Emerging psychopathology and clinical staging in adolescent offspring of parents with bipolar disorder or schizophrenia—A longitudinal study

Offspring of parents with bipolar disorder (BDo) and schizophrenia (SZo) are at increased risk for these disorders and general psychopathology. Little is known about their (dis)similarities in risk and developmental trajectories during adolescence. A clinical staging approach may help define the developmental course of illness.


| INTRODUC TI ON
Schizophrenia (SZ) and bipolar disorder (BD) are severe mental illnesses.SZ affect 0.33%-0.7% 1,2and BD affect 0.81%-1.29% 3of the worldwide population.A positive family history of SZ or BD is the strongest predictor for the development of these severe mental illnesses. 4Prospective familial high-risk studies may inform us on the early trajectories and general risk of psychopathology.In addition to genetic predisposition, having a parent with SZ or BD has been associated with an increased burden of environmental stressors in childhood that further increases the risk of developing a psychiatric disorder. 5Using a prospective cross-disorder approach, studies on offspring of parents with SZ (SZo) or BD (BDo) can inform us on the similarities and dissimilarities in neurodevelopmental trajectories of their risk for psychopathology. 6e development of psychopathology during adolescence is of great interest since it is known that neurodevelopmental deviation prior to clinical onset plays an important role in the pathophysiology of the severe mental illnesses.Additionally, the risk for developing psychopathology is high during adolescence 7 and prodromal symptoms start even earlier. 8Moreover, during adolescence, typical adolescent behavior and prodromal symptoms of severe mental illness overlap and are difficult to differentiate. 9,10Familial high-risk studies show that both SZo and BDo have an elevated risk to develop homotypic psychopathology, that is psychotic and mood disorders, as well as heterotypic psychopathology, that is anxiety or developmental disorders. 4,11Importantly, subthreshold symptoms of mania and depression are highly predictive of future manic and depressive episodes in BDo, 12,13 as well as anxiety disorders, which is confirmed by the clinical staging perspective. 14 date, four studies using a cross-disorder approach have compared the risk of psychopathology in offspring at familial risk for SZ or BD (see Table S1).6][17][18][19][20] Some studies reported higher prevalence rates of mood 15 or more anxiety symptoms in BDo compared to SZo. 17 Psychotic experiences and developmental and disruptive disorders were more prevalent in SZo compared to BDo. 15,16,21 Others found no significant differences between these groups, 18,20 although one of these studies showed an overall increase in dimensional psychopathology in 7-year-old SZo compared to BDo at baseline. 19The mean age of these studies varied from 7.8 to 17.5 (see Table S1).Although a higher prevalence of psychopathology in both familial risk groups is evident in these studies, whether developmental trajectories are similar across groups or diverge during adolescence is not fully understood due to the relatively young age [15][16][17]19 and/or lack of longitudinal design 17,18 of these cross-disorder studies.
Clinical staging models have the potential to improve the logic and timing of interventions in psychiatry but are rarely used to enhance early recognition in mental illnesses, [22][23][24] despite the known benefits from its use for physical illnesses. 25They define the progression of illness rather than the diagnosis itself. 22Studies on the current clinical staging models for SZ or BD focused on the clinical implications of the models, cognitive and neurobiological biomarkers, and on offspring at familial risk [26][27][28][29][30][31][32] but do not use a crossdisorder approach and rarely have a developmental perspective.
Moreover, among present cross-disorder studies, a clinical staging model perspective is lacking.Incorporating both approaches may shed light on (dis)similarities in illness age of onset, course/progression and hetero/homotypic development of psychopathology.
This study aims to investigate the natural trajectories of psychopathology in SZo, BDo, and Co from late childhood/early adolescence onward (mean age of 13.2 years; range: 8-18 years) up to a mean age of 17.1 years (range 11-22), using a unique prospective cohort study.
We will examine the natural course of emerging psychopathology by comparing the three groups, using three approaches.First, the presence of lifetime categorical psychopathology will be compared between groups.Second, studying the timing and development of these diagnoses using a clinical staging model will help identify windows of opportunity for early interventions in both at risk groups.
Finally, as (young) adolescent offspring may show increased symptomatology without a full-blown diagnoses, we will also compare levels of (sub)clinical symptoms by taking a dimensional approach.

