Psychiatric Polygenic Risk Scores and Week-by-Week Symptomatic Status in Youth with Bipolar Disorder: An Exploratory Study

Abstract

Introduction:

Prior studies have demonstrated that, in both adults and youth, bipolar disorder (BD) is a polygenic illness. However, no studies have examined polygenic risk scores (PRSs) in relation to the longitudinal course of mood symptoms in youth with BD.

Methods:

This study included 246 youth of European ancestry with BD (7–20 years old at intake) from the Course and Outcome of Bipolar Youth study and Centre for Youth Bipolar Disorder. Mood symptom severity was assessed at intake and, for 168 participants, prospectively for a median of 8.7 years. PRSs for BD, schizophrenia (SCZ), major depressive disorder (MDD), and attention-deficit/hyperactivity disorder (ADHD) were constructed using genome-wide summary statistics from independent adult cohorts.

Results:

Higher BD-PRS was significantly associated with lower most severe lifetime depression score at intake (β = −0.14, p = 0.03). Higher SCZ-PRS and MDD-PRS were associated with significantly less time spent in euthymia (SCZ-PRS: β = −0.21, p = 0.02; MDD-PRS: β = −0.22, p = 0.01) and more time with any subsyndromal mood symptoms (i.e., any mania, mixed, or depression symptoms; SCZ-PRS: β = 0.15, p = 0.04; MDD-PRS: β = 0.17, p = 0.01) during follow-up. PRSs for BD and ADHD were not significantly associated with any longitudinal mood variable.

Conclusions:

This exploratory analysis was the first to examine psychiatric PRSs in relation to the prospective course of mood symptoms among youth with BD. Results from the current study can serve to guide future youth BD studies with larger sample sizes on this topic.

Introduction

Bipolar disorder (BD) is a severe recurrent mood disorder characterized by episodes of mania and/or hypomania and depression (Goldstein et al., 2017). BD affects 2%–3% of youth and is the fourth-leading cause of disability among youth (Gore et al., 2011; Van Meter et al., 2019). Compared to adult-onset BD, early-onset of BD is particularly associated with a worse course of illness (e.g., less euthymia, more rapid cycling, high rates of recurrences) (Leverich et al., 2007). Prior prospective naturalistic studies have shown that youth with BD spend more time symptomatic (∼60% of the follow-up time), particularly with subsyndromal symptoms of depression and mixed polarity, and experience more changes in symptomatic status compared to adults with BD (Birmaher et al., 2009, 2006; McKnight et al., 2017).

BD is also highly heritable and polygenic (Gordovez and McMahon, 2020). Heritability estimates of BD from twin and family studies range between 60% and 85% (Barnett and Smoller, 2009). Recently, polygenic risk score (PRS) for BD (BD-PRS), a weighted sum of single nucleotide polymorphisms (SNPs) identified from genome-wide associated studies (GWAS) in adults with BD (Mullins et al., 2021), has been shown to significantly differentiate individuals with BD, at high risk for BD, and controls in both adult and youth cohorts (Aminoff et al., 2015; Andlauer et al., 2021; Birmaher et al., 2022; Boies et al., 2018; Charney et al., 2017; Fullerton et al., 2015; Jiang et al., 2024; Power et al., 2015). It is noteworthy that heritability estimates of BD from GWAS (∼23%) are lower than the estimates from family, twin and adoption studies, and that BD-PRS only explains 8% of phenotypic variance in both adults and youth with BD (Birmaher et al., 2022; Jiang et al., 2024; Mullins et al., 2021; Owen and Williams, 2021). The observed gap in heritability suggests that additional genetic factors such as complex interactions among SNPs or rare genetic variants, and environmental factors also contribute to the heritability of BD, and should be taken into consideration to address the heritability gap and enhance the precision of genomic studies in BD and related symptoms.

Higher BD-PRS has been found to associate with greater burden of mania/hypomania in general adult population sample or adults with BD, SCZ, and MDD (Bipolar Disorder and Schizophrenia Working Group of the Psychiatric Genomics Consortium, 2018; Mistry et al., 2019; Richards et al., 2022; Ruderfer et al., 2014; Wiste et al., 2014). One prior study in adults with BD also reported a positive association between BD-PRS and manic episodes during the follow-up period (median 5 years) (Hasseris et al., 2023). However, no prior studies have examined the association between BD-PRS and longitudinal course of mood symptoms in youth with BD.

Combining data from the Brown University and University of Pittsburgh sites of the Course and Outcome of Bipolar Youth (COBY) study and the Centre for Youth Bipolar Disorder (CYBD) in Toronto (Birmaher et al., 2009; Jiang et al., 2024), with a median 8.7 years of follow-up of the weekly mood symptoms, this exploratory study aimed to investigate the association between BD-PRS and the longitudinal mood trajectories of youth with BD. We also examined PRSs for schizophrenia (SCZ), major depressive disorder (MDD), and attention-deficit/hyperactivity disorder (ADHD) due to the high rate of comorbidity as well as overlapping clinical symptoms and genetics with BD (Cross-Disorder Group of the Psychiatric Genomics Consortium et al., 2013; Marangoni et al., 2015; Mullins et al., 2021; Vöhringer and Perlis, 2016; Yamada et al., 2020). Finally, the associations between BD-PRS and key clinical characteristics at intake were examined in secondary analyses.

