Abstract
Commonly comorbid early onset psychiatric disorders might reflect the varying expression of overlapping risk factors. The mediating processes remain poorly understood, but three factors show some promise: adolescent externalizing traits, early life adversity, and midbrain dopamine autoreceptors. To investigate whether these features acquire greater predictive power when combined, a longitudinal study was conducted in youth who have been followed since birth. Cohort members were invited to participate based on externalizing scores between 11 to 16 years of age. At age 18 (age 18.5 ± 0.6 y.o.), 52 entry criteria meeting volunteers had a 90-min positron emission tomography scan with [18F]fallypride, completed the Childhood Trauma Questionnaire, and were assessed with the Structured Clinical Interview for DSM-5. The three-factor model identified those with a lifetime history of DSM-5 disorders with an overall accuracy of 90.4% (p = 2.4 × 10−5) and explained 91.5% of the area under the receiver operating characteristic curve [95% CI: .824, 1.000]. Targeting externalizing disorders specifically did not yield a more powerful model than targeting all disorders (p = 0.54). The model remained significant when including data from participants who developed their first disorders during a three-year follow-up period (p = 3.5 × 10−5). Together, these results raise the possibility that a combination of temperamental traits, childhood adversity, and poorly regulated dopamine transmission increases risk for diverse, commonly comorbid, early onset psychiatric problems, predicting this susceptibility prospectively.
Introduction
Converging epidemiological [1,2,3,4,5] and molecular genetic evidence [6,7,8] raises the possibility that many psychiatric disorders reflect the varying expression of overlapping developmental trajectories [4, 9]. One of the largest trajectories is characterized by diverse externalizing (EXT) behaviors [5, 10,11,12], including poor impulse control, dysregulated affect, and altered responses to rewards and punishments [13, 14].
EXT traits are likely mediated by multiple neurobiological systems [15, 16]. One implicated system has shown associations between low midbrain dopamine autoreceptor levels [17, 18], increased dopamine release [19, 20], and various EXT-related features, including impulsivity, sensitivity to punishment and novelty seeking [20, 21]. In laboratory animals too, low midbrain autoreceptors and elevated dopamine transmission can increase the salience of both positive and negative stimuli [22,23,24,25,26,27,28,29,30], promote novelty seeking and motor impulsivity [16, 28,29,30,31], and reduce the ability to disambiguate optimal choices [32,33,34]. Stressful events can aggravate these features, including the fomenting of dopaminergic [35, 36] and behavioral hyperreactivity [35, 37] and susceptibility to mental health problems [38, 39].
Prompted by the above findings, we tested, in young adults who have been followed since birth, whether the combination of all three factors (EXT, childhood trauma, dopamine regulation) predicts clinical outcomes. It was hypothesized that higher EXT traits, greater early life adversity, and lower midbrain dopamine autoreceptor availability would predict the presence of early onset commonly comorbid psychiatric conditions.
Methods
Participants
All participants were born between 1996 and 1998, lived in the Montreal or Quebec City area in Canada, and had been followed since birth [40,41,42]. Fifty-eight volunteers (36 F/22 M) underwent magnetic resonance imaging (MRI) and [18F]fallypride positron emission tomography (PET) scans. Six were excluded from the current analyses: two due to not having Childhood Trauma Questionnaire (CTQ) data and four due to BPND values being outliers (±3 SD) (Table 1 and Supplementary Table 1). The final sample consisted of 52 participants (30 F/22 M), mean age at the time of scanning was 18.5 (SD = 0.6 years). Forty-one participants (22 F/19 M) had one or more follow-up interview (age at last assessment: 21.0 ± 0.9 years). Ethics approval was obtained from McGill University and Sainte-Justine University Hospital Research Ethics Boards, and all participants provided written informed consent. Further details are in Jaworska et al. [21].
Externalizing (EXT) scores
Externalizing trait scores were measured annually between ages 10 and 16 through self-report (Quebec Longitudinal Study of Child Development; QLSCD, n = 53) or teacher ratings (Quebec Newborn Twin Study; QNTS, n = 5) using the Social Behavior Questionnaire (SBQ) [43, 44]. Mean scores were calculated for the following SBQ subscales: hyperactivity, impulsivity, oppositional behavior, non-aggressive behavioral problems, physical aggression, proactive aggression, indirect aggression and reactive aggression. Composite EXT trait scores were aggregated using a minimum of two years’ data between 10 and 16 years [5, 40]. Mean EXT scores during these years correlated with those obtained earlier in life (1–5 and 6–10 years) [40]. For the current study, cohort members were invited to participate if they scored in either the top or bottom 30% of EXT trait scores, as established in the first wave of QLSCD cohort members (n = 242, born in 1996) [40].
Assessments
At their index assessment prior to the PET and MRI scans, all participants were interviewed face-to-face with the Structured Clinical Interview for DSM-5 (SCID) [45] and administered the CTQ [46]. Follow-up telephone assessments were then conducted annually for the next three years. The CTQ measures early life adversity, including emotional abuse, physical abuse, sexual abuse, emotional neglect and physical neglect [46]. It has been validated in French [47] and has good psychometric properties [48]. Elevated CTQ scores are associated with poor cognitive function across multiple domains [49] and diverse psychiatric disorders [50, 51].