| Participants and procedure
The Dutch Bipolar and Schizophrenia Offspring study (DBSOS) is an ongoing prospective cohort study, investigating the development of brain, genetics, cognitive functioning, and environment, that contribute to risk and resilience in SZo and BDo.In total, 58 SZo (38 families), 94 BDo (60 families), and 56 Co (34 families), aged 8-18 years old, were included between 2010 and 2017.4][35] Figure S1 presents detailed recruitment and drop-out characteristics.Overall, 88.5% of the offspring participated at 4-year follow-up: 51 SZo (87.9%), 84 BDo (89.4%), and 49 Co (87.5%).Families with at least one firstdegree relative (mostly one affected parent; one family with two parents affected with BD-I and two families with a BD sibling), or two second-degree relatives with SZ (four families) or BD (one family) were recruited via adult psychiatrists and child-and-adolescent psychiatrists in the Netherlands, advertisements (e.g., lectures, patient advocacy groups, and study website) or hear-say (e.g., hospital staff or other participants).Control families were recruited via advertisements on schools and leisure clubs (i.e., lectures and flyers) or hear-say.Parental diagnoses were confirmed with the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I 36 ).In case of psychiatric admission, death or unknown whereabouts of the index parent (15.8% of BD and 14.7% of SZ), parental diagnosis was retrieved from the Family Interview for Genetic Studies (FIGS). 37,38Co-parents and control parents were interviewed using the mini Schedule for Clinical Assessment in Neuropsychiatry (mini-SCAN) interview, 39 followed by the SCID-I in case of reported psychopathology.
Exclusion criteria were an IQ < 70, a major medical history or history of neurological illness, and additionally for Co, a first-degree relative with a severe mood or psychotic disorder.
Written informed consent was obtained from all offspring older than 12 years and both parents or legal caregivers of children younger than 18 years old.The study was conducted at the University Medical Center Utrecht (UMCU) in the Netherlands.The medical ethics committee of the UMCU approved the study protocol.All participants received a financial compensation for their time participating in several measurements (i.e., interviews, questionnaires, neuropsychological tests, blood withdrawal, and MRI) and travel expenses were reimbursed.

| Measures
Clusters of Lifetime DSM-IV axis-I disorders: The presence of current and lifetime DSM-IV diagnoses in offspring was examined with the Kiddie Schedule for Affective Disorders and Schizophrenia for School Age Children Present and Lifetime Version (K-SADS-PL). 40is interviewer-oriented semi-structured face-to-face interview was administered to offspring and parents separately.Screen items cover symptoms of depression, mania, psychosis, anxiety, developmental, substance, and other non-mood disorders, allowing severity scores of 1 (absent), 2 (mild), and 3 (severe).Supplementary items were always administered for the depression, mania, and psychosis domains, and for the remaining domains when at least one of the screen items in that specific domain was scored with a severity score of 3. Symptoms were assessed using specific criteria of the instrument, 40 based on the criteria of the DSM-IV.
Consensus scores were based on both child and parent perspectives to establish lifetime DSM-IV axis-I diagnoses, giving a greater weight to the parent's report for observed behavior and to the child's report for subjective experiences in case of disagreement.Autism spectrum diagnoses were based on psychiatric evaluations from outpatient clinics.During the K-SADS interview, information on age of onset, medication use, and treatment history was obtained.The Children's Global Assessment Scale (CGAS) was used to assess the level of functioning (i.e., social, school/occupational, and personal functioning). 41inical Staging Model: Duffy 29 proposed a staging model for BDo with stages that take into account the developmental perspective and early precursors of the classical episodic BD as well as psychotic features of the bipolar spectrum: 0 (well, with familial risk); 1 (nonspecific syndromes/non-specific syndromes and developmental disorders); 2 (mild mood or single depressive episode/negative syndrome); 3 (recurrent major depressive disorder/attenuated psychotic syndrome); and 4 (BD/mixed [mania/psychotic/cyclic] or psychotic disorder).This study slightly adapted this model, not only profiting from the offspring approach, but also allowing a cross-disorder approach that acknowledges the heterotypic nature of SZ and BD. 42age 0 (well) was divided into stage 0 a (well, without familial risk) and 0 b (well, with familial risk).Stage 1 is referred to as the non-mood stage.Within stage 1, we distinguished "developmental," "anxiety," "alcohol/substance," and "other non-mood" disorders.Psychosis was clustered with manic episodes in stage 4.Each transition through the stages and corresponding age of onset were based on the DSM-IV diagnoses retrieved from the K-SADS-PL interviews (at either baseline, follow-up, or both).The absence of lifetime DSM-IV diagnoses at either baseline or both waves was registered as stage 0 a (Co) or 0 b (SZo and BDo).Comorbid diagnoses with the same age of onset are registered as the highest stage according to the hierarchy of this model.
(Sub)clinical symptoms: In order to evaluate the symptomatology that may precede the onset of categorical psychopathology, summary scores for (sub)clinical symptoms were calculated at baseline and follow-up of both screen and supplementary items.To prevent missing values, we only summarized data that was administered to all participants; thus, both screen and supplementary items were used for depression, mania, and psychosis domains, and only screen items were used for anxiety and behavioral problems.
Additionally, parent-, self-, and teacher-reports of the Achenbach System of Empirically Based Assessment (ASEBA) 43,44 questionnaires were administered at both waves.Parents reported on their offspring's behavior using the Child Behavior Checklist (CBCL: 8-18 years) or the Adult Behavior Checklist (ABCL; 19+).Offspring filled out the Youth Self Report (YSR; 12-18 years) or the Adult Self Report (ASR; 19+).The Teacher's Report Form (TRF) was administered to primary or secondary school teachers of participants.
T-scores of internalizing and externalizing scales were calculated with ASEBA PC (https://aseba.org/aseba-pc/).Both parents of SZo and BDo and one parent (by choice) of Co were asked to fill out the parent-report.Mothers were included as main informant in this study and fathers were included in the absence of the mother's questionnaire (baseline: 4% SZo, 2% BDo, and 2% Co; follow-up: 9% SZo, 12% BDo, and 24% Co).

