensated by a desensitization of DAT. In turn, both DA synthesis and DAT binding were “normal” in the target region of nigrostriatal projections. This, consistently, also held for striatal D₂R.

Unfortunately, so far, no in vivo imaging studies on DAT function and DA synthesis are available on entirely medication-naïve as well as acutely medicated adolescents. Moreover, no investigation on DA release has been conducted in adolescents, as of yet. For the time being, however, it may be concluded that, in adolescents, the DAergic dysfunction is limited compared to adults with striatal DA synthesis still intact. Apparently, in adolescents, the sensitization of DAT is sufficient to normalize striatal DA levels without requiring a compensatory curbing of DA synthesis. Striatal DA neurons possess autoreceptors of the D₂R subtype, which are localized at the presynaptic terminal and modulate DA synthesis and release via inhibitory feedback loops (for review, see Chiodo et al. 1995). In the present survey, D₂R binding showed a significant effect of “age”. Hence, a key function might be ascribed to D₂ autoreceptor function in adolescents: possibly, the higher amount of D₂R (including D₂ autoreceptor) binding sites in adolescents compared to adults accounts for a higher capacity to adapt to shifting DA concentrations already at the synaptic level.

Appraisal

As of yet, ADHD in adults can be characterized by (1) enhancement of striatal DA reuptake, (2) decline of mesencephalic, striatal and neocortical DA synthesis, (3) enhancement of striatal DA release, (4) decline of D₁R and D₂R function in limbic and striatal regions, respectively, and (5) decline of striatal and mesencephalic/pontine 5-HT reuptake. In adolescents, ADHD can be characterized by (1) enhancement of striatal DA reuptake, (2) decline of mesencephalic DA reuptake (in subjects, who were either medication naïve or free of MPH for 7 days) and (3) decline of mesencephalic DA synthesis (also in subjects, who were either medication naïve or free of MPH for 7 days). The medication with MPH can be assumed to relieve the shortage of (primarily) DA by normalizing the excessive DA reuptake as well as by facilitating DA efflux in the STR. A compensatory induction of overage DAT expression might also account for the efficacy of MPH in case of postsynaptic DA hyperfunction.

Mesencephalic DAT was unaltered in adult, but decreased in adolescent patients, while, contrarily, striatal and mesencephalic DA synthesis were unaltered in adolescents, but decreased in adults. This suggests that, in adolescent ADHD, the shortage of mesencephalic DA is compensated at the synaptic level by the desensitization of DAT. In contrast, in adult ADHD, DA shortage (or DA surplus) in the STR are due to (or compensated by) feedback input to the origin of DAergic fibers. This functional detour – likely via neocortex and limbic system – may be related to behavioral and cognitive coping strategies evolving with increasing disease duration. Moreover, it may account for the age-dependent predominance of different key symptoms (hyperactivity/impulsivity in childhood/adolescence versus inattentiveness and emotional problems in adulthood; Hirsch et al. 2018; Larsson et al. 2011).

Generally, against our mode of analysis the objection can be raised that the pooling of all in vivo imaging studies of a given synaptic constituent might lead to flawed results, since all available investigations were included irrespective of their “quality”. Ideally, any investigator should confine his or her study to patients of the same sex and age, the same disease duration, the same disease severity and the same key symptoms (either hyperactivity/impulsivity or inattention). Besides, all patients should be medication-naïve, non-smokers and otherwise healthy. Table 1 shows that reality looks quite different. Therefore, the conjecture of region-specific alterations of receptor and transporter functions in the individual conditions of adult and adolescent, as derived from the available in vivo imaging studies, must be put into perspective. Firstly, in the majority of investigations, medication naïve and medication-free patients were pooled (Table 1). Likely, the inherent inconsistency of the data set account for the failure to detect any effect of the medication state on the outcome of in vivo imaging studies in the present survey. Secondly, only a limited amount of investigations is available on individual constituents of the DAergic, 5-HTergic and NEergic synapse. In fact, numerous findings were obtained in merely onea investigation (Table 1). As evidenced by the percentile ranges in Figures 1 and 2, results on the individual synaptic constituents were widely discrepant, which renders it difficult to gauge the relevance of one single finding without any knowledge about the possible range of alterations relative to healthy controls. This, in particular, holds for the findings on D₁R, DA release and SERT in adults and DAT as well as DA synthesis in adolescent patients. Thirdly, the individual findings underlying the obtained differences between ADHD patients and controls were inconsistent, with reports of significant and non-significant increases and decreases of radioligand binding (Figures 1 and 2). In the three studies on adolescent patients, the inconsistency of findings may be accounted for by the fact that the mean ages of patients in the individual studies (10, 14 and 17 y; Table 1) were widely different; besides, in one study, binding data in adolescents (mean age: 14 y) were compared to binding data obtained in adult controls (mean age: 30 y). In both adults and adolescents, inconclusive results may, furthermore, be due to varying comorbidities, varying disease durations and varying disease severities, which, in addition, were assessed with a variety of different inventories (Table 1). Moreover, the individual findings may be confounded by the assessment of brain regions, which express only limited amounts of the respective transporter or receptor subtypes (e.g., DAT in MB or D₂R in neocortex) or whose small sizes fall short of the spatial resolution of the employed imaging systems (e.g., raphe nuclei or AMYG).