Magnetic resonance imaging (MRI)
MRI scans were acquired on each participant for anatomical co-registration using a 3 T Siemens Trio TIM scanner (McConnell Brain Imaging Center, Montreal Neurological Institute) with a Magnetization Prepared Rapid Acquisition sequence (slice: 1 mm, TR: 2300 ms, TE: 3.42 ms, flip angle: 9°, FOV: 256 mm, Matrix: 256 × 256). MRI and PET data were coregistered to aid anatomical precision when defining the regions of interest (ROIs).
PET imaging
Dopamine D2 receptor levels were measured using the high affinity dopamine D2/D3 receptor ligand, [18F]fallypride, and a 90-minute PET scan on a high-resolution research tomograph (HRRT). Following cannula insertion into the left antecubital vein for tracer administration, a 6-min 137Cs transmission scan was obtained for attenuation correction. The [18F]fallypride tracer (prepared as previously described [19]) was administered as a 1-min intravenous bolus, with emission scans acquired concurrently in list mode over 90 minutes (participants were instructed to remain awake). The mean ± SD injected [18F]fallypride dose was 123.39 ± 7.59 MBq, injected mass was 1.46 ± 3.90 nmol, and molar activity at time of injection was 381.08 ± 819.99 GBq/µmol. There were no statistically significant differences in injected dose, mass or molar activity between participants with and without psychiatric diagnoses (p values > 0.4; Supplementary Table 2).
PET images were reconstructed using the Ordinary Poisson Ordered Subset Expectation Maximization (OP-OSEM) reconstruction algorithm (10 iterations, 16 subsets). This included correction for non-uniformities, attenuation, scattered and random coincidences, and motion. To reduce partial volume effects, resolution modeling using the point spread function was implemented in image reconstruction. Motion correction was based on a data-driven motion estimation and correction method that estimates rigid-body motion between dynamic frames. Further details are in Jaworska et al. 2020 [21].
MRI and PET analyses
The imaging data set was obtained from previously analyzed images [21]. Binding potential non-displacement (BPND) values were derived from the primary region of interest (ROI), the midbrain dopamine cell body region, consisting of the ventral tegmental area and substantia nigra, and seven exploratory regions of interest (ROIs) further implicated in the regulation of mood and motivational states and impulsive behaviors (superior frontal gyrus, medial frontal gyrus, medial orbito-frontal gyrus, middle frontal gyrus, insula, amygdala, and hippocampus) [21]. ROI masks were defined using standard masks on the MNI152 template which were coregistered to each individual’s MRI scan using linear and nonlinear transformations [52]. These ROI masks were then applied to each summed PET image using nonlinear co-registration. Time-activity curves were extracted from each ROI in native PET space using tools developed by the Turku PET Center (http://www.turkupetcentre.net/). BPND values (i.e., equilibrium ratio of specifically bound to non-displaceable radioligand in tissue) were derived from ROIs using the simplified reference tissue model [53] with cerebellar gray matter as the reference region.
Statistical analyses
Binomial logistic regression analyses were conducted to test whether midbrain [18F]fallypride BPND values, EXT and CTQ scores, converted to Z scores, predicted the presence of lifetime DSM-5 diagnoses. All analyses were run for diagnoses obtained at the time of or prior to the PET scan and again including diagnoses obtained during the follow-up interviews.
Linearity of the continuous variables with respect to the logit of the dependent variable was assessed via the Box-Tidwell (1962) procedure [54]. A Bonferroni correction was applied using all seven terms (midbrain BPND, EXT, CTQ, midbrain*ln_midbrain, EXT*ln_EXT, CTQ*ln_CTQ, and constant) in the overall model resulting in statistical significance being defined as p < 0.0071 [55]. Based on this assessment, all continuous independent variables were linearly related to the logit of the dependent variable.
Receiver operating characteristic (ROC) curves were created for each model, and the area under the curve (AUC) was used to evaluate the overall model’s strength. ROC curves illustrate the ability of a binary classifier to discriminate groups as the threshold is varied. The graph is created by plotting the true positive rate or sensitivity on the y axis, and the false positive rate or (1—specificity) on the x axis. The area under the ROC curve measures how well the model can identify positive and negative cases. The AUC ranges between 0 and 1. The closer the AUC is to 1, the better the model. An AUC close to 1 is able to correctly identify patients with and without diagnoses, an AUC close to 0 identifies group designation incorrectly, and an AUC close to 0.5 indicates that the model has no classification ability [56].
All hypotheses were tested using two-tailed statistics. Unless otherwise noted, all analyses were performed using IBM SPSS version 26 with the following add-ons: Complex Sampling & Testing, Forecasting & Decision Trees, and Custom Tables & Advanced Statistics.
