Can polygenic risk scores help identify pediatric bipolar spectrum and related disorders?: A systematic review

Abstract

The genetic basis of mood disorders can, theoretically, provide diagnostic information in scenarios of clinical uncertainty. Therefore, we examined the available body of knowledge on the association between polygenic risk scores for bipolar disorder (BP-PRSs) and pediatric bipolar spectrum and related disorders. We performed a literature search through PubMed in March of 2020. The following variables were extracted from relevant studies: population age, study sample size, source of polygenic risk scores, source of data, the primary goal of the study, the assessments used during the course of the study, and the main findings/outcomes of each study. BP-PRSs were associated with deficits in executive functioning and the diagnosis of attention deficit/hyperactivity disorder (ADHD). Three studies included in our analysis directly compared major depressive disorder (MDD)-PRSs to BP-PRSs in youth. Results showed that MDD-PRSs, and not BP-PRSs, were associated with ADHD symptoms, internalizing problems, and social problems. ADHD-PRSs were associated with conduct problems, depressive symptomatology, and externalizing disorders symptoms. Findings revealed that ADHD-PRSs were more clearly associated with emotional reactivity, emotional dysregulation, and irritability—frequent correlates of pediatric BP disorder. These findings suggest that ADHD-PRSs may have an important contribution to the development of mood related problems in youth.

1. Introduction

Despite a few decades of research, the diagnosis of pediatric bipolar disorder (BP) continues to be viewed with skepticism. Although the reasons for this state of affairs are not entirely clear, they could stem from the complexity of the clinical picture, the frequent presence of aggression and dyscontrol, and the high rate of comorbidity with ADHD and disruptive behavior disorders including oppositional defiant and conduct disorder (Doyle and Faraone, 2002; Faraone et al., 2012; Kowatch, 2016; Spencer et al., 2001; Uchida et al., 2014; Yule et al., 2019).

Among the most consequential comorbidities of pediatric BP is the comorbidity with conduct disorder (CD) that has been well documented to be high and bidirectional: youth with BP frequently meet criteria for CD and vice versa (Masi et al., 2008; Wozniak et al., 2019). Likewise, high levels of severe irritability, aggression, and emotional dysregulation are the main abnormalities of the mood disturbance in pediatric BP (Uchida et al., 2014). We previously documented that high levels of emotional dysregulation captured through aggregate scores on a unique profile of the CBCL consisting of the Aggressive Behavior, Attention Problems and Anxious/Depressed scales are highly correlated with a structured diagnostic interview of pediatric BP (Uchida et al., 2014; Yule et al., 2019). Consistent with these clinical findings, neuroimaging studies using spectroscopy (Wozniak et al., 2012) and DTI (Hung et al., 2020) methodology found significant correlations between the severity of emotional dysregulation and levels of glutamate in the anterior cingulate as well as deficits in white matter tracts in regions of the brain connecting the limbic system with the frontal cortex. Furthermore, as we recently documented in a systematic review of the extant literature, even subsyndromal manifestations of BP are highly morbid (Vaudreuil et al., 2019; Wozniak et al., 2011).

Unfortunately, the complexity of the clinical picture of pediatric BP can lead to neglecting the diagnosis of BP and a consequential misdiagnosis in some patients, such as conduct disorder or ADHD, resulting in inappropriate treatments. Thus, the identification of external validators for this complex disorder could aid in the assessment of pediatric BP patients.

The strong genetic basis of BP can, theoretically help provide external validation of pediatric BP. Because the genetic basis of complex disorders such as BP is polygenic (Smoller et al., 2018; Stahl et al., 2019; Wray et al., 2007), clinically relevant information could emerge from genetic markers in aggregate in the form of polygenic risk scores (PRSs) (Purcell et al., 2009; Wray et al., 2007). PRSs for an individual are calculated as weighted counts of thousands of risk variants, where the risk variants and their weights have been identified in genome-wide association studies (Wray et al., 2020). To the best of our knowledge the potential utility of PRS has not been examined in pediatric BP research.