Methods

Participants

Data for this study were obtained from a subset of the COBY study and CYBD cohorts with records of genetics and longitudinal mood symptoms. Methods for these studies were previously described in detail elsewhere (Birmaher et al., 2009; Jiang et al., 2024). Briefly, the sample included 246 (n = 84 COBY; n = 162 CYBD) youth of European ancestry aged 7–20 years. Participants’ primary diagnoses were Diagnostic and Statistical Manual of Mental Disorders-4 (DSM-4) BD-I, BD-II, or operationally defined BD not otherwise specified (BD-NOS) (Axelson et al., 2006; Birmaher et al., 2006). Participants and a parent/guardian both provided written informed consent. The institutional review board for each study reviewed and approved the study protocol before the enrollment of participants.

Diagnostic interview and symptom ratings

At intake, the participant and parent/guardian were directly interviewed for the presence of current and lifetime psychiatric disorders using the Schedule for Affective Disorders and Schizophrenia for School Age Children, Present and Lifetime Version (K-SADS-PL), a semistructured interview used to assess present and lifetime history of psychiatric diagnoses (Kaufman et al., 1997). All interviews were performed by trained study personnel who had completed comprehensive K-SADS-PL training under the supervision of licensed child and adolescent psychiatrists. All diagnostic and symptom ratings were reviewed and confirmed by the licensed child and adolescent psychiatrists. Details regarding the collection of study data and clinical measures at intake are described in the Supplementary Data.

Mood symptom severity at intake was assessed through the K-SADS Mania Rating Scale and the K-SADS Depression Rating Scale (Axelson et al., 2003; Chambers et al., 1985). Week-by-week mood symptoms were assessed prospectively using the Adolescent Longitudinal Interval Follow-up Evaluation (A-LIFE) and quantified using the instrument’s Psychiatric Status Rating (PSR) scale, which is linked to DSM-4 criteria (Keller et al., 1987; Warshaw et al., 2001). DSM-4 criteria information is gathered in the interview and then translated into ratings for each week of the follow-up period. The PSR scores range from 1 for no symptoms, to 2–4 for increasing levels of subthreshold symptoms and impairment levels, to 5–6 full criteria with different degrees of severity or impairment. Consistent with the prior COBY studies (Birmaher et al., 2009, 2006), the percentage of follow-up weeks being euthymic or symptomatic in the different mood symptom categories (mania/hypomania, depression, or mixed) during the entire follow-up period were computed, based on the PSR scores for each subject. Change in mood polarity was defined as a switch between depression (rating ≥3) and mania/hypomania (rating ≥3) or vice versa with or without intervening weeks at the euthymic status or when a week had both mania/hypomania and depression rating scores of 3 or more (Birmaher et al., 2009, 2006; Judd et al., 2002).

Polygenic risk scores

Saliva collection, DNA extraction, and genotyping were performed separately for the CYBD and COBY study and have been previously described in detail (Birmaher et al., 2022; Jiang et al., 2024). Detailed methods regarding genetic quality control and imputation can be found in the Supplementary Data.

PRS was derived from GWAS summary statistics from the latest Psychiatric Genomics Consortium GWAS for BD (Mullins et al., 2021), SCZ (Trubetskoy et al., 2022), MDD (Howard et al., 2019), and ADHD (Demontis et al., 2023). PRSs were calculated using PRS-CS-auto, a Bayesian-based method that places a continuous shrinkage prior to the effect sizes of SNPs in the discovery GWAS summary statistics (Ge et al., 2019). PRS-CS-auto was implemented using the software default settings and with the 1000 Genomes European samples and HapMap3 SNPs as the LD reference panel provided with the PRS-CS software. PLINK 1.9 was used to weight all SNPs by their effect sizes calculated using PRS-CS-auto and sum all SNPs into PRS for each individual in the target cohort. The derived PRSs were subsequently standardized to a mean of 0 and an standard deviation (SD) of 1.

Statistical analysis

Analyses were performed using IBM SPSS (version 27) and R software (version 4.4.1). Beta regression models were used to examine the association between each psychiatric PRS and the percentage of weeks spent in different mood symptom categories. Linear regression models were used to examine the associations of each psychiatric PRS with mood symptom severity at intake and the number of changes in mood polarity per year. Covariates included age, sex, the first two genetic principal components, and study sites (Brown University, University of Pittsburgh, CYBD). Due to the paucity of prior literature and the exploratory nature of this analysis, we also presented significant results from univariable models (i.e., not adjusted for covariates) to provide a comprehensive evaluation that could guide future studies.

Results

The demographic and clinical characteristics of 246 youth with BD are presented in Table 1. Participants were 15.3 ± 3.2 years old and were predominantly female (61.8%). Among these participants, 168 participants had at least one follow-up assessment, and the majority (75.6%) were followed up for 3 years or more. Follow-up period ranged from 41 weeks to 19.8 years, with a median of 8.7 years (interquartile range: 3.2–16.5 years).

Table 1.