IQ: Four subtasks of the Wechsler Intelligence Scale for
Children-III (WISC-III) 45 for participants aged 8-16 or Wechsler Adult Intelligence Scale-III (WAIS-III) 46 for participants 17 years or older were administered (i.e., picture arrangement, block design, vocabulary, and information) to estimate IQ.

All assessments were conducted by trained interviewers with a
Bachelor's or Master's degree in medicine/psychology, supervised by a licensed psychologist (EM) and a psychiatrist certified in adult and child-and-adolescent psychiatry (MH).

| Statistical analyses
Statistical analyses were checked for assumptions and replaced with appropriate alternatives in case of violations.
Demographic information: Group differences at baseline and follow-up on dichotomous demographic variables, treatment and level of functioning were tested with chi-squared (X 2 ) tests.Differences on continuous demographic variables were tested with one-way ANOVA's, or their non-parametric equivalents.Post hoc pairwise comparisons were done with a Tukey's HSD in case of significant main effects for group.

Clusters of Lifetime DSM-IV axis I disorders:
Lifetime prevalence of DSM-IV axis I disorders in each cluster was compared between groups with chi-squared (X 2 ) tests.Survival curves using Kaplan-Meier survival analyses 47 were calculated for each cluster, determining the probability of illness onset, given the age of the latest known information.Survival analyses were extended using cox proportional hazard regression analyses including a correction for age.Finally, the number of comorbid diagnoses was compared between groups with a Kruskal-Wallis test.

Clinical staging:
We registered all lifetime DSM-IV psychopathology and corresponding ages of onset for all subjects, assigning each of them to the stages in our clinical staging model.First, the prevalence of each latest known stage and number of transitions was compared between groups with chi-squared (X 2 ) tests.Second, survival curves using Kaplan-Meier survival analyses were calculated for the prevalence of stage 1 (non-mood), stage 2 (single/mild mood), and stage 3 and 4 combined (recurrent depressive mood and manic/psychotic), thereby determining the probability of transitions to a later stage.Cox proportional hazard analyses, correcting for age at the interview and sex were done to calculate the hazard ratio.Finally, exploratory analyses were done to investigate the differences in precursors of stage 2 (i.e., first single mood episode) between SZo and BDo.
(Sub)clinical symptoms: K-SADS-PL summary scores of depression, mania, psychosis, anxiety, and behavioral problems (i.e., symptoms of attention-deficit [hyperactivity], oppositional defiant or conduct disorders), and t-scores from the ASEBA questionnaires were compared between groups using linear mixed models, with group and age at the waves as fixed factors and subject-id as random factor.Evaluation of AIC showed that an additional age*group interaction effect did not improve the models.
Sensitivity analyses: All analyses were repeated after excluding subjects with multiple second-degree family members, but without an affected parent (BDo: N = 4, three families; SZo: N = 6, four families).
Additionally, to correct for the effect of age and sex, the analyses for the cluster of DSM-IV categories any diagnosis, any non-mood and any mood diagnosis are repeated for offspring below and above the age of 17.5 years (median age at follow-up), and for the sexes separately.
Correction for multiple testing was done for 32 analyses (main analyses of lifetime diagnoses, clinical staging model, and (sub)clinical symptoms) using the false rate discovery-Benjamini-Hochberg procedure, with a false discovery rate of 5% (α < 0.034 for main analyses and α < 0.05 for post hoc pairwise comparisons).