In vivo evidence is conflicting as to either general monoaminergic hypofunction or hyperfunction in ADHD. For the time being, it might be helpful to, rather, assume regional stages of synaptic DA shortage or DA surplus, with the former due to increased DA reuptake and/or decreased DA synthesis, and the latter due to increased DA efflux. It may be hypothesized that these stages are transient and elicit presynaptic as well as postsynaptic compensatory mechanisms, striving to approach functional homeostasis as closely as possible. Moreover, an interrelation with associated brain regions and other neurotransmitter systems can be inferred, which may, additionally, depend on the stage of disease (adolescent vs adult).

In an intriguing dynamic PET study (Badgaiyan et al. 2015), D₂R binding data were acquired both in a state of rest (tonic release experiments) and during performance of Eriksen’s flanker tests (phasic release experiments). In the resting condition, D₂R binding was significantly higher in the right CAUD of ADHD patients, suggesting reduced tonic DA release compared to controls. During the performance of the response inhibition tasks, D₂R binding in the right CAUD was significantly lower, indicating enhanced phasic DA release relative to healthy individuals. This outcome pinpoints an underlying methodological problem: in the study of Badgaiyan et al. (2015), D₂R binding was reduced (indicative of increased competition between radioligand and DA molecules for D₂R binding sites and/or D₂R desensitization in response to the increased availability of synaptic DA) during cognitive activation, but elevated (indicative of decreased competition between radioligand and DA molecules for D₂R binding sites and/or D₂R sensitization in response to the decreased availability of synaptic DA) in the state of rest, whereas all other in vivo imaging studies of DAT, D₂R and DA synthesis de facto were conducted under resting conditions, but not unanimously provided evidence of DA hypofunction in ADHD. If the manifestation of a general DAergic hypo- or hyperfunction in the STR of ADHD patients, truly, be related to the state of activation during the performance of in vivo imaging studies, further investigations must be conducted under controlled conditions of resting and cognitive activation in adult and adolescent as well as in medicated and not (acutely) medicated patients in order to scrutinize this matter.

Studies on animal models of ADHD have evidenced an interdependence with age: DA concentration in STR and PFC were increased in juvenile spontaneously hypertensive rats (SHR), but lowered in mature animals (for review see Viggiano et al. 2004). Furthermore, in juvenile SHR, D₂R densities were elevated in pre- and infralimbic cortices as well as CING (Kozlowska et al. 2019). Comparison of SHR and Wistar-Kyoto controls revealed a lower density of D_2/3 autoreceptors in the VSTR and a higher density of D_1/5R in dorsal STR, VSTR and olfactory tubercle of the former, implying anterior hypo- and posterior hyperfunctioning in the mesolimbocortical system (Papa et al. 2002). Thus, basically, in human ADHD, it remains to be further explored, exactly which synaptic DAergic (and 5-HTergic) functions are hypo- or hyperactive in the context of motor, cognitive and emotional functioning after a given duration of disease and/or pharmacological treatment. Thereby, we, first and foremost, are in need of supplementary evidence on regional DA release as well as DA and 5-HT receptor binding in both adults and adolescents.

It must also be taken into account that glutamate (GLU) and GABA are increasingly implicated in the pathobiochemistry of ADHD (for review, see Purkayastha et al. 2015). An increased concentration of GLU was found in the CING of adult ADHD patients (Bauer et al. 2018). Moreover, GLU release was elevated in STR and PFC of the SHR (Miller et al. 2014), while chronic MPH decreased the amplitude of phasic GLU signaling (Miller et al. 2019). In turn, diminished GABA levels were found in primary sensory and motor cortices (Edden et al. 2012) as well as in the STR (Puts et al. 2020) of children with ADHD. Likewise, SHR displayed reduced tonic levels of GABA in the HIPP (Sterley et al. 2013). It is, thus, evident that the patterns of neurochemic dysfunction obtained in the present survey of in vivo imaging studies are preliminary. Future efforts must be directed towards procuring in vivo evidence on GLU and GABA receptor function in adolescent and adult patients with ADHD, which may either reveal the source underlying DA and/or 5-HT dysfunction(s), or even entirely shift the focus as to the causation of ADHD.