Results
Lifetime psychiatric diagnoses
At their index interview, 23% of participants met criteria for one or more lifetime DSM-5 disorders. By their last interview, this had increased to 31%. The specific diagnoses included substance use disorders (SUDs), unipolar mood disorders, attention deficit-hyperactivity disorder (ADHD), panic disorder, generalized anxiety disorder, adjustment disorders, dyslexia, binge eating disorder, and conduct disorder (Supplementary Tables 3A and 3B).
Three-factor model predicts lifetime psychiatric disorders at index interview
The three-factor model was statistically significant (p = 2.4 × 10−5), correctly classifying 90% of cases and explaining 56% (Nagelkerke R2) of the variance in DSM-5 diagnoses. Sensitivity was 75%, specificity was 95%, positive predictive value was 82%, negative predictive value was 93%, and ROC AUC was 92% [95% CI: .824, 1.000], considered an outstanding level of discrimination (Table 2) [57].
All three factors contributed to the model. Adding CTQ total scores as the third factor strengthened the model (compared to EXT + midbrain BPND, p = 0.05). Adding midbrain [18F]fallypride BPND values as the third factor strengthened the model (compared to EXT + CTQ, p = 0.03). Adding EXT scores as the third factor strengthened the model (compared to CTQ + midbrain BPND, p = 5.0 × 10−6). In comparison, the model was not improved by the addition of BPND values from the seven exploratory ROIs (p ≥ 0.37).
Within the three-factor model, standardized EXT scores (Wald = 9.880, p = 0.002) and midbrain [18F]fallypride BPND values (Wald = 3.833, p = 0.05) were statistically significant after controlling for the influence of the two other factors. One standard deviation (SD) increases in CTQ and EXT scores were associated with two- and 10-times greater odds of having a lifetime DSM diagnosis, respectively. A one SD decrease in midbrain [18F]fallypride BPND values was associated with three times higher odds (Table 3).
Three-factor model predicts lifetime EXT disorders at index interview
The three-factor model predicted lifetime EXT disorders alone at index interview. The model was primarily driven by EXT traits (p = 0.008). When adding midbrain BPND values and CTQ scores, the former strengthened the model (compared to EXT + CTQ, p = 0.027) while the latter did not (compared to EXT + midbrain BPND, p = 0.225). More importantly, this three-factor model for EXT disorders was not stronger than the one capturing all expressed DSM-5 disorders (AUC ROC curves for EXT vs. all disorders, t(102) = 0.6191, p = 0.54) [58], plausibly reflecting the propensity of people with EXT disorders to develop internalizing disorders and vice versa [59] (Table 4).
Three-factor model predicts lifetime psychiatric disorders at follow-up
At the last follow-up interview, the logistic regression model was statistically significant (p = 3.5 × 10−5), explained 51% (Nagelkerke R2) of the variance in DSM-5 diagnoses, and correctly classified 81% of cases. Sensitivity was 63%, specificity was 89%, positive predictive value was 71% and negative predictive value was 84%. Of the three predictor variables only EXT trait score was statistically significant (Wald = 11.948, p = 0.001), but inclusion of CTQ scores and midbrain BPND values yielded what is considered an excellent level of discrimination [57] (EXT alone, AUC ROC = 0.855 [95% CI: .713, 0.998]; EXT + CTQ + midbrain BPND, AUC ROC = 0.898 [95% CI: 0.795, 1.000]).
Three-factor model is accurate in men and women
Male participants were more likely to have diagnoses of ADHD and SUDs while female participants more frequently met criteria for mood and anxiety disorders (Supplementary Table 4). Overall, lifetime psychopathology was observed in similar proportions of males and females (6/22 vs. 10/30; no significant difference, Fisher’s exact test, p = 0.76). At the index interview, the model remained significant with the addition of sex as a fourth factor (p = 3.4 × 10−5). Sex itself was not a significant predictor of a lifetime DSM-5 diagnosis (Wald = 0.374, p = 0.54), nor was the four-factor model (EXT + CTQ + midbrain BPND + sex) significantly different from the three-factor model (p = 0.54) (Table 5). When additional diagnoses from the follow-up interviews were included, the model with sex as a fourth variable remained significant (p = 7.4 × 10−5). Again, however, sex itself was not a significant predictor of lifetime DSM-5 diagnosis (Wald = 0.843, p = 0.36) and the four-factor model (EXT + CTQ + midbrain BPND + sex) was not significantly different from the three-factor model (p = 0.35).
Discussion
Our study’s primary objective was to test whether a three-factor model, composed of features with modest transdiagnostic predictive value when tested in isolation, could more powerfully predict the presence of lifetime psychiatric disorders in young adults who have been followed from birth. As hypothesized, the combination of higher EXT traits, greater levels of childhood adversity, and lower levels of midbrain D2 receptors identified, with high accuracy and statistical robustness, participants with any lifetime history of psychiatric illness, both at the index interview and at follow-up.
