SHEN ET AL The hypothesis is that thes e small to mediun dopamine on the prefrontal region(59-62).a 2018 study However,as shown in the results,these findings were sta- stically gray ma ngs ma olume of the posterior occipital cluster as an intermediate Our comparison between medicated and never medica phenotype of ADHD,especially the inattention symptoms. ted patients with AD may sugge ati pos ure associated ADID Stimulant me s the brain in ical manifold ways,including in its structure (64.65)and function(66)and in neurotransmission(67).Therefore, of future inattention symptoms;and5)is preserved inclinical lower gray mae f th DHD and cannot b e simply explain Dy a nla tha ed fo ntermediate phenotyne of a mental health disorder (27) matter volumes bet n medicated pa develops early in life and functionally and typ lly dev ol subj suggest th 77, nd th .on ADHD.which of ADHD.As ement in visual attention by we also found that the ADHD-associated prefrontal cluster stimulant treatment for ADHD (68). was indeed correlated with delay dis ounting,which is ual-path HD has been ava nsisten vith the A 078 11 ith independent-pathways model (ie.the cognitive circuit be degree ofindividual variation (80).itma he m ntal cortex and orbito m⑤bti this tudy should be and motivation ADHD.Thisin m e and not only is supported by the associations between working clinical diagnosis.our ing and between intrasubjec fully captured ADHD symptoms. This study identified a g in the IM ample bu e ate for wor attennor supp hanced oitive hias-enhanced mahenoioc2,&rminStrie in ADHD.However,the study did not identify any significant this interaction atter netwo eciall in th are in a but the de ital and reported in task-based NR ments (70) gely completed during Its association with delay discounting is not com letel imaging in an ADHD cohort during childhood will be nec surprising given that both hyperactivation(71,72)and higher essary to confirm whether any abnormal development of this gray matter ave cortex can be observed leading to the manifestation immediate reward its association with workine performance (i.e.,2-back a ccuracy)has also been observed in an fMRI CONCLUSIONS at th ual path ays in A ted t which athways for the dev ment of ADHD Given that the sterior occipital region develops and matures much earlie envir rected tasks(5) bee even a sm se re sizes of the identified gosis and p ntion strategies for ADHD. .September 2020 The hypothesis is that these regions may be modulated by dopamine activity. In addition to the well-known effect of dopamine on the prefrontal region (59–62), a 2018 study found that the dopamine transporter gene (DAT1)-related reduction of gray matter volume in the left posterior occipital region may contribute to visual memory performance in children with ADHD (63). Our comparison between medicated and never￾medicated patients with ADHD may suggest a positive effect of medication on gray matter atrophy in patients with ADHD. Stimulant medication for ADHD affects the brain in manifold ways, including in its structure (64, 65) and function (66) and in neurotransmission (67). Therefore, lower gray matter volumes of the identified clusters in never-medicated patients with ADHD may exclude one alternative explanation, namely, that lower gray matter volumes were caused secondarily by medication, while comparable gray matter volumes between medicated pa￾tients and typically developed control subjects suggest that medication may have a remedial effect on brain structure in ADHD, which provides a possible neuroanatomical basis for the behavioral improvement in visual attention by stimulant treatment for ADHD (68). The dual-pathway model of ADHD has been a valuable model for our understanding of the neuropsychopathology of this disorder (5–7). Our findings do not support the independent-pathways model (i.e., the cognitive circuit be￾tween dorsolateral prefrontal cortex and dorsal striatum and the motivational circuit between orbitofrontal cortex and ventral striatum) (5) but instead demonstrate an interaction between cognition and motivation in ADHD. This interaction not only is supported by the associations between working memory and delay discounting and between intrasubject variability and delay discounting in the IMAGEN sample but also is supported by previous reports of both monetary incentive–enhanced cognition and cognitive bias–enhanced avoidance motivation (69). Our findings further suggest that this interaction may have its neural basis within the visual attention network, especially the left cuneus in the posterior occipital cortex, and hyperactivation of this region has pre￾viously been reported in task-based fMRI experiments (70). Its association with delay discounting is not completely surprising given that both hyperactivation (71, 72) and higher gray matter volume (73) of the posterior occipital cortex have already been associated with choosing delayed gain over immediate reward. Its association with working memory performance (i.e., 2-back accuracy) has also been observed in an fMRI experiment (74). Together, these findings may suggest that the dual pathways in ADHD are likely related to dysfunction of these cognitive and motivational processes, particularly in the visual attentional system, which supports the top-down selection of relevant information from the environment during goal-directed tasks (75). It has been reported that even a small improvement in memory score (10%) can make a significant difference in school performance (76). The effect sizes of the identified neural associations in our study were small to medium, partially owing to the multifactorial nature of ADHD. However, as shown in the results, these findings were sta￾tistically robust and empirically replicable using an in￾dependent clinical sample. Therefore, the findings may improve the accuracy of the diagnosis by using gray matter volume of the posterior occipital cluster as an intermediate phenotype of ADHD, especially the inattention symptoms. This neuroanatomical feature 1)is associated withinattention in the general population; 2) is associated with neuro￾psychological endophenotypes of ADHD; 3) is associated with genetic risk for ADHD; 4) contributes to the explanation of future inattention symptoms; and 5) is preserved in clinical patients with ADHD and cannot be simply explained by a confounding effect of medication. These are exactly the lines of evidence that are required for the identification of an intermediate phenotype of a mental health disorder (27). The occipital cluster develops early in life and functionally matures during childhood (77), and therefore it may also be used as a neuroimaging biomarker of disrupted brain de￾velopment for the early diagnosis of ADHD. As expected, we also found that the ADHD-associated prefrontal cluster was indeed correlated with delay discounting, which is consistent with the frontostriatal model of ADHD (78). However, given that this prefrontal area is under significant development during adolescence (79), with a significant degree of individual variation (80), it may be difficult to use gray matter volume of this prefrontal cluster as an imaging trait marker for ADHD. Several limitations of this study should be mentioned. Because both parent and teacher ratings are needed for clinical diagnosis, our study using parent ratings may not have fully captured ADHD symptoms. This study identified a common neural correlate for working memory—attention regulation and delay discounting—which represents partic￾ular aspects of the broader cognitive and motivational deficits in ADHD. However, the study did not identify any significant association of response inhibition with the ADHD symptoms. Our findings are in adolescents, but the development and maturation of the posterior occipital cortex are believed to be largely completed during childhood. Longitudinal neuro￾imaging in an ADHD cohort during childhood will be nec￾essary to confirm whether any abnormal development of this occipital cortex can be observed leading to the manifestation of ADHD symptoms. CONCLUSIONS Using a comprehensive approach, we revealed a common neuroanatomical correlate of both cognitive and motivational pathways for the development of ADHD. Given that the posterior occipital region develops and matures much earlier than the prefrontal areas previously focused on in the frontostriatal model of ADHD, these results may provide new clues to discovering novel imaging markers for early di￾agnosis and preemptive intervention strategies for ADHD. Am J Psychiatry 177:9, September 2020 ajp.psychiatryonline.org 851 SHEN ET AL