| Clusters of lifetime DSM-IV axis I disorders
The prevalence of each cluster of DSM-IV psychopathology at baseline and follow-up is presented in Table 2. Overall, an increase of psychopathology is seen between baseline and follow-up.At the mean age of 17.1 years, the lifetime prevalence of any DSM-IV axis-I disorder was highest in SZo (72.5%), followed by BDo (64.3%) and Co (30.6%); the difference between SZo and BDo was not significant (X 2 = 0.99, p = 0.349).Significant differences were found in the "non-mood cluster" and "mood cluster."The "nonmood cluster" was most prevalent in SZo (70.6%) and significantly more prevalent than BDo (47.6%;X 2 = 6.81, p < 0.012).Subsequent analyses within the "non-mood cluster" showed significantly more developmental disorders in SZo (43.1%) as compared to controls (16.3%)(X 2 = 8.55, p = 0.004).Substance disorders were more frequent in BDo (10.7%) as compared to Co (0.0%) (X 2 = 5.63, p = 0.026).Both SZo and BDo reported more comorbidity (49.0% and 50.0%) compared to Co.
Taking into account age of onset, survival distributions per group and DSM-IV clusters were plotted using Kaplan-Meier survival curves (Figure 1).As shown in Figure 1, for both clusters, the survival distributions of the familial high-risk groups were significantly different from Co.Only in the "non-mood cluster" SZo (median age of onset: 18.0; CI95%: 16.6-19.4),survival distribution was significantly different from BDo (median age of onset: 18.0; CI95%: 16.6-19.4)(Exp(B) = 2.17, p < 0.001) after correction for age and sex .

| Clinical staging
The distribution of stages and number of transitions are presented in Table 2.At the mean age of 17.1 years, stage 0 a /0 b was most common in Co (69.4%) followed by BDo (35,7%) and SZo (27.5%).We found no significant difference in prevalence of stage 1-4 at this age.As shown in Table 2, the total number of transitions between age 13.2 and 17.1 was significantly higher in SZo (One transition: 52.9%, two or more: 19.6%) and BDo (One transition: 35.7%, two or more: 28.6%) as compared to Co (One transition: 24.5%, two or more: 6.1%).Individual transitions between stages over time are depicted in Figure 2. Exploratory analyses showed no significant difference in prevalence of precursors of stage 3 (mild or single mood) between SZo and BDo (data not shown).Figure 1B shows the survival distribution per stage and group.Significant group differences were found for stage 1 and stage 2, indicating that the timing and risk to reach one of these stages differed between groups during adolescence.Again, differences between BDo and SZo were only significant for stage 1.

| Sensitivity analyses
Results after the sensitivity analyses excluding subjects with multiple second-degree family members did not alter findings (data not shown).The results of the sensitivity analyses for age and sex are presented in Table S2.Sensitivity analyses for age suggest that the higher prevalence rates for the different types of psychopathology was only statistically significant between BDo and Co in the subgroups older than 17.5 years old.Differences between SZo and Co for any diagnosis and non-mood disorders was statistically significant in the subgroups younger as well as older than 17.5 years old.
Sensitivity analyses for sex suggest that especially male offspring at familial risk are more vulnerable for psychopathology.These results should be interpreted with care due to small sample sizes per subgroup.