Demographic and Clinical Characteristics of 246 Youth with Bipolar Disorder at Intake

	Mean ± SD/n (%)
Demographics characteristics
Age	15.3 ± 3.2
Sex (n, % female)	152 (61.8)
SES	3.8 ± 1.1
Intact family	134 (54.5)
Tanner stage (1–5)	3.9 ± 1.3
Clinical characteristics
Age of onset	12.6 ± 4.3
BD subtype (BD-I/BD-II/BD-NOS)	86 (35.0)/55 (22.4)/10 (42.7)
Mood symptom severity and functioning
DRS—Depression score at intake	18.4 ± 11.9
DRS—Most severe lifetime depression score	29.4 ± 11.3
MRS—Mania score at intake	18.7 ± 13.0
MRS—Most severe lifetime Mania score	31.0 ± 8.7
CGAS—Most severe past episode	41.0 ± 11.7
CGAS—Highest past year	63.7 ± 11.7
CGAS—Past month	55.5 ± 11.7
Other characteristics
Lifetime psychosis	53 (21.5)
Lifetime suicide attempts	67 (27.2)
Lifetime self-injurious behavior	125 (50.8)
Lifetime suicidal ideation	174 (70.7)
Lifetime physical or sexual abuse	49 (19.9)
Lifetime psychiatric hospitalization	118 (48.0)
Lifetime comorbid diagnoses
ADHD	129 (52.4)
Any anxiety	180 (73.2)
SUD	58 (23.6)
ODD	76 (30.9)
CD	15 (6.1)
Family psychiatric history
BD	119 (48.4)
Depression	206 (83.7)
ADHD	96 (39.0)
Anxiety	163 (66.3)
Lifetime medications
SGA	136 (55.3)
Lithium	49 (19.9)
SSRI antidepressants	119 (48.4)
Non-SSRI antidepressants	55 (22.4)
Stimulants	86 (35.0)
Lamotrigine	12 (4.9)
Any medications	211 (85.8)
Current medications
SGA	82 (33.3)
Lithium	32 (13.0)
SSRI antidepressants	38 (15.4)
Non-SSRI antidepressants	17 (6.9)
Stimulants	38 (15.4)
Lamotrigine	10 (4.1)

Values for all continuous variables are presented as mean ± standard deviation (SD) and categorical variables are presented as n (% within group). All lifetime variables refer to lifetime occurrence at the time of intake.

ADHD, attention-deficit/hyperactivity disorder; BD, bipolar disorder; CD, conduct disorder; CGAS, Children’s Global Assessment Scale; DRS, Depression Rating Scale; MRS, Mania Rating Scale; NOS, not otherwise specified; ODD, oppositional defiant disorder; SES, socioeconomic status (range 1–5, with 1 a reference to unskilled/menial labor and 5 a reference to major professionals); SGA, second generation antipsychotic; SSRI, selective serotonin reuptake inhibitor; SUD, substance use disorder.

At intake, higher BD-PRS was associated with significantly lower most severe lifetime depression score (Table 2; β = −0.14, p = 0.03). Higher BD-PRS and SCZ-PRS were also associated with higher most severe lifetime mania score in univariable models (BD-PRS: β = 0.16, p = 0.02; SCZ-PRS: β = 0.14, p = 0.04), albeit not significant when adjusted for covariates (BD-PRS: β = 0.10, p = 0.13; SCZ-PRS: β = 0.06, p = 0.36). MDD-PRS and ADHD-PRS were not significantly associated with the mood symptom severity at intake.

Table 2.

Association Between Psychiatric PRSs and Mood Symptom Severity at Intake Among 246 Youth with Bipolar Disorder

Mood symptom severity at intake	β	SE	p	R²
Association with BD-PRS
DRS—Depression score at intake	−0.11	0.07	0.08	0.16
DRS—Most severe lifetime depression score	−0.14	0.06	0.03	0.28
MRS—Mania score at intake	0.05	0.07	0.51	0.05
MRS—Most severe lifetime Mania score	0.10	0.07	0.13	0.14
Association with SCZ-PRS
DRS—Depression score at intake	0.01	0.07	0.87	0.15
DRS—Most severe lifetime depression score	0.01	0.06	0.88	0.27
MRS—Mania score at intake	0.08	0.07	0.24	0.05
MRS—Most severe lifetime Mania score	0.06	0.07	0.36	0.14
Association with MDD-PRS
DRS—Depression score at intake	0.11	0.07	0.09	0.16
DRS—Most severe lifetime depression score	−0.02	0.06	0.77	0.27
MRS—Mania score at intake	−0.01	0.07	0.93	0.05
MRS—Most severe lifetime Mania score	−0.01	0.07	0.90	0.14
Association with ADHD-PRS
DRS—Depression score at intake	−0.04	0.06	0.57	0.15
DRS—Most severe lifetime depression score	−0.06	0.06	0.31	0.27
MRS—Mania score at intake	0.10	0.07	0.13	0.06
MRS—Most severe lifetime Mania score	0.01	0.07	0.94	0.14

All lifetime variables refer to lifetime occurrence at the time of intake.

Findings were adjusted for age, sex, two genetic principal components, and study site.