In this study we aimed to systematically review the available body of knowledge about the association between polygenic risk scores for pediatric BP and pediatric bipolar spectrum (BP-PRS). Because of the extremely high comorbidity with depressive disorders and ADHD documented in pediatric BP (Faraone et al., 2012; Kowatch, 2016), we also examined the contribution of PRSs for ADHD and depression to this risk.

2. Methods

2.1. Database Search

We performed a literature search through PubMed in March of 2020 using the following search algorithms: (1) [Polygenic Risk or Polygenic] AND [Bipolar Disorder or Bipolar or Bipolar Depression], or (2) [Polygenic Risk or Polygenic] AND [Depression or MDD or MD], or (3) [Polygenic Risk or Polygenic] AND [Attention or ADHD]. Because of their frequent presence in youth with BP, we also included in our search criteria the presence of conduct disorder, emotional dysregulation, and aggression. In addition to the PubMed search, the authors screened, reviewed, and assessed the reference lists of the retrieved papers. Our systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

2.2. Selection Criteria

All articles were screened for inclusion/exclusion criteria by authors JB and AG. Articles were included if they met the following criteria: (1) original research in a peer-reviewed journal, (2) research focus on pediatric populations and/or involved a pediatric sample, (3) primary research focus on the application of polygenic risk scores to BP, depression, or associated psychopathology (emotional dysregulation, irritability, conduct symptoms), and (4) the polygenic risk score calculated from a general population sample or specifically from samples of probands with BP, MDD, or associated psychopathologies. Articles were excluded if they met the following exclusion criteria: (1) focused on other psychiatric, medical, or neurological disorders that were not BP, depression, or associated psychopathologies, (2) reported no clinical findings or outcomes (i.e., neuroimaging studies), (3) published in a language other than English, and (4) performed a literature review or meta-analysis.

2.3. Data Extraction

The following variables were extracted from the relevant studies: population age, study sample size, source of polygenic risk scores, source of clinical and genetic data, the primary goal of the study, the assessments used during the course of the study, and the main findings/outcomes of each study.

2.4. Qualitative Analysis

Our analysis focused on associations between polygenic risk for BP, MDD, and ADHD with bipolar spectrum disorder and related disorders.

3. Results

Fig. 1 details the screening and evaluation process. After the removal of duplicates, 755 records were identified. Of these, 160 were considered relevant to our research and were carefully examined. One hundred forty-six of these 160 articles were excluded because they did not meet our inclusion and exclusion criteria. The primary reasons for exclusion were: (1) the research focus was on adults, and (2) polygenic risk was calculated from populations of different psychiatric disorders that were not specifically ADHD, BP, or associated psychopathology (emotional dysregulation, irritability, conduct symptoms). Ultimately, 13 studies met our inclusion/exclusion criteria and were included in our analysis.

Fig. 1.

(PRISMA Diagram and Checklist).

From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(7): e1000097. doi:10.1371/journal.pmed1000097

For more information, visit www.prisma-statement.org.

3.1. Bipolar Disorder Polygenic Risk Scores (BP-PRSs)

As shown in Table 1, BP-PRSs were significantly associated with the diagnosis of ADHD and associated executive function deficits. ADHD children were more likely to have higher BP-PRSs than controls (Schimmelmann et al., 2013) and higher BP-PRSs were correlated with increased hypomania scores on the Hypomania Checklist (HCL) and poorer IQ (Mistry et al., 2019a, b). Quantification of these risks revealed that BP-PRSs increased the risk for ADHD (OR: 1.13–1.51) (Mistry et al., 2019a; Schimmelmann et al., 2013), hypomania (OR=1.33), and executive functioning deficits (ß= −0.03) (Mistry et al., 2019b) (Table 1) (Supplementary Table 1).

Table 1.

Polygenic Risk Review Tabulation Table.