| DISCUSS ION
This prospective study among offspring of parents with SZ (SZo) or BD (BDo) cover the window of early childhood until late adolescence using a longitudinal design.The aim was to examine the (dis) similarities in the course of emerging psychopathology and clinical staging among this high-risk population.In sum, this study showed that both SZo and BDo are at increased risk for psychopathology.
Moreover, compared to BDo, SZo were more likely to develop nonmood psychopathology at young age.At a dimensional level, more (sub)clinical symptoms of developmental and mood disorders were present in SZo as compared to BDo and Co.The multi-informant approach showed that information gathered via interviews, parents, or teachers, in addition to the self-report, were important, especially for externalizing symptoms.
Importantly, Maziade and authors included a sample of selected families with high familial load for SZ and BD, while our sample may have been more heterogeneous.The higher risk of non-mood psychopathology in SZo as compared to BDo, confirms findings of the Spanish cross-diagnostic cohort study that developmental and disruptive disorders were more prevalent in SZo compared to BDo. 15,16 Unlike, De la Serna, et al. 15 this study did not replicate higher prevalence rates of mood disorders in BDo than in SZo.However, the Spanish cohort is approximately 3 years younger than the DBSOS cohort, and within their study, SZo were significantly younger than BDo (see Table S1).The survival distribution of the mood cluster in Figure 1 illustrates that age may be a possible explanation for this divergent finding.Finally, this study did not replicate the finding of higher levels of anxiety 17 in prevalence rate nor in the dimensional data.Cultural or methodological factors may explain the difference in findings.

Clinical staging
Stage 0 a /0 b : well with(out) familial risk Fisher's exact 0.305

TA B L E 2 (Continued)
The clinical staging approach allowed us to explore differences in disorders or stages preceding first mood episodes between groups, which may provide a tool for prognostic orientation or opportunities for intervention.As expected, both familial risk groups have showed more transitions between stages, but staging did not reveal clear differences between SZo and BDo.By evaluating survival curves of the DSM-IV diagnoses clusters as well as clinical stages, we could confirm that non-mood disorders are more prevalent at younger age in SZo than BDo.Although the rates of non-mood disorders (stage 1) did not differ in frequency, our staging perspective revealed the importance of risk profiling using a longitudinal course.
Taken together, this study underscores the heterotypic character of SZ and BD have been identified as life-course illnesses with different clinical manifestations from an at-risk to a late stage. 28ether regression to previous stages, as a result of remission of symptoms, is possible is still under debate in the literature.
Although symptoms can go in remission (current clinical state), as underlying disease processes may be irreversible, clinical staging is considered unidirectional. 42Offspring will therefore retain their risk for progression to further stages, irrespective of their current clinical state.
In line with previous studies, 16,18,19 both SZo and BDo reported more (sub)clinical symptoms in the domains of depression, psychosis, anxiety, and developmental symptoms than Co.SZo reported more symptoms of developmental disorders and depression than BDo; this was reflected in interviewer-based-, parent-, and teacher-, but not self-report, stressing the importance of a multi-informant approach.
Mania symptoms were most prevalent in BDo, and have been previously marked as important risk factor for BD in BDo. 12,13,53Mania symptoms may belong to the specific prodrome for BD, yet longer follow-up of these cross-disorder studies are needed.
Overall, results show that SZo and BDo follow a similar developmental course of emerging psychopathology, although SZo are more vulnerable for non-mood disorders at a younger age and for symptoms of developmental and mood disorders.This may point toward a different ethiopathophysiology.In an earlier study on brain structure in a subsample of the current cohort, we found evidence of smaller intracranial volumes in SZo compared to BDo (a difference of 1%). 33As ICV is considered a marker for neurodevelopment, 54,55 a smaller ICV is suggestive of neurodevelopmental factors that are likely more at play in relation to familial risk for SZ than BD. 56Future studies are essential to further address the relationship between brain developmental and clinical trajectories.Long-term follow-up of these offspring is essential to provide a complete and reliable prediction of the developmental course toward SZ and BD.
Several limitations should be considered.we adapted the staging model by Duffy 29 to compare the developmental trajectories between SZo and BDo.Importantly, other staging models suggest the perspective of cognitive development and the importance of biomarkers. 23,26,57The model by Duffy was chosen here, because of its focus on bipolar disorder and the psychotic features of the bipolar spectrum.Additionally, its developmental perspective includes symptoms that are not disorder-specific.It has been argued that variabilities in clinical outcome, prodrome and onset, and comorbid disorders complicate the predictive use of clinical staging models for SZ and BD. 58Additionally, clinical staging models that include pre-onset stages for individuals at risk are nonspecific, therefore caution is needed to prevent inappropriate (early) treatment that are not proven to prevent severe mental illnesses and may increase fear and stigma. 59However, this study used the clinical staging framework to further unravel the developmental course of BD and SZ in at-risk offspring, rather than to predict the clinical outcome of offspring at familial risk for mental illness.Finally, the majority of BDo had a parent with BD-I (82%), while genetic overlap with SZ is greater for BD-I than BD-II. 60Differences between SZo and BDo may have been more pronounced with a more heterogeneous BDo sample.
Demographic and clinical characteristics of participating families and offspring are summarized in Table 1.In total, 58 SZo (38 families), 94 BDo (60 families), and 56 Co (34 families), aged 8-18 years old participated at baseline.At 4-year follow-up (mean age = 17.1, SD = 2.7), the overall retention rate was 88.5%.Across groups, BDo were slightly older than SZo and Co. Mean IQ was significantly lower in the familial high-risk groups than Co.

2
The development of psychopathology and staging at baseline and follow-up.

F I G U R E 1
(A) Survival distributions across groups per domain of DSM-IV axis I psychopathology cluster.(B) Survival distributions across groups per stage.Cox regressions, corrected for age at interview and sex.Asterisks depict significant pairwise comparisons: *p < 0.05, **p < 0.01, and ***p < 0.001.psychopathology in offspring at familial high risk and suggests that problems (especially developmental disorders) occur at a younger age in SZo than in BDo, well before the onset of a first mood episode.As expected, between the mean age of 13 and 17 years, the majority of Co remained at the same stage, and SZo and BDo progressed one or two stages.

F I G U R E 2
Individual trajectories of clinical staging.Stage 0a = control offspring, no diagnosis; stage 0b= at risk, no diagnosis; stage 1 = non mood disorders; stage 2 = single mood disorders; stage 3 = recurrent mood disorders; stage 4 = manic/psychotic disorders; C = cyclothymic; H = hypomania; M = Mania; P = psychotic; each connecting line between dots represents a transition between stages, offspring without any diagnosis at both waves are represented by a single dot at their latest know ages (i.e., at follow-up or at baseline in case of drop-outs).TA B L E 3 Dimensional psychopathology reported at baseline and follow-up.Bold indicates statistically significant p-values Abbreviations: BDo, bipolar disorder offspring; SZo, schizophrenia offspring; Co, Control offspring; M, mean; SD, standard deviation.Note: Statistics are based on latest known information (baseline or follow-up) and corrected for multiple testing using false discovery rate-Benjamini-Hochberg procedure, with ⍺ = 0.034.To conclude, SZo and BDo show a comparable course of emerging psychopathology during childhood/adolescence.However, the earlier onset of developmental disorders and elevated developmental and mood symptoms in SZo as compared to BDo suggest a more severe early course in SZo compared to BDo that may reflect a difference in the at-risk phenotype.Having an affected parent has great impact on the development of offspring's mental well-being.Both offspring groups could benefit from close monitoring of clinical symptoms, to assess their clinical risk and provide early interventions to prevent illness transitions.

Table 3
presents the dimensional psychopathology at age 13.2 and 17.1 per group and domain using a multi-informant approach.Both SZo and BDo reported significantly more symptoms than co, independent of age, on the interview-based summary scales as well as the internalizing and externalizing scales of the ASEBA questionnaires as reported by all informants (offspring, parent, and teacher).BDo scored significantly higher on the mania scale than Co (t(205) = 3.43, p < 0.001).Compared to BDo, SZo reported higher levels on the depression (t(205) = 2.08, p = 0.039) and behavioral problems (t(205) = 2.13, p = 0.034) scales of the K-SADS-PL.Interestingly, self-report showed no significant differences between the familial high-risk groups on both internalizing or externalizing problem scales.Parent-report indicated that SZo experienced more internalizing symptoms than BDo (t(199) = 2.23, p = 0.027).Also, more externalizing problems were observed by parents (t(199) = 2.27, p = 0.025) and teachers (t(151) = 3.14, p = 0.0.002) in SZo than in BDo.
51,529 BD diagnoses and 6 SZ diagnoses were based on the Family Interview for Genetic Studies; There are four families with two second degree family members affected with schizophrenia, one family with two second degree family members affected with BD, and two families with first degree family members (sisters) affected with BD that were not included in this table; one family had two affected parent with BD, only the parent with the highest number of episodes was included in this table.Co: one co-parent with a chronic pain disorder; BDo: one co-parent with a gambling addiction and one co-parent with a adjustment disorder; comorbidities are not included in this table.d 6.9% of SZo and 5.3% of BDo used a combination of two types of medication.