To our knowledge, this is the first investigation to study all three factors together. In comparison, each individual factor has been found to increase risk for psychopathology. As a start, both epidemiological [1,2,3,4,5] and molecular genetic [6,7,8] studies suggest that EXT traits reflect a developmental trajectory from which diverse psychiatric disorders can emerge. This includes substance use and cluster B personality disorders, as well as problems commonly considered internalizing disorders [1, 4, 5, 59].
Childhood trauma is arguably the prototypical transdiagnostic risk factor, increasing the probability of both internalizing and EXT disorders in adolescence and adulthood [51, 60]. The replicated associations with early life adversity noted, some people exhibit striking resilience [61, 62]. The factors accounting for these different responses to adversity are thought to include variability in coping skills [63] and pre-existing temperament [64]. The present study raises the possibility that a third contributing factor is midbrain dopamine cell reactivity.
Midbrain D2 receptors are primarily autoreceptors [65]. In humans, somatodendritic autoreceptors are present on dopamine cells that project to the striatum but not cells that project to the cerebral cortex [66,67,68,69]. Despite this, there is evidence from studies in laboratory animals that altered mesostriatal dopamine transmission affects functioning within the larger mesocorticolimbic circuitry. In these studies, poorly regulated dopamine cell reactivity can lead to cognitive deficits [32, 70, 71], sub-optimal choices [32, 33], impulsivity [29, 72], and susceptibility to depressive phenotypes [23]. Adverse outcomes might be particularly profound when elevated mesostriatal dopamine transmission is combined with reduced or asynchronous cortical dopamine function [23, 32, 72]. Indeed, across diagnostic categories, there is evidence of poorly regulated dopamine transmission [20, 21, 73,74,75,76,77] with DRD2 variants constituting a transdiagnostic risk gene [78].
The study’s fourth finding was related to sex. As expected, males were more likely than females to be diagnosed with attention deficit and substance use disorders and less likely to meet criteria for mood and anxiety disorders. This noted, both in our study and in the general population, the lifetime prevalence of psychiatric illness overall is similar for men and women and adding sex as a predictor did not improve the model’s strength. These observations suggest that the three investigated factors influence susceptibility to psychiatric disorders in both males and females irrespective of how the susceptibility comes to be expressed.
The breadth of diagnoses predicted by the model raises the possibility that it contributes to the hypothesized p factor that has been proposed to affect risk for all common psychiatric disorders [4,5,6, 79]. The specific processes composing p have been little studied, but some work suggests that they include poor impulse control over emotions and difficulties in reality testing [4]. The findings reported here might identify sources of these alterations.
Strengths and limitations
The present research benefitted from prospectively collected behavioral traits and clinical outcomes. It is also, to our knowledge, the first PET study in a longitudinally followed birth cohort. These strengths noted, the results should be interpreted in light of the following considerations. First, all three factors contributed to the model’s strength after controlling for the influence of the two other factors indicating that they are not interchangeable proxies of each other. This noted, only EXT was statistically significant as a standalone variable. This observation suggests that CTQ scores and midbrain D2 receptor availabilities contribute through interactions with the other variables. Although the present sample is reasonably large for a PET study, it is not large enough to interrogate further these potential statistical interactions; we can therefore primarily conclude that high statistical power is provided by the model overall. Second, the three-factor model also predicted the presence of EXT disorders uniquely. In this analysis, the smaller number of affected cases decreased further the ability to disentangle contributions of each individual factor, but the effect appeared to be driven by EXT scores and midbrain [18F]fallypride BPND values. Additional work will be needed to better capture the model’s strength for EXT vs. internalizing disorders, but longitudinal research indicates that a history of disorders from one cluster predicts higher risk for both “homotypic” and “heterotypic” conditions [59]. Consistent with the proposition that the studied risk factors are not specific to a single cluster, the three-factor model was not significantly better at capturing EXT disorders than all occurring DSM-5 disorders. Third, much of the data was collected prospectively, but definitive statements about causality are not possible. Instead, the study identifies a model that predicts clinical outcomes. Fourth, the sample was relatively homogenous, composed primarily of francophone youth of European descent. Future studies should test whether the model generalizes to other populations. Fifth, the index interviews were conducted face-to-face, but the follow-up interviews were by telephone, potentially diminishing the quality of information from these later assessments. Sixth, participant attrition may not have been random. However, these sub-samples did not differ on key variables, including age, EXT score, or proportion with a lifetime history of DSM-5 disorders (Supplementary Table 5). Finally, the project’s PET tracer, [18F]fallypride, binds to both D2 and D3 receptors [80]. Midbrain D2 receptors are primarily located on mesostriatal dopamine neurons where they act to inhibit mesostriatal dopamine release [17, 19, 27]. The majority of midbrain D3 receptors, in comparison, are on terminal regions of descending striatal GABAergic neurons and their activation inhibits cortical dopamine release [18]. Since [18F]fallypride has preferential affinity for D2 [80], our PET data are likely dominated by effects relevant for mesostriatal dopamine transmission. This noted, to the extent that the effect could also include changes to D3 receptors, poorly synchronized striatal and cortical dopamine transmission erodes a wide range of cognito-affective processes [23, 32].