ADHD, attention-deficit/hyperactivity disorder; BD, bipolar disorder; DRS, Depression Rating Scale; MDD, major depressive disorder; MRS, Mania Rating Scale; SCZ, schizophrenia.

The boldface type indicates statistical significance with a p value <0.05.

Longitudinal course of mood symptoms and changes in mood polarity of 168 youth with BD, and their associations with psychiatric PRSs, are presented in Tables 3 and 4. BD-PRS was not significantly associated with any mood variable during follow-up. Higher SCZ-PRS and MDD-PRS were associated with significantly lower percentage of the follow-up time being euthymic (SCZ-PRS: β = −0.21, p = 0.02; MDD-PRS: β = −0.22, p = 0.01) and higher percentage of the follow-up time with any subsyndromal symptoms (SCZ-PRS: β = 0.15, p = 0.04; MDD-PRS: β = 0.17, p = 0.01). Additionally, higher SCZ-PRS was associated with higher percentage of the time with subsyndromal depression in the univariable model (β = 0.15, p = 0.03), albeit not significant when adjusted for covariates (β = 0.12, p = 0.11). Finally, higher ADHD-PRS was associated with higher percentage of the follow-up time with any subsyndromal symptoms in the univariable model (β = 0.14, p = 0.03), particularly subsyndromal mixed symptoms (β = 0.19, p = 0.02), albeit not significant when adjusted for covariates (any subsyndromal symptoms: β = 0.13, p = 0.06; subsyndromal mixed symptoms: β = 0.13, p = 0.12).

Table 3.

Week-by-Week Symptomatic Status and Changes in Mood Polarity Among 168 Youth with Bipolar Disorder

Symptomatic status	Mean ± SD
Euthymic (%)	53.56 ± 22.91
Syndromal (%)
Any	14.89 ± 14.03
Mania	0.84 ± 1.88
Hypomania	1.41 ± 2.83
Mixed/cycling	1.27 ± 2.81
Depression	11.67 ± 12.95
Subsyndromal (%)
Any	31.55 ± 16.78
Mania	6.79 ± 8.94
Mixed/cycling	6.70 ± 9.40
Depression	18.26 ± 12.23
Changes in mood polarity (times/year)	8.76 ± 6.90

MDD, major depressive disorder; SD, standard deviation.

Table 4.

Association Between Psychiatric PRSs and Week-by-Week Symptomatic Status Among 168 Youth with Bipolar Disorder

Symptomatic status	β	SE	p	R²
Association with BD-PRS
Euthymic (%)	0.001	0.08	0.99	0.03
Syndromal (%)
Any	0.003	0.08	0.97	0.04
Mania	0.01	0.08	0.89	0.02
Hypomania	−0.13	0.08	0.09	0.10
Mixed/cycling	−0.06	0.08	0.45	0.27
Depression	0.07	0.08	0.38	0.05
Subsyndromal (%)
Any	0.02	0.07	0.76	0.03
Mania	−0.01	0.08	0.95	0.13
Mixed/cycling	0.01	0.08	0.89	0.12
Depression	0.03	0.07	0.63	0.06
Changes in mood polarity (times/year)	−0.03	0.08	0.72	0.26
Association with SCZ-PRS
Euthymic (%)	−0.21	0.08	0.02	0.06
Syndromal (%)
Any	0.08	0.08	0.36	0.04
Mania	0.01	0.08	0.93	0.02
Hypomania	−0.06	0.08	0.49	0.08
Mixed/cycling	0.01	0.08	0.91	0.27
Depression	0.11	0.09	0.21	0.05
Subsyndromal (%)
Any	0.15	0.07	0.04	0.05
Mania	0.02	0.08	0.84	0.13
Mixed/cycling	0.11	0.09	0.19	0.14
Depression	0.12	0.07	0.11	0.06
Changes in mood polarity (times/year)	0.10	0.07	0.16	0.26
Association with MDD-PRS
Euthymic (%)	−0.22	0.08	0.01	0.08
Syndromal (%)
Any	0.10	0.08	0.19	0.05
Mania	0.03	0.08	0.68	0.02
Hypomania	0.04	0.08	0.60	0.08
Mixed/cycling	0.06	0.08	0.45	0.27
Depression	0.13	0.08	0.11	0.06
Subsyndromal (%)
Any	0.17	0.07	0.01	0.07
Mania	0.01	0.08	0.87	0.13
Mixed/cycling	0.07	0.08	0.40	0.13
Depression	0.02	0.07	0.78	0.05
Changes in mood polarity (times/year)	0.07	0.07	0.36	0.26
Association with ADHD-PRS
Euthymic (%)	−0.11	0.08	0.17	0.04
Syndromal (%)
Any	0.02	0.08	0.84	0.04
Mania	0.06	0.08	0.38	0.04
Hypomania	0.12	0.08	0.15	0.10
Mixed/cycling	0.03	0.08	0.70	0.27
Depression	−0.08	0.08	0.35	0.05
Subsyndromal (%)
Any	0.13	0.07	0.06	0.04
Mania	0.03	0.08	0.69	0.13
Mixed/cycling	0.13	0.08	0.12	0.14
Depression	0.01	0.07	0.85	0.05
Changes in mood polarity (times/year)	−0.004	0.07	0.95	0.26

Findings were adjusted for age, sex, two genetic principal component, and study site.