Author (Year)	Study Population	Study Goals	Methods	Main Findings and Comments
Bipolar Disorder Polygenic Risk Scores (BP-PRSs)
Mistry et al. (2019) Investigating associations between genetic risk for bipolar disorder and cognitive functioning in childhood	Population Sample N = 5,613 – 5,936 (ages 7–8) Clinical and Genetic Data collected from ALSPAC BP-PRSs calculated from PGC summary data (adult population)	Examine the association between BP-PRSs and cognitive variables in pediatric BP disorder	Assessments Weschler Intelligence Scale - III Freedom from Distractibility Index Test of Everyday Attention for Children Diagnostic Analysis of Nonverbal Accuracy	High BP-PRSs were associated with impaired executive functioning, lower performance IQ and poorer processing speed
Mistry et al. (2019) Genetic risk for bipolar disorder and psychopathology from childhood to early adulthood.	Population Sample N = 3,448 (all assessments completed between ages 7–11 except the HCL-32, which was collected when the cohort was 22–23) Clinical and Genetic Data collected from ALSPAC BP-PRSs calculated from the PGC GWAS for Bipolar Disorder	Examine the impact of BD-PRSs on the development of childhood psychopathology in early adulthood.	Assessments Hypomania Checklist – 32 (HCL-32) Strengths and Difficulties Questionnaire (SDQ) Development and Wellbeing Assessment (DAWBA) Childhood Interview for DSM-IV Borderline Personality Disorder (CI-BPD)	BP-PRSs strongly associated with the diagnosis of ADHD
Schimmelmann et al (2013) Bipolar disorder risk alleles in children with ADHD	ADHD Sample N = 495 (ages 6–18) Healthy Controls N = 4,433 BP-PRSs calculated from Hinney, et al (2011). Genetic and Clinical data collected from the same study.	Interaction between BP-PRSs and childhood ADHD	Assessments DSM-IV Diagnosis made according to K-SADS, Kinder-DIPS, or PACS Other psychiatric diagnoses made from K-SADS, Kinder-DIPS, or PACS	Children with ADHD have a higher probability of being a BP disorder risk allele carrier than controls
Summary of Findings: BP-PRSs are strongly associated with the diagnosis of ADHD, impaired executive functioning, and lower IQ in childhood populations.
Major Depressive Disorder Polygenic Risk Scores (MDD-PRSs)
Halldorsdottir et al (2019) Polygenic risk: predicting depression outcomes in clinical and epidemiological cohorts of youths	MDD Sample N = 279 (ages 7–17) Sub-threshold MDD Children N = 694 (ages 12–17) Healthy Controls N = 187 (ages 7–17) PRSs calculated from PPGC GWAS study on MDD	Examine whether MDD-PRSs in adults generalize to childhood depression and depressive symptoms.	Assessments Diagnostic Interview for Mental Disorders in Childhood and Adolescence Depression Inventory for Children and Adolescents Beck Depression Inventory-II Life Event Survey Munich Life Event List Children’s Depression Inventory Childhood Trauma Questionnaire Child Behavior Checklist	MDD-PRSs predicted significant levels of childhood MDD and earlier age of onset of MDD Greater MDD-PRSs increased the risk of future development of depressive symptoms or a clinical diagnosis of MDD BP-PRSs did not predict depression severity or age at onset of depressive symptoms