Conclusions
This first study of the three-factor model supports proposals that many common forms of early onset psychopathology emerge from the varying expression of shared intersecting dimensions [3,4,5]. Obtaining a better understanding of the identified features could provide targets for early interventions with implications for both nosology and etiology.
References
Krueger RF. The structure of common mental disorders. Arch Gen Psychiatry. 1999;56:921.
Kapur S, Phillips AG, Insel TR. Why has it taken so long for biological psychiatry to develop clinical tests and what to do about it? Mol Psychiatry. 2012;17:1174–9.
Lee SH, Ripke S, Neale BM, Faraone SV, Purcell SM, Perlis RH, et al. Genetic relationship between five psychiatric disorders estimated from genome-wide SNPs. Nat Genet. 2013;45:984–94.
Caspi A, Moffitt TE. All for one and one for all: mental disorders in one dimension. Am J Psychiatry. 2018;175:831–44.
Castellanos-Ryan N, Brière FN, O’Leary-Barrett M, Banaschewski T, Bokde A, Bromberg U, et al. The structure of psychopathology in adolescence and its common personality and cognitive correlates. J Abnorm Psychol. 2016;125:1039–52.
The Brainstorm Consortium, Anttila V, Bulik-Sullivan B, Finucane HK, Walters RK, Bras J, et al. Analysis of shared heritability in common disorders of the brain. Science. 2018;360:eaap8757. https://doi.org/10.1126/science.aap8757.
Smoller JW, Andreassen OA, Edenberg HJ, Faraone SV, Glatt SJ, Kendler KS. Psychiatric genetics and the structure of psychopathology. Mol Psychiatry. 2019;24:409–20.
Zhou H, Sealock JM, Sanchez-Roige S, Clarke T-K, Levey DF, Cheng Z, et al. Genome-wide meta-analysis of problematic alcohol use in 435,563 individuals yields insights into biology and relationships with other traits. Nat Neurosci. 2020;23:809–18.
Krueger RF, Kotov R, Watson D, Forbes MK, Eaton NR, Ruggero CJ, et al. Progress in achieving quantitative classification of psychopathology. World Psychiatry. 2018;17:282–93.
Krueger RF, Tackett JL. Personality and Psychopathology: Working Toward the Bigger Picture. J Personal Disord. 2003;17:109–28.
Kendler KS, Ohlsson H, Edwards AC, Sundquist J, Sundquist K. A developmental model for alcohol use disorders in Swedish men. Psychol Med. 2016;46:2759–70.
Foster KT, Hicks BM, Zucker RA. Positive and negative effects of internalizing on alcohol use problems from childhood to young adulthood: the mediating and suppressing role of externalizing. J Abnorm Psychol. 2018;127:394–403.
Byrd AL, Loeber R, Pardini DA. Antisocial behavior, psychopathic features and abnormalities in reward and punishment processing in youth. Clin Child Fam Psychol Rev. 2014;17:125–56.
Bjork JM, Pardini DA. Who are those ‘risk-taking adolescents’? Individual differences in developmental neuroimaging research. Dev Cogn Neurosci. 2015;11:56–64.
Miczek KA, Meyer-Lindenberg A, editors. Neuroscience of Aggression. 1st ed. 2014. Berlin, Heidelberg: Springer Berlin Heidelberg: Imprint: Springer; 2014.
Dalley JW, Robbins TW. Fractionating impulsivity: neuropsychiatric implications. Nat Rev Neurosci. 2017;18:158–71.
Meador-Woodruff JH, Damask SP, Watson SJ. Differential expression of autoreceptors in the ascending dopamine systems of the human brain. Proc Natl Acad Sci. 1994;91:8297–301.
Sokoloff P, Le Foll B. The dopamine D3 receptor, a quarter century later. Eur J Neurosci. 2017;45:2–19.
Milella MS, Fotros A, Gravel P, Casey KF, Larcher K, Verhaeghe JAJ, et al. Cocaine cue–induced dopamine release in the human prefrontal cortex. J Psychiatry Neurosci. 2016;41:322–30.
Buckholtz JW, Treadway MT, Cowan RL, Woodward ND, Li R, Ansari MS, et al. Dopaminergic Network Differences in Human Impulsivity. Science. 2010;329:532–532.
Jaworska N, Cox SML, Tippler M, Castellanos-Ryan N, Benkelfat C, Parent S, et al. Extra-striatal D2/3 receptor availability in youth at risk for addiction. Neuropsychopharmacology. 2020;45:1498–505.
Berridge KC, Robinson TE. Parsing reward. Trends Neurosci. 2003;26:507–13.
Chaudhury D, Walsh JJ, Friedman AK, Juarez B, Ku SM, Koo JW, et al. Rapid regulation of depression-related behaviours by control of midbrain dopamine neurons. Nature. 2013;493:532–6.
da Silva JA, Tecuapetla F, Paixão V, Costa RM. Dopamine neuron activity before action initiation gates and invigorates future movements. Nature. 2018;554:244–8.
Mohebi A, Pettibone JR, Hamid AA, Wong J-MT, Vinson LT, Patriarchi T, et al. Dissociable dopamine dynamics for learning and motivation. Nature. 2019;570:65–70.
Heymann G, Jo YS, Reichard KL, McFarland N, Chavkin C, Palmiter RD, et al. Synergy of distinct dopamine projection populations in behavioral reinforcement. Neuron. 2020;105:909–20.
Bello EP, Mateo Y, Gelman DM, Noaín D, Shin JH, Low MJ, et al. Cocaine supersensitivity and enhanced motivation for reward in mice lacking dopamine D2 autoreceptors. Nat Neurosci. 2011;14:1033–8.
Marinelli M, White FJ. Enhanced vulnerability to cocaine self-administration is associated with elevated impulse activity of midbrain dopamine neurons. J Neurosci. 2000;20:8876–85.
Bellés L, Dimiziani A, Tsartsalis S, Millet P, Herrmann FR, Ginovart N. Dopamine D2/3 receptor availabilities and evoked dopamine release in striatum differentially predict impulsivity and novelty preference in roman high- and low-avoidance rats. Int J Neuropsychopharmacol. 2021;24:239–51.
Tournier BB, Steimer T, Millet P, Moulin-Sallanon M, Vallet P, Ibañez V, et al. Innately low D2 receptor availability is associated with high novelty-seeking and enhanced behavioural sensitization to amphetamine. Int J Neuropsychopharmacol. 2013;16:1819–34.
D’Amour-Horvat V, Leyton M. Impulsive actions and choices in laboratory animals and humans: effects of high vs. low dopamine states produced by systemic treatments given to neurologically intact subjects. Front Behav Neurosci. 2014;8:432. https://doi.org/10.3389/fnbeh.2014.00432.
Duvarci S, Simpson EH, Schneider G, Kandel ER, Roeper J, Sigurdsson T. Impaired recruitment of dopamine neurons during working memory in mice with striatal D2 receptor overexpression. Nat Commun. 2018;9:2822.
Olivetti PR, Balsam PD, Simpson EH, Kellendonk C. Emerging roles of striatal dopamine D2 receptors in motivated behaviour: Implications for psychiatric disorders. Basic Clin Pharm Toxicol. 2020;126:47–55.
Holroyd KB, Adrover MF, Fuino RL, Bock R, Kaplan AR, Gremel CM, et al. Loss of feedback inhibition via D2 autoreceptors enhances acquisition of cocaine taking and reactivity to drug-paired cues. Neuropsychopharmacology. 2015;40:1495–509.
Matuszewich L, McFadden LM, Friedman RD, Frye CA. Neurochemical and behavioral effects of chronic unpredictable stress. Behav Pharm. 2014;25:557–66.
Booij L, Welfeld K, Leyton M, Dagher A, Boileau I, Sibon I, et al. Dopamine cross-sensitization between psychostimulant drugs and stress in healthy male volunteers. Transl Psychiatry. 2016;6:e740–e740.
Leyton M, Stewart J. Preexposure to foot-shock sensitizes the locomotor response to subsequent systemic morphine and intra-nucleus accumbens amphetamine. Pharm Biochem Behav. 1990;37:303–10.
Keyes KM, Eaton NR, Krueger RF, McLaughlin KA, Wall MM, Grant BF, et al. Childhood maltreatment and the structure of common psychiatric disorders. Br J Psychiatry. 2012;200:107–15.
Jaekel J, Sorg C, Breeman L, Baumann N, Bilgin A, Bäuml JG, et al. Early regulatory problems and parenting: life-long risk, vulnerability or susceptibility for attention, internalizing and externalizing outcomes? Eur Child Adolesc Psychiatry. 2020. 5 September 2020. https://doi.org/10.1007/s00787-020-01632-2.
Cox SML, Castellanos-Ryan N, Parent S, Benkelfat C, Vitaro F, Pihl RO, et al. Externalizing risk pathways for adolescent substance use and its developmental onset: A Canadian birth cohort study. Can J Psychiatry. 2021:070674372098242.
Orri M, Boivin M, Chen C, Ahun MN, Geoffroy M-C, Ouellet-Morin I, et al. Cohort Profile: Quebec Longitudinal Study of Child Development (QLSCD). Soc Psychiatry Psychiatr Epidemiol. 2020. 13 November 2020. https://doi.org/10.1007/s00127-020-01972-z.
Boivin M, Brendgen M, Dionne G, Dubois L, Pérusse D, Robaey P, et al. The Quebec newborn twin study into adolescence: 15 years later. Twin Res Hum Genet. 2013;16:64–69.
Tremblay RE, Vitaro F, Gagnon C, Piché C, Royer N. A prosocial scale for the preschool behaviour questionnaire: concurrent and predictive correlates. Int J Behav Dev. 1992;15:227–45.
Behar L, Stringfield S. A behavior rating scale for the preschool child. Dev Psychol. 1974;10:601–10.
Glasofer DR, Brown AJ, Riegel M Structured Clinical Interview for DSM-IV (SCID). In: Wade T, editor. Encycl. Feed. Eat. Disord., Singapore: Springer Singapore; 2015. p. 1–4.
Bernstein DP, Fink L Childhood trauma questionnaire: a retrospective self-report: manual. San Antonio, Tex.: Harcourt Brace & Co.; 1998.
Paquette D, Laporte L, Bigras M, Zoccolillo M. [Validation of the French version of the CTQ and prevalence of the history of maltreatment]. Sante Ment Que. 2004;29:201–20.
Initial reliability and validity of a new retrospective measure of child abuse and neglect. Am J Psychiatry. 1994;151:1132-6.
Goltermann J, Redlich R, Grotegerd D, Dohm K, Leehr EJ, Böhnlein J, et al. Childhood maltreatment and cognitive functioning: the role of depression, parental education, and polygenic predisposition. Neuropsychopharmacology. 2020. 14 August 2020. https://doi.org/10.1038/s41386-020-00794-6.
McLaughlin KA, Sheridan MA. Beyond Cumulative Risk: A Dimensional Approach to Childhood Adversity. Curr Dir Psychol Sci. 2016;25:239–45.
McLaughlin KA, Colich NL, Rodman AM, Weissman DG. Mechanisms linking childhood trauma exposure and psychopathology: a transdiagnostic model of risk and resilience. BMC Med. 2020;18:96.
Murty VP, Shermohammed M, Smith DV, Carter RM, Huettel SA, Adcock RA. Resting state networks distinguish human ventral tegmental area from substantia nigra. NeuroImage. 2014;100:580–9.
Lammertsma AA, Hume SP. Simplified reference tissue model for PET receptor studies. NeuroImage. 1996;4:153–8.
Box GEP, Tidwell PW. Transformation of the independent variables. Technometrics. 1962;4:531–50.
Tabachnick BG, Fidell LS Using multivariate statistics. 6th ed. Boston: Pearson Education; 2013.
Narkhede S Understanding AUC - ROC Curve. Medium. 2021. https://towardsdatascience.com/understanding-auc-roc-curve-68b2303cc9c5. Accessed 21 June 2021.
Hosmer DW, Lemeshow S, Sturdivant RX Applied logistic regression. Third edition. Hoboken, New Jersey: Wiley; 2013.
Zhou X, Obuchowski NA, McClish DK, editors. Statistical methods in diagnostic medicine. New York: Wiley-Interscience; 2002.
Lahey BB, Zald DH, Hakes JK, Krueger RF, Rathouz PJ. Patterns of heterotypic continuity associated with the cross-sectional correlational structure of prevalent mental disorders in adults. JAMA Psychiatry. 2014;71:989.
Copeland WE, Shanahan L, Hinesley J, Chan RF, Aberg KA, Fairbank JA, et al. Association of childhood trauma exposure with adult psychiatric disorders and functional outcomes. JAMA Netw Open. 2018;1:e184493.
Rutter M. Stress, coping and development: some issues and some questions. J Child Psychol Psychiatry. 1981;22:323–56.
Amstadter AB, Maes HH, Sheerin CM, Myers JM, Kendler KS. The relationship between genetic and environmental influences on resilience and on common internalizing and externalizing psychiatric disorders. Soc Psychiatry Psychiatr Epidemiol. 2016;51:669–78.
Asselmann E, Wittchen H-U, Lieb R, Beesdo-Baum K. A 10-year prospective-longitudinal study of daily hassles and incident psychopathology among adolescents and young adults: interactions with gender, perceived coping efficacy, and negative life events. Soc Psychiatry Psychiatr Epidemiol. 2017;52:1353–62.
Rioux C, Castellanos-Ryan N, Parent S, Vitaro F, Tremblay RE, Séguin JR. Differential susceptibility to environmental influences: Interactions between child temperament and parenting in adolescent alcohol use. Dev Psychopathol. 2016;28:265–75.
Chen R, Ferris MJ, Wang S. Dopamine D2 autoreceptor interactome: Targeting the receptor complex as a strategy for treatment of substance use disorder. Pharm Ther. 2020;213:107583.
Holloway ZR, Freels TG, Comstock JF, Nolen HG, Sable HJ, Lester DB. Comparing phasic dopamine dynamics in the striatum, nucleus accumbens, amygdala, and medial prefrontal cortex. Synapse. 2019;73:e22074.
Galloway MP, Wolf ME, Roth RH. Regulation of dopamine synthesis in the medial prefrontal cortex is mediated by release modulating autoreceptors: studies in vivo. J Pharm Exp Ther. 1986;236:689–98.
Wolf ME, Galloway MP, Roth RH. Regulation of dopamine synthesis in the medial prefrontal cortex: studies in brain slices. J Pharm Exp Ther. 1986;236:699–707.
Hoffmann IS, Talmaciu RK, Ferro CP, Cubeddu LX. Sustained high release at rapid stimulation rates and reduced functional autoreceptors characterize prefrontal cortex dopamine terminals. J Pharm Exp Ther. 1988;245:761–72.
Arnsten AFT, Vijayraghavan S, Wang M, Gamo NJ, Paspalas CD Dopamine’s Influence on Prefrontal Cortical Cognition: Actions and Circuits in Behaving Primates. In: Iversen L, Iversen S, Dunnett S, Bjorklund A, editors. Dopamine Handb., Oxford University Press; 2009. p. 230-48.
Vijayraghavan S, Major AJ, Everling S. Neuromodulation of prefrontal cortex in non-human primates by dopaminergic receptors during rule-guided flexible behavior and cognitive control. Front Neural Circuits. 2017;11:91.
Leyton M, Vezina P. Dopamine ups and downs in vulnerability to addictions: a neurodevelopmental model. Trends Pharm Sci. 2014;35:268–76.
Buckholtz JW, Treadway MT, Cowan RL, Woodward ND, Benning SD, Li R, et al. Mesolimbic dopamine reward system hypersensitivity in individuals with psychopathic traits. Nat Neurosci. 2010;13:419–21.
Cherkasova MV, Faridi N, Casey KF, O’Driscoll GA, Hechtman L, Joober R, et al. Amphetamine-induced dopamine release and neurocognitive function in treatment-naive adults with ADHD. Neuropsychopharmacology 2014;39:1498–507.
Vosberg DE, Leyton M, Flores C. The Netrin-1/DCC guidance system: dopamine pathway maturation and psychiatric disorders emerging in adolescence. Mol Psychiatry. 2020;25:297–307.
Wang G-J, Wiers CE, Shumay E, Tomasi D, Yuan K, Wong CT, et al. Expectation effects on brain dopamine responses to methylphenidate in cocaine use disorder. Transl Psychiatry. 2019;9:93.
Sequeira SL, Silk JS, Ladouceur CD, Hanson JL, Ryan ND, Morgan JK, et al. Association of Neural Reward Circuitry Function With Response to Psychotherapy in Youths With Anxiety Disorders. Am J Psychiatry. 2021:appiajp202020010094.
Lee PH, Anttila V, Won H, Feng Y-CA, Rosenthal J, Zhu Z, et al. Genomic relationships, novel loci, and pleiotropic mechanisms across eight psychiatric disorders. Cell. 2019;179:1469–1482.e11.
Lahey BB, Applegate B, Hakes JK, Zald DH, Hariri AR, Rathouz PJ. Is there a general factor of prevalent psychopathology during adulthood? J Abnorm Psychol. 2012;121:971–7.
Mukherjee J, Yang ZY, Brown T, Lew R, Wernick M, Ouyang X, et al. Preliminary assessment of extrastriatal dopamine D-2 receptor binding in the rodent and nonhuman primate brains using the high affinity radioligand, 18F-fallypride. Nucl Med Biol. 1999;26:519–27.
Acknowledgements
We thank David Zald for feedback on an earlier version of this manuscript.
Funding
The project has been made possible with the financial support of Health Canada, through the Canada Brain Research Fund, an innovative partnership between the Government of Canada (through Health Canada) and Brain Canada, and the Montreal Neurological Institute, with further support from the Canadian Institutes for Health Research (CIHR) MOP-133537 (Leyton, Séguin, Boivin), MOP-44072 (Séguin) and MOP-97910 (Séguin, Parent); from the Fonds de Recherche du Québec—Santé (FRQS) numbers 981055, 991027 (Séguin) and 35282 (Castellanos-Ryan); Fonds de Recherche du Québec - Société et Culture numbers 2002-RS-79238, 2009-RG-124779 (Séguin, Boivin); and from the Social Sciences and Humanities Research Council of Canada numbers 410-99-1048 and 839-2000-1008 (Séguin, Boivin). Salary awards were provided by CIHR to Jaworska and Castellanos-Ryan and by FRQS to Tippler. The QLSCD cohort born 1997–1998 is led by the Institut de la Statistique du Québec in collaboration with several departments and agencies of the Government of Quebec and many collaborating researchers including the authors of this article. Project completion with QLSCD respondents was authorized by the QLSCD Steering Committee. The views expressed herein do necessarily represent the views of the Minister of Health or the Government of Canada.
Author information
Authors and Affiliations
Contributions
ML conceived of the study and designed it with JS and SMLC. MT, SMLC and NJ collected the data. MI, SMLC, NJ and NCR analyzed the data. MI, SMLC, NJ and ML interpreted the data and wrote the manuscript with input from the other authors (MT, SP, AD, FV, MRB, MB, ROP, SMC, RET, JRS).
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
About this article
Cite this article
Iqbal, M., Cox, S.M.L., Jaworska, N. et al. A three-factor model of common early onset psychiatric disorders: temperament, adversity, and dopamine. Neuropsychopharmacol. 47, 752–758 (2022). https://doi.org/10.1038/s41386-021-01187-z
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1038/s41386-021-01187-z