ADHD, attention-deficit/hyperactivity disorder; BD, bipolar disorder; MDD, major depressive disorder; SCZ, schizophrenia; SE, standard error.

The boldface type indicates statistical significance with a p value <0.05.

Exploratory analyses examined the association between each psychiatric PRS and the percentage of the follow-up time with any mania/hypomania symptoms (i.e., any syndromal mania, syndromal hypomania, or subsyndromal mania symptoms), and any depression symptoms (i.e., any syndromal MDD or subsyndromal depression symptoms). No significant associations were found between PRSs and percentage of the follow-up time with any mania/hypomania symptoms. Higher SCZ-PRS was associated with lower percentage of the follow-up time with any depression symptoms (β = 0.17; p = 0.04), whereas PRS for BD, MDD, and ADHD were not. Findings remained unchanged when additionally controlling for duration of illness and length of follow-up.

Discussion

The present study adds to the sparse literature regarding the prospective clinical course of BD in relation to psychiatric PRSs and particularly bridges a gap on this topic among youth with BD. The results showed that higher BD-PRS was associated with significantly lower most severe lifetime depression score at intake. Higher BD-PRS and SCZ-PRS were also associated with higher most severe lifetime mania score at intake, although these findings were not statistically significant when adjusted for covariates. BD-PRS was not associated with the longitudinal course of mood symptoms during follow-up, whereas higher SCZ-PRS and MDD-PRS were associated with significantly less time in euthymia and more time with any subsyndromal mood symptoms (i.e., any mania, mixed, or depression symptoms) during follow-up. Additionally, higher SCZ-PRS and ADHD-PRS were also associated with more time with subsyndromal depression and subsyndromal mixed symptoms, respectively, although these findings were not statistically significant when controlling for covariates.

BD-PRS was associated with less severe lifetime depressive symptoms at intake, aligning with findings from adults with BD that higher BD-PRS was associated with lower lifetime depression scores (Richards et al., 2022). Similarly, the significant univariable finding that higher BD-PRS and SCZ-PRS were associated with more severe lifetime mania symptoms at intake is also consistent with prior cross-sectional studies in adults with BD, SCZ, and MDD, and the general adult population (Bipolar Disorder and Schizophrenia Working Group of the Psychiatric Genomics Consortium, 2018; Mistry et al., 2019; Richards et al., 2022; Ruderfer et al., 2014; Wiste et al., 2014).

However, we did not find significant associations of BD-PRS with the longitudinal burden of mood symptoms, whereas in adults with BD higher BD-PRS was found to be significantly associated with increased odds of experiencing any episodes with manic polarity and more manic episodes during follow-up time (Hasseris et al., 2023). A number of factors may contribute to this discrepancy. The lack of findings in the current may be due to smaller sample size compared to prior study in adults (N = 2705). In addition, greater prospective burden of mood symptoms in youth versus adults with BD may have constrained variability in mood symptom burden (Birmaher et al., 2006). These two considerations could also have acted synergistically to reduce signal detection.

SCZ-PRS and MDD-PRS were associated with significantly less time euthymic and more time with any subsyndromal symptoms during follow-up. There was also evidence of higher SCZ-PRS and ADHD-PRS associated with more time with depression and mixed symptoms, respectively. Similarly in adults with BD, MDD-PRS has been correlated with greater burden of depressive and mixed symptoms and less burden of manic episodes during follow-up period, whereas SCZ-PRS has been associated with a greater burden of mania (Hasseris et al., 2023). Prior reports based on partially overlapping samples with the current study have reported that youth BD is characterized mainly by subsyndromal episodes, particularly with depressive and mixed episodes, and less frequently by syndromal episodes (Birmaher et al., 2009, 2006; Karthikeyan et al., 2022). The current study further suggests that the course of mood symptoms may partly relate to the polygenic liability to SCZ and MDD.