Kwong, et al (2019) Genetic and environmental risk factors associated with trajectories of depression symptoms from adolescence to young adulthood	Population Sample N = 9,394 (ages 10–24) Genetic and Clinical data collected from the ALSPAC MDD-PRSs calculated from Okbay, et al (2016).	Predictive utility of MDD-PRSs in the childhood trajectory of MDD	Assessments Short Mood and Feelings Questionnaire	Higher MDD-PRSs were associated with a trajectory of early-adult onset depression and childhood persistent depression BP outcomes not examined
Nigg et al (2020) Evaluating chronic emotional dysregulation and irritability in relation to ADHD and depression genetic risk in children with ADHD	Population Sample N = 495 (ages 7–11) Clinical and Genetic Data collected from community recruitment. ADHD-PRSs and MDD-PRSs calculated from Demontis et al., 2019	Clarify whether emotional dysregulation and irritability are a result of increased ADHD or MDD-PRSs	Assessments Temperament in Middle Childhood Questionnaire (TMCQ) ADHD Rating Scale Conner’s Parent Rating Scale-Third Edition Strengths and Difficulties Questionnaire	Increased ADHD-PRSs but not MDD-PRSs were associated with irritability and emotional reactivity ADHD-PRSs but not MDD-PRSs predicted ADHD, emotionally dysregulation, and irritability BP-PRSs unrelated to irritability or symptoms of ADHD
Riglin et al (2017) Investigating the genetic underpinnings of early-life irritability	ADHD Sample N = 678 (ages 6–18) Population Sample N = 43,398 Clinical and Genetic Data collected from ALSPAC, NCDS, and SAGE ADHD-PRS and MDD-PRSs calculated from PGC	Relationship between ADHD and MDD-PRSs and irritability	Assessments DAWBA Rutter A Scale in NCDS	ADHD-PRSs but not MDD-PRSs were associated with irritability at multiple time points throughout childhood (age 7, 10, and 15) MDD-PRSs not associated with irritability
Akingbuwa et al (2020) Genetic Associations Between Childhood Psychopathology and Adult Depression and Associated Traits in 42,998 Individuals	Population Sample N = 42,998 Clinical and Genetic Data collected from ALSPAC, CATS, Generation R, NMCCS, NFB, NTR, and TED. BPD-PRSs, MDD-PRSs, Subective Well-Being-PRSs, Neuroticism PRSs, Insomnia PRSs, Education Attainment-PRSs, and BMI-PRSs calculated from combined GWAS data	Identify associations between childhood psychopathology and adult mood disorders.	Assessments Strength and Difficulties Questionnaire Autism-Tics, ADHD, and Other Comorbidities Inventory Screen for Child Anxiety Related Emotional Disorders Short Mood and Feelings Questionnaire Child Behavior checklist Rating Scale for Disruptive Behvaior Disorders Youth Self Report Conners’ Parent Rating Scale	MDD-PRSs but NOT BP-PRSs were associated with ADHD symptoms, internalizing problems, and social problems BP-PRSs not associated with childhood psychopathology
Summary of Findings: MDD-PRSs predict the severity and future trajectory of MDD (BP outcomes not examined), and ADHD-PRSs, but not MDD-PRSS, are associated with increased irritability and emotional dysregulation.
Attention/Deficit-Hyperactivity Disorder Polygenic Risk Scores (ADHD-PRSs) and Proxies for BP Disorder Illness
Hamshere et al. (2013) High loading of polygenic risk for ADHD in children with comorbid aggression	ADHD Sample N = 452 (ages 6–17) Controls N = 5,081 Clinical and Genetic Data collected from Cardiff Sample. Controls recruited from WTCCC2 ADHD-PRSs calculated from Neale, 2010 (adult population)	Examine whether ADHD-PRSs are more significant in a sample of children with ADHD and comorbid conduct disorder (CD), compared to those with ADHD only.	Assessments Child and Adolescent Psychiatric Assessment	High ADHD-PRSs were associated with increased number of aggressive CD symptoms
Brikell et al. (2018) The contribution of genetic risk variants for ADHD to a general factor of childhood psychopathology	Population Sample N = 13,457 (ages 9 or 12) Clinical and Genetic Data collected from CATSS ADHD-PRSs calculated from Demontis, 2017.	Association between ADHD- PRSs and general childhood psychopathology.	Assessments Autism-Tics, ADHD, and Other Comorbidities Inventory (ATAC) Short Mood and Feelings Questionnaire (SMFQ) Screen for Child Anxiety Related Emotional Disorders (SCARED)	High ADHD-PRSs were associated with increased risk for depressive symptoms, ADHD, and childhood neurodevelopmental and externalizing disorders There was a significant association between ADHD-PRS and general psychopathology
Vuijk, et al (2019) Translating discoveries in attention deficit hyperactivity disorder genomics to an outpatient child and adolescent psychiatric cohort	ADHD Sample N = 433 (ages 7–18) Clinical and Genetic Data collected from the Longitudinal Study of Genetic Influences on Cognition ADHD-PRS calculated from GWAS study on ADHD, Schizophrenia, and Autism Spectrum Disorder (adult population) Results replicated in 5,140 adults	Consider the effects of ADHD-PRSs on general childhood psychopathology and identify whether there is a unique ADHD clinical profile in those with high ADHD-PRS risk	Assessments DSM-IV-TR Axis 1 diagnosis Child Behavior Checklist Social Responsiveness Scale Weschler Children/Adult Intelligence Scale-Fourth Edition	High ADHD-PRSs were associated with aggression and more severe psychopathology Analysis in adults replicated the findings in youth.
Riglin et al (2016) Association of genetic risk variants with attention deficit hyperactivity disorder trajectories in the general population	Population Sample N = 9,757 (ages 4–17) Clinical and Genetic Data collected from ALSPAC ADHD-PRSs calculated from PGC	Examine the effect of ADHD-PRSs on the persistence and trajectory of ADHD symptoms into adulthood.	Assessments Strengths and Difficulties Questionnaire Wechsler Intelligence Scale for Children Social and Communication Disorders Checklist Children’s Communication Checklist	Persistent ADHD was associated with lower IQ, social communication problems, language impairment, and conduct problems
Demontis et al. (2020) Identification of risk variants and characterization of the polygenic architecture of disruptive behavior disorders in the context of ADHD	Population Sample N = 31,305 Clinical and Genetic Data collected from Danish iPSYCH cohort N=3,803 (ADHD, DBD, and/or CD) ADHD, DBD, and CD-PRSs calculated from Danish iPSYCH cohort	Determine the genetic basis of ADHD and comorbid Disruptive Behavior Disorders (DBDs)	All cases included in the analysis had a diagnosis of ADHD, DBDs or hyperkinetic conduct disorder	Subjects with ADHD+DBD had significantly increased PRSs for aggressive behavior and ADHD Cognitive performance negatively associated with ADHD+DBD-PRSs compared to ADHD without DBDs.
Summary of Findings: ADHD-PRSs are associated with the development and severity of comorbid psychopathology. ADHD-PRSs are specifically related to increased conduct disorder symptoms, aggressive behavior, and depressive symptomatology.

3.2. Major Depressive Disorder Polygenic Risk Scores (MDD-PRSs)

Three studies included in our analysis directly compared MDD-PRSs to BP-PRSs in youth. In a population sample of more than 40,000 children aged 6–17, only MDD-PRSs and not BP-PRSs, were associated with ADHD symptoms, internalizing problems, and social problems (Akingbuwa et al., 2020).

Similarly, when comparing MDD-PRSs and BP-PRSs in youth with diagnosed depression, MDD-PRSs were predictive of depression severity and age at onset of depressive symptoms. However, BP-PRSs were not predictive of these outcomes (Halldorsdottir et al., 2019). Other MDD-PRS studies reported that higher MDD-PRSs were predictive of more severe MDD trajectories including an earlier onset of MDD and a childhood-persistent trajectory of MDD (Halldorsdottir et al., 2019; Kwong et al., 2019). However, these studies did not examine BP spectrum outcomes associated with these traits. A small population sample showed that ADHD-PRSs, but not MDD-PRSs or BP-PRSs, were associated with emotional reactivity, emotional dysregulation (Nigg et al., 2020), and irritability (Riglin et al., 2017).

Quantification of these risks showed that the MDD-PRSs were significantly associated with depressive symptomatology (ß = 0.557) (Halldorsdottir et al., 2019), ADHD (ß = 0.0495), social problems (ß = 0.0403), internalizing problems (ß = 0.016) (Akingbuwa et al., 2020), childhood-persistent MDD (OR = 1.47), and an early-adult-onset MDD (OR=1.29) (Kwong et al., 2019) (Table 1) (Supplementary Table 1).

3.3. Attention/Deficit-Hyperactivity Disorder Polygenic Risk Scores (ADHD-PRSs)

ADHD-PRSs were associated with conduct problems, depressive symptoms and externalizing disorders symptoms. In a sample of youth with ADHD, ADHD-PRSs were significantly higher in participants who had a comorbid diagnosis of ADHD and conduct disorder (CD) compared with ADHD alone. The relationship was primarily driven by aggressive symptoms, rather than covert conduct symptoms (Hamshere et al., 2013). Similarly, aggression was also associated with increased ADHD-PRSs in a separate sample of ADHD children aged 7–18 (Vuijk et al., 2019). The relationship between ADHD and CD was also seen in a population sample of children; ADHD-PRSs were strongly correlated to conduct problems, specifically in ADHD youth with a persistent trajectory of ADHD. This same study, however, found no relationship between ADHD-PRSs and BP (Riglin et al., 2016). A third study analyzed ADHD-PRSs, CD-PRSs, and disruptive behavior disorder (DBD)-PRSs and showed that subjects with comorbid ADHD and DBD showed higher genetic correlations with aggressive and anti-social behaviors (Demontis, 2021). Finally, one population study in youth looked at the association between ADHD-PRSs and depressive symptoms and found that higher ADHD-PRSs were correlated with increased depressive symptoms. While the ADHD-PRSs in this study were significantly associated with depressive symptoms, they were even more associated with externalizing symptoms, such as those captured by oppositional defiant disorder (ODD) and CD questionnaires (Brikell et al., 2018). Multiple studies showed that higher ADHD-PRSs were predictive of more psychiatric symptoms in subjects with ADHD (Brikell et al., 2018; Riglin et al., 2016; Vuijk et al., 2019). Brikell et al. (Brikell et al., 2018) reported that ADHD-PRSs were significantly associated with general psychopathology suggesting that the genetic variants associated with ADHD can also be associated with other forms of childhood psychopathology (Brikell et al., 2018).

Quantification of these risks showed that ADHD-PRSs significantly increased the risk for ADHD diagnosis (OR = 1.39) (Vuijk et al., 2019), multimorbidity (OR=1.16) (Riglin et al., 2016), CD (ß = 0.118) (Hamshere et al., 2013), hyperactivity/impulsivity (ß = 0.06), and general psychopathology (ß = 0.09–0.24) (Brikell et al., 2018; Vuijk et al., 2019) (Table 1) (Supplementary Table 1).

Studies that assessed the risk contribution of multiple PRSs showed that MDD-PRSs, and not BP-PRSs, were associated with ADHD (ß = 0.0495) (Akingbuwa et al., 2020) and MDD severity and age-of-onset (Halldorsdottir et al., 2019). Other studies reported that only ADHD-PRSs, and not BP-PRSs nor MDD-PRSs, increased the risk for ADHD (Nigg et al., 2020; Riglin et al., 2016) and irritability (Nigg et al., 2020; Riglin et al., 2017) (Supplementary Table 1).

4. Discussion

This literature review aimed to examine the available body knowledge on the association between PRSs of BP, MDD and ADHD with the development of pediatric bipolar spectrum and related disorders. Findings revealed that while higher BP-PRSs were correlated with increased hypomania scores in a single study (Mistry et al., 2019a ADHD-PRSs were more clearly associated with emotional reactivity, emotional dysregulation (Nigg et al., 2020) and irritability (Riglin et al., 2017), frequent correlates of pediatric BP spectrum disorder. Although MDD-PRSs were predictive of more severe pediatric MDD trajectories, their contribution to the development of pediatric BP spectrum and related disorders remains unknown since the available studies did not examine the risk for BP and BP spectrum disorder.

ADHD-PRSs were associated with conduct problems, depressive symptoms, and aggressive symptoms (Brikell et al., 2018; Hamshere et al., 2013; Riglin et al., 2016; Vuijk et al., 2019). These findings are consistent with a body of literature linking ADHD with CD, ODD, emotional dysregulation, MDD and BP (Abikoff and Klein, 1992; Biederman et al., 2009; Biederman et al., 2012; Faraone et al., 2012). Taken together, these findings suggest that the genetic variants associated with ADHD are associated with related forms of serious childhood psychopathology (Brikell et al., 2018) in general and mood disorders in particular.

The limited contribution of BP-PRSs to the development of pediatric mood disorders is surprising. These pediatric findings are inconsistent with findings reported by Liebers et al. (Liebers et al., 2020) in adults showing that PRSs discriminated modestly between BP and MDD cases. More work is needed to further investigate the contributions of BP-PRSs in pediatric mood disorders.

Also noteworthy are the findings limiting the contribution of MDD-PRSs to depressive symptoms. However, as noted previously, none of the available studies reviewed examined the risk for BP and BP spectrum disorders. Further examination as to whether MDD-PRSs are selectively associated with pediatric MDD and not BP could have important clinical and scientific implications considering the critical differential diagnosis between pediatric unipolar depression and BP forms of mood disorders (Geller, 1994; Salvatore et al., 2013; Weissman et al., 1999).

Despite their limited role in pediatric psychiatry, as discussed by Liebers et al., PRSs are increasingly studied in medicine in the predictions of common medical disorders such as heart disease (Khera et al., 2018), type-II diabetes (Udler et al., 2019), obesity (Khera et al., 2019), and cancer (Mavaddat et al., 2019; Seibert et al., 2018). In these studies, PRSs were found to be largely independent of clinically-based risk estimation tools, suggesting that they could have an independent and perhaps incremental role in prediction models that could inform clinical decision making (Lee et al., 2019; Lewis and Vassos, 2020). Also, because unlike clinical predictors, genetic predictors can be measured before the onset of symptom, PRSs may contribute to early identification and preventive efforts in disease prediction and management, a most important issue in pediatric psychiatry.

As recently discussed by Wray et al. (Wray et al., 2020), although PRSs on their own are unlikely to be able to establish or definitively predict future diagnoses of common complex conditions, they could play a role as part of multivariable predictive algorithms or risk calculators and could have utility to help triage patients at risk to appropriate screening programs and contribute to clinical decision-making (Wray et al., 2020) in conjunction with other data in order to provide the most accurate prediction of risk.

Our findings need to be seen in light of some methodological limitations. Because the literature on the subject is limited, we are unable to make strong inferences. We relied on expanded phenotypes that included disruptive behavior disorders, emotional dysregulation and irritability to assess the contribution of PRS to pediatric mood disorders. We also expanded the search for PRS beyond pediatric BP ones to include PRS for ADHD and depression as well considering the high comorbidity between pediatric BP with these disorders. While we believe that this approach is highly informative considering the diagnostic uncertainties associated with pediatric BP, whether a more narrow approach would have been preferrable remains an open question. Another limitation stems from the fact that there are many more studies of ADHD PRS than of pediatric MDD or pediatric BP which could have influenced the results. A final concern surrounding the interpretation of PRS findings is that the scores have largely been calculated from studies of patients of European ancestry (Martin et al., 2019), and may not generalize to other ancestral groups.

Despite these caveats, our review of the available literature suggests that ADHD-PRSs may have the largest contribution to the development of pediatric psychopathology in general and mood disorders in particular. More work is needed to further evaluate the contributions of PRSs to disease prevention and management in pediatric psychiatry.