Limitations

There are some limitations to consider in the present study. First, the discovery GWAS summary statistics were predominantly derived from individuals of European ancestry, and PRS has lower accuracy in cross-population prediction when the target sample is genetically distant from the discovery GWAS sample. Therefore, this study only included individuals of European ancestry, and present findings may not be generalizable to individuals with other ancestral origins (Kachuri et al., 2024). Given that ancestry is an important factor in genetic structure, it is important to undertake GWAS in the ancestral populations to which the findings would be applied (Martin et al., 2019; Peterson et al., 2019; Wray et al., 2021). We echo other calls for well-powered GWAS in globally diverse populations, which would be essential to improve the potential clinical utility of PRS and ancestry-related health care disparities (Polygenic Risk Score Task Force of the International Common Disease Alliance et al., 2021; Ruan et al., 2022). Second, the sample size did not provide sufficient power to detect small effect sizes. Given the lack of studies on psychiatric PRSs and mood symptoms in youth with BD, we also reported findings unadjusted for covariates as a descriptive approach to provide comprehensive results that may help guide future studies. However, it is noteworthy that time-related covariates, such as age and duration of illness, are important factors that may impact the course of mood symptoms, and therefore should be considered when interpreting findings from the exploratory univariable models. Small sample size also limited controlling for additional covariates (e.g., BD subtypes, medication use, comorbidity, family history) or to examine each psychiatric PRS with key clinical characteristics (e.g., lifetime comorbidities). Third, although we controlled duration of illness and length of follow-up as covariates, there may still be residual effects of variations in the time of intake, duration of follow-up, and duration of illness. Future larger studies with a more controlled design (i.e., consistent timing of intake assessment and duration of follow-up across participants) are needed to limit the potential confounding effects of these factors on the association between PRS and the clinical course of mood symptoms. Fourth, youth with BD could have any level of mood symptom severity, and indeed the sample included youth recruited from differing sources, ranging from inpatient units to outpatient clinics. Although this decision increases the representativeness of findings, it also introduces greater heterogeneity which may have impacted present findings. Fifth, PRS does not include other sources of genetic variations (e.g., rare, complex, sex chromosome, mitochondrial genetic variants) and only explains a relatively small proportion of phenotypic variance. PRS therefore may have limited explanatory power on the heterogeneous and complex clinical course of BD (Lewis and Vassos, 2020). Additionally, the psychiatric PRSs in this study were derived from GWAS in adults with the full spectrum of onset age. Given that early-onset BD is especially severe and heritable (Faraone et al., 2003), and it has been shown that the effects of genetic risk may vary with age (Jiang et al., 2021), PRSs derived from GWAS that specifically focuses on adults with early age of onset, or ideally youth, would enhance the potential clinical utility of PRS in prediction of the clinical mood course or personalized treatment in youth with BD.

Future directions and conclusions

There remain considerable gaps in prognostication of key issues such as symptomatic burden precision, functioning, efficacy, and safety/tolerability of treatments (Bauer et al., 2018). The current study represents an important initial step that examines the usefulness of PRS in predicting prospective symptomatic status among youth with BD. The design and results from the current study can serve to guide future youth BD studies that further integrate PRS to advance the long-term clinical management of youth BD. Larger samples would enable more precise prediction taking into account sources of variability such as age, sex, BD subtype, comorbidities, and environmental factors. Additionally, future analyses are needed to evaluate the relevance of genetics to factors such as mood fluctuations, polarity switches, specific symptom domains, and less common events such as suicide attempts and hospitalizations. Finally, future studies are warranted to evaluate the concordance of PRS and symptoms in domains/comorbidities such as psychosis, ADHD, anxiety, and addictions. Maximizing clinical utility will require that PRSs explain more variance, apply to a broader segment of the population, and encompass nonmood aspects of BD. Toward this end, it will be important to generate PRS from GWAS focused specifically on early-onset BD and that apply to multiple ancestries. Additionally, adding the BD-PRS to clinical/demographic predictors in risk calculators has yielded significantly improved its accuracy in individual risk prediction (Hafeman et al., 2023). Future studies that combine PRS with demographic, clinical, familial, and environmental variables, as well as incorporate additional genetic factors have the potential to enhance the contribution of genetics to optimize diagnostic classification, risk prediction, and prognostic precision, with the ultimate goal of greater benefit to a broader proportion of individuals living with BD.

Footnotes

Acknowledgments

The authors would like to thank all staff, study participants, and families for their time and contributions to this study.

Author’s Contributions

X.J. primarily wrote the article and performed statistical analyses. C.C.Z. conducted GWAS quality control and PRS calculations. J.M. assisted calculation of the longitudinal mood variables. B.I.G. contributed to study conception, design, and assisted with article preparation. All authors contributed to revisions of the article and have approved the final article.

Availability of Data and Material

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request. The data are not publicly available due to privacy or ethical restrictions.

Disclosures

B.I.G. acknowledges his position as RBC Investments Chair in Children’s Mental Health and Developmental Psychopathology at CAMH, a joint Hospital–University Chair between the University of Toronto, CAMH, and the CAMH Foundation. All other authors report no actual or potential conflicts of interest.

Supplementary Material

Supplementary Data

References

Aminoff

, Tesli

, Bettella

, et al. Polygenic risk scores in bipolar disorder subgroups. J Affect Disord, 2015; 183:310–314.

Andlauer

TFM

, Guzman-Parra

, Streit

, et al; Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium. Bipolar multiplex families have an increased burden of common risk variants for psychiatric disorders. Mol Psychiatry, 2021; 26(4):1286–1298.

Axelson

, Birmaher

, Brent

, et al. A preliminary study of the kiddie schedule for affective disorders and schizophrenia for school-age children mania rating scale for children and adolescents. J Child Adolesc Psychopharmacol, 2003; 13(4):463–470.

Axelson

, Birmaher

, Strober

, et al. Phenomenology of children and adolescents with bipolar spectrum disorders. Arch Gen Psychiatry, 2006; 63(10):1139–1148.

Barnett

, Smoller

. The genetics of bipolar disorder. Neuroscience, 2009; 164(1):331–343.

Bauer

, Andreassen

, Geddes

, et al. Areas of uncertainties and unmet needs in bipolar disorders: Clinical and research perspectives. Lancet Psychiatry, 2018; 5(11):930–939.

Bipolar Disorder and Schizophrenia Working Group of the Psychiatric Genomics Consortium. Genomic dissection of bipolar disorder and schizophrenia, including 28 Subphenotypes. Cell, 2018; 173(7):1705–1715.e16.

Birmaher

, Axelson

, Goldstein

, et al. Four-Year longitudinal course of children and adolescents with bipolar spectrum disorders: The Course and Outcome of Bipolar Youth (COBY) Study. Am J Psychiatry, 2009; 166(7):795–804.

Birmaher

, Axelson

, Strober

, et al. Clinical course of children and adolescents with bipolar spectrum disorders. Arch Gen Psychiatry, 2006; 63(2):175–183.

10.

Birmaher

, Hafeman

, Merranko

, et al. Role of polygenic risk score in the familial transmission of bipolar disorder in youth. JAMA Psychiatry, 2022; 79(2):160–168.

11.

Boies

, Mérette

, Paccalet

, et al. Polygenic risk scores distinguish patients from non-affected adult relatives and from normal controls in schizophrenia and bipolar disorder multi-affected kindreds. Am J Med Genet B Neuropsychiatr Genet, 2018; 177(3):329–336; doi: 10.1002/ajmg.b.32614

12.

Chambers

, Puig-Antich

, Hirsch

, et al. The assessment of affective disorders in children and adolescents by semistructured interview. Test-retest reliability of the schedule for affective disorders and Schizophrenia for school-age children, present episode version. Arch Gen Psychiatry, 1985; 42(7):696–702.

13.

Charney

, Ruderfer

, Stahl

, et al. Evidence for genetic heterogeneity between clinical subtypes of bipolar disorder. Transl Psychiatry, 2017; 7(1):e993.

14.

Cross-Disorder Group of the Psychiatric Genomics Consortium, Lee

, Ripke

, et al. Genetic relationship between five psychiatric disorders estimated from genome-Wide SNPs. Nat Genet, 2013; 45(9):984–994.

15.

Demontis

, Walters

, Athanasiadis

, et al; iPSYCH-Broad Consortium. Genome-wide analyses of ADHD identify 27 risk loci, refine the genetic architecture and implicate several cognitive domains. Nat Genet, 2023; 55(2):198–208.

16.

Faraone

, Glatt

, Tsuang

. The genetics of pediatric-onset bipolar disorder. Biol Psychiatry, 2003; 53(11):970–977.

17.

Fullerton

, Koller

, Edenberg

, et al; Bipolar High Risk Study Group, BiGS Consortium. Assessment of first and second degree relatives of individuals with bipolar disorder shows increased genetic risk scores in both affected relatives and young at-risk individuals. Am J Med Genet B Neuropsychiatr Genet, 2015; 168(7):617–629.

18.

, Chen

C-Y

, Ni

, et al. Polygenic prediction via bayesian regression and continuous shrinkage priors. Nat Commun, 2019; 10(1):1776.

19.

Goldstein

, Birmaher

, Carlson

, et al. The international society for bipolar disorders task force report on pediatric bipolar disorder: Knowledge to date and directions for future research. Bipolar Disord, 2017; 19(7):524–543.

20.

Gordovez

FJA

, McMahon

. The genetics of bipolar disorder. Mol Psychiatry, 2020; 25(3):544–559.

21.

Gore

, Bloem

PJN

, Patton

, et al. Global burden of disease in young people aged 10-24 years: A systematic analysis. Lancet, 2011; 377(9783):2093–2102.

22.

Hafeman

, Uher

, Merranko

, et al. Polygenic risk score and individual risk prediction of bipolar disorder. Biol Psychiatry, 2023; 93(9):S66–S67.

23.

Hasseris

, Albiñana

, Vilhjalmsson

, et al. Polygenic risk and episode polarity among individuals with bipolar disorder. Am J Psychiatry, 2023; 180(3):200–208.

24.

Howard

, Adams

, Clarke

T-K

, et al; Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium. Genome-wide meta-analysis of depression identifies 102 independent variants and highlights the importance of the prefrontal brain regions. Nat Neurosci, 2019; 22(3):343–352.

25.

Jiang

, Holmes

, McVean

. The impact of age on genetic risk for common diseases. PLoS Genet, 2021; 17(8):e1009723.

26.

Jiang

, Zai

, Dimick

, et al. Psychiatric polygenic risk scores across youth with bipolar disorder, youth at high risk for bipolar disorder, and controls. J Am Acad Child Adolesc Psychiatry, 2024; 63(11):1149–1157; doi: 10.1016/j.jaac.2023.12.009

27.

Judd

, Akiskal

, Schettler

, et al. The long-term natural history of the weekly symptomatic status of Bipolar I disorder. Arch Gen Psychiatry, 2002; 59(6):530–537.

28.

Kachuri

, Chatterjee

, Hirbo

, et al; Polygenic Risk Methods in Diverse Populations (PRIMED) Consortium Methods Working Group. Principles and methods for transferring polygenic risk scores across global populations. Nat Rev Genet, 2024; 25(1):8–25.

29.

Karthikeyan

, Dimick

, Fiksenbaum

, et al. Inflammatory markers, brain-derived neurotrophic factor, and the symptomatic course of adolescent bipolar disorder: A prospective repeated-measures study. Brain Behav Immun, 2022; 100:278–286.

30.

Kaufman

, Birmaher

, Brent

, et al. Schedule for affective disorders and Schizophrenia for School-Age Children-Present and Lifetime Version (K-SADS-PL): Initial reliability and validity data. J Am Acad Child Adolesc Psychiatry, 1997; 36(7):980–988.

31.

Keller

, Lavori

, Friedman

, et al. The longitudinal interval follow-up evaluation. a comprehensive method for assessing outcome in prospective longitudinal studies. Arch Gen Psychiatry, 1987; 44(6):540–548.

32.

Leverich

, Post

, Keck

Jr , et al. The poor prognosis of childhood-onset bipolar disorder. J Pediatr, 2007; 150(5):485–490.

33.

Lewis

, Vassos

. Polygenic risk scores: From research tools to clinical instruments. Genome Med, 2020; 12(1):44.

34.

Marangoni

, De Chiara

, Faedda

. Bipolar disorder and ADHD: Comorbidity and diagnostic distinctions. Curr Psychiatry Rep, 2015; 17(8):604.

35.

Martin

, Kanai

, Kamatani

, et al. Clinical use of current polygenic risk scores may exacerbate health disparities. Nat Genet, 2019; 51(4):584–591.

36.

McKnight

, Bilderbeck

, Miklowitz

, et al. Longitudinal mood monitoring in bipolar disorder: Course of illness as revealed through a short messaging service. J Affect Disord, 2017; 223:139–145.

37.

Mistry

, Escott-Price

, Florio

, et al. Genetic risk for bipolar disorder and psychopathology from childhood to early adulthood. J Affect Disord, 2019; 246:633–639.

38.

Mullins

, Forstner

, O’Connell

, et al; HUNT All-In Psychiatry. Genome-Wide association study of more than 40,000 bipolar disorder cases provides new insights into the underlying biology. Nat Genet, 2021; 53(6):817–829.

39.

Owen

, Williams

. Explaining the missing heritability of psychiatric disorders. World Psychiatry, 2021; 20(2):294–295.

40.

Peterson

, Kuchenbaecker

, Walters

, et al. Genome-Wide association studies in ancestrally diverse populations: Opportunities, methods, pitfalls, and recommendations. Cell, 2019; 179(3):589–603.

41.

Polygenic Risk Score Task Force of the International Common Disease Alliance, Adeyemo

, Balaconis

, et al. Responsible use of polygenic risk scores in the clinic: Potential benefits, risks and gaps. Nat Med, 2021; 27(11):1876–1884.

42.

Power

, Steinberg

, Bjornsdottir

, et al. Polygenic risk scores for schizophrenia and bipolar disorder predict creativity. Nat Neurosci, 2015; 18(7):953–955.

43.

Richards

, Cardno

, Harold

, et al. Genetic liabilities differentiating bipolar disorder, schizophrenia, and major depressive disorder, and phenotypic heterogeneity in bipolar disorder. JAMA Psychiatry, 2022; 79(10):1032–1039.

44.

Ruan

, Lin

Y-F

, Feng

Y-CA

, et al; Stanley Global Asia Initiatives. Improving polygenic prediction in ancestrally diverse populations. Nat Genet, 2022; 54(5):573–580.

45.

Ruderfer

, Fanous

, Ripke

, et al; Cross-Disorder Working Group of the Psychiatric Genomics Consortium. Polygenic dissection of diagnosis and clinical dimensions of bipolar disorder and Schizophrenia. Mol Psychiatry, 2014; 19(9):1017–1024.

46.

Trubetskoy

, Pardiñas

, Qi

, et al; Schizophrenia Working Group of the Psychiatric Genomics Consortium. Mapping genomic loci implicates genes and synaptic biology in Schizophrenia. Nature, 2022; 604(7906):502–508.

47.

Van Meter

, Moreira

ALR

, Youngstrom

. Updated meta-analysis of epidemiologic studies of pediatric bipolar disorder. J Clin Psychiatry, 2019; 80(3); doi: 10.4088/JCP.18r12180

48.

Vöhringer

, Perlis

. Discriminating between bipolar disorder and major depressive disorder. Psychiatr Clin North Am, 2016; 39(1):1–10.

49.

Warshaw

, Dyck

, Allsworth

, et al. Maintaining reliability in a long-term psychiatric study: An ongoing inter-rater reliability monitoring program using the longitudinal interval follow-up evaluation. J Psychiatr Res, 2001; 35(5):297–305.

50.

Wiste

, Robinson

, Milaneschi

, et al. Bipolar polygenic loading and bipolar spectrum features in major depressive disorder. Bipolar Disord, 2014; 16(6):608–616.

51.

Wray

, Lin

, Austin

, et al. From basic science to clinical application of polygenic risk scores: A primer. JAMA Psychiatry, 2021; 78(1):101–109.

52.

Yamada

, Matsumoto

, Iijima

, et al. Specificity and continuity of Schizophrenia and bipolar disorder: Relation to biomarkers. Curr Pharm Des, 2020; 26(2):191–200.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB