Bipolar Disorder and Major Depression – Genetic Architecture and Approach in Medical Genetics

Literature and regulatory information update: 31 October 2026.

Genetic architecture and approach in medical genetics: heritability, genome-wide association studies, rare variants, syndromic forms, recurrence risk, counselling and pharmacogenetics

0. Initial Reference Table

1. Introduction

1.1. Two mood disorders with a genetic component

Bipolar disorder and major depressive disorder are high-frequency mood disorders in which medical genetics intervenes for reasons different from those that bring it to Mendelian diseases. There is no gene whose variant explains the disorder in most affected people, nor a genetic test that confirms or excludes the diagnosis. The genetic architecture is multifactorial: the sum of thousands of common variants of very small effect, a minor contribution of rare variants of moderate effect and interaction with environmental factors [16, 17, 24]. The role of the geneticist in this context includes estimating empirical risks for relatives, psychiatric genetic counselling, recognising the syndromic and chromosomal forms in which the mood disorder is part of a broader picture, excluding inborn errors of metabolism with psychiatric presentation and interpreting pharmacogenetic tests that already have formal recommendations [29–31, 40].

1.2. Definition of the entities in scope

Bipolar disorder type I (BD-I) is defined by the occurrence of at least one manic episode, usually accompanied by depressive episodes; bipolar disorder type II (BD-II) combines episodes of hypomania with major depressive episodes, without mania; the bipolar spectrum also includes subthreshold forms, in which hypomanic symptoms do not meet all criteria, and cyclothymia, in which neither the hypomanic nor the depressive symptoms reach the criteria for an episode [1, 26]. Schizoaffective disorder, bipolar type, occupies a borderline position with schizophrenia and, from a genetic standpoint, behaves as a distinct subgroup, with a greater burden of copy number variants [41]. Major depressive disorder (MDD) is defined by major depressive episodes in the absence of any manic or hypomanic episode; this definition by exclusion has practical consequences, because some people initially diagnosed with MDD later reveal BD, and MDD is the most frequent misdiagnosis in BD [1]. In Online Mendelian Inheritance in Man (OMIM), BD appears as Major affective disorder 1 (MAFD1, 125480), with historical synonyms such as manic-depressive psychosis, and MDD as Major depressive disorder (608516), always with the susceptibility notation, which indicates the absence of an established Mendelian causal gene [3].

1.3. From family studies to linkage studies

The foundation of the genetics of these disorders is epidemiological. Familial aggregation and twin concordance were demonstrated before any molecular identification: the Brazilian handbook Síndromes Genéticas: Reconhecer para Tratar summarises BD concordance as 40–60% in monozygotic twins and 4–8% in dizygotic twins [42]. The liability threshold model proposed by D. S. Falconer in 1965, presented in the Oxford Desk Reference, interprets common discontinuous diseases as the expression of a continuous liability, normally distributed in the population, with a threshold above which the disease becomes manifest; relatives of an affected person have a distribution shifted towards higher values and therefore a greater probability of exceeding the threshold [16]. From the 1980s onwards, linkage studies in multiplex families searched for genes of major effect. OMIM lists nine susceptibility loci for major affective disorder (MAFD1 to MAFD9, from 18p to 12p13.3) and two for MDD (MDD1 at 12q22-q23.2 and MDD2 at 15q25.3-q26.2), none of which led to a replicated causal gene [3]. OMIM cites a concordance of 57% in monozygotic twins and 14% in dizygotic twins (Cadoret, 1978) and records the sequence of these hypotheses [4]. The hypothesis of an X-linked form, suggested since the 1930s by the excess of affected women and the rarity of father-to-son transmission, gained strength with the linkage to colour blindness described by Reich in 1969 and with a LOD score of 3.10 with the factor IX locus at Xq27 (Mendlewicz, 1987), but later studies excluded the region in seven families and X-linked dominant inheritance was never formally demonstrated [5]. There followed proposals of genetic anticipation, with onset 8.9 to 13.5 years earlier in the second generation of 34 families, and of a parent-of-origin effect, with an excess of affected mothers, which led to searches for trinucleotide expansions and mitochondrial variants without confirmed results [4]. Linkage studies in isolated populations, such as the Old Order Amish and the population of the Central Valley of Costa Rica, descended from Spanish and Amerindian founders of the 16th and 17th centuries, produced signals at 18p, 18q and 4p that were not consistently replicated [4].

1.4. The Scottish family with the t(1;11) translocation and the DISC1 gene

The most influential finding of this phase was a large Scottish family identified in the Medical Research Council Cytogenetics Register in Edinburgh. In 1990, St Clair and colleagues examined 282 pedigrees with familial chromosomal abnormalities and described a family with 23 cases of mental or behavioural disorder, in which 34 of the 77 members studied carried a balanced t(1;11) translocation; there were psychiatric diagnoses in 16 of the 34 carriers and in 5 of the 43 non-carriers [43]. In 2001, Blackwood and colleagues described the breakpoints as (1;11)(q42;q14.3) and obtained a LOD score of 3.6 when the phenotype was restricted to schizophrenia, 4.5 when restricted to affective disorders and 7.1 when relatives with recurrent major depression, BD or schizophrenia were considered affected; the breakpoint on chromosome 1 directly disrupts the DISC1 and DISC2 genes [44]. The band designation differs between the two publications (q43;q21 in 1990 and q42;q14.3 in 2001). The result belongs to the translocation in this family and not to a point defect of DISC1: an imputation meta-analysis with 11,626 cases and 15,237 controls found no association between common variants at the locus and schizophrenia, and the review of 25 historical candidate genes concluded that this literature brought no clear knowledge about the genetic basis of schizophrenia, mainly owing to lack of statistical power [45, 46].

1.5. The candidate gene era

During the 2000s, case-control association studies of genes chosen on the basis of a pathophysiological hypothesis multiplied. The Brazilian handbook presents, based on Alves and colleagues (2012), five candidate genes with polymorphisms studied in BD: DISC1 (1q42), ANK3 (10q21), CACNA1C (12p13), TPH2 (12q21) and DAOA (13q33) [42]. CACNA1C was among the first associations replicated in genome-wide association studies (GWAS) and common polymorphisms of ANK3 maintained association in a case-control meta-analysis, both with small individual effects and without predictive value for the individual; the candidate gene literature, as a whole, did not withstand replication in large samples [46–48].

1.6. Genome-wide association studies and the Psychiatric Genomics Consortium

In 2009, the International Schizophrenia Consortium showed that a score built from thousands of common variants associated with schizophrenia was also associated with BD, the first demonstration of a shared polygenic basis [49]. The review by Craddock and Sklar, in 2013, summarised the initial phase: replicated associations at CACNA1C, ODZ4 and NCAN, polygenic contribution, overlap with schizophrenia and a minor role of structural variation [47]. The BD working group of the Psychiatric Genomics Consortium (PGC) published 30 loci in 2019, 64 loci in 2021 with 41,917 cases and 298 loci in 2025 with 158,036 cases from four major ancestry groups [17, 25, 50]. In depression, the 2019 meta-analysis with 246,363 cases identified 102 independent variants, and the PGC trans-ancestry meta-analysis published in 2025, with 688,808 cases from 29 countries, identified 697 associations at 635 loci [24, 51]. The joint analysis of eight psychiatric disorders, in 2019, showed that genetic influence transcends diagnostic boundaries, with 109 loci associated with at least two disorders [52].

1.7. Rare variants and sequencing

The second line of discovery is rare variation. In 2011, Malhotra and colleagues found, in 788 trios, a higher frequency of de novo copy number variants (CNV) in BD than in controls, especially with onset before the age of 18 [53]. Later CNV studies confirmed three loci shared with schizophrenia and showed that the CNV burden in BD is concentrated in schizoaffective disorder [41, 54]. Exome sequencing by the Bipolar Exome (BipEx) consortium, in 2022, identified the first definitive risk gene from rare variants, AKAP11, shared with schizophrenia [18]. A September 2026 preprint, with 64,435 cases, raises this number to 13 genes with exome-wide significance [55].

2. Molecular Genetics and Pathophysiology

2.1. Genetic architecture: a spectrum of causation

The Oxford Desk Reference describes the causation of human diseases as a continuum ranging from strictly genetic diseases, caused by fully penetrant variants, to predominantly environmental ones, with incompletely penetrant and multifactorial diseases in the intermediate zone [16]. BD and MDD lie in this intermediate zone. A clinical label such as "bipolar disorder" probably covers a heterogeneous set of conditions: in some people a rare variant of major effect or a structural chromosomal change predominates; in most, susceptibility results from the accumulation of many small effects [16]. The classical measure of familial aggregation is λ (lambda), the ratio between the incidence in relatives of an affected person and the incidence in the population, which works as a relative risk and an indirect indicator of heritability [16].

2.2. Heritability and familial aggregation

Scandinavian national registries have allowed population estimates free of the recruitment bias of multiplex families. In Sweden, linkage of the multigeneration and hospital discharge registries between 1973 and 2004 included 9,009,202 individuals and 40,487 probands with BD; the relative risk of BD was 7.9 in full siblings, 4.5 in maternal half-siblings and 2.4 in paternal half-siblings, and the estimated heritability was 59%, with a shared environment effect of 3.4% [19]. The same study showed an increased risk of schizophrenia in all degrees of relationship to probands with BD, including biological children adopted by other families, and attributed 63% of the comorbidity between the two disorders to common additive genetic effects [19]. A second Swedish analysis, with 54,723 patients among 8,141,033 offspring, estimated familial risks of 5.8 to 7.9 in first-degree relatives, decreasing with genetic distance, and a heritability of 58%; the genetic correlation with depression (0.62) was higher than that with schizophrenia (0.44), which places BD on both the mood and psychotic spectra, closer to the former [20]. In the Swedish Twin Registry, with 804 affected twins, heritability was 60.4% (95% CI 50.3–70.5%), with no difference between sexes despite 62.2% of those affected being women [21]. For psychosis in the broad sense (schizophrenia and BD), the Swedish STAR twin study estimated a heritability of 0.73 and showed that twins with a higher polygenic score were more likely to have an affected co-twin, similarly in monozygotic and dizygotic pairs [56].

In MDD, the meta-analysis by Sullivan and colleagues, with five family studies and five twin studies, estimated an odds ratio of 2.84 (95% CI 2.31–3.49) for the disorder in first-degree relatives, a heritability of 37% (31–42%), shared environment close to 0% and individual environment (including measurement error) of 63%; recurrence of episodes was the best predictor of familial aggregation [22]. In the Swedish Twin Registry, with 42,161 twins interviewed, heritability was 42% in women and 29% in men, with genetic factors only moderately correlated between sexes and stable across cohorts born between 1900 and 1958 [23].

2.3. Common variation: genome-wide association studies

GWAS compare millions of single nucleotide polymorphisms (SNP) between cases and controls. The Oxford Desk Reference stresses that factors identified by GWAS in complex diseases confer typical relative risks of 1.05 to 1.3 and that, even combined, they rarely allow clinically useful predictions; the difference between the heritability estimated in families and that explained by the markers constitutes the "missing heritability" [16]. In BD, the 2019 PGC study (20,352 cases and 31,358 controls in discovery) identified 30 loci and showed that 8 of the 19 variants significant in discovery ceased to be so in the combined analysis, as expected with small effects and heterogeneity [50]. The 2021 study, with 41,917 cases and 371,549 controls of European ancestry, identified 64 loci, estimated the heritability explained by SNP (h²SNP) at 18.6% on the liability scale (assuming a prevalence of 2%) and found enrichment in synaptic signalling genes, expressed in neurons of the prefrontal cortex and hippocampus, and in targets of antipsychotics, calcium channel blockers, antiepileptics and anaesthetics [17]. The h²SNP varied between recruitment waves, from 24.6% in the initial clinical cohorts to 11.9% in biobanks with cases identified by diagnostic codes [17].

The 2025 study, with 158,036 cases and 2.8 million controls of European, East Asian, African American and Latino ancestries, identified 298 loci, an association specific to the East Asian cohort and 36 credible genes by fine-mapping; these genes were enriched for ultra-rare damaging missense variants and truncating variants in cases, which shows convergence between common and rare variation [25]. The architecture differed according to the mode of recruitment: h²SNP was 0.22 in clinical samples, 0.05 in community samples and 0.08 in self-reported samples, and the self-reported samples correlated more with the community ones (genetic correlation, rg, 0.79) than with the clinical ones (0.47), which the authors attribute to a high proportion of BD-II in the self-reported samples [25].

In MDD, the 2019 meta-analysis of 807,553 individuals identified 102 independent variants, 87 of which were replicated in an independent sample of 1,306,354 individuals [51]. The 2025 PGC trans-ancestry meta-analysis brought together 688,808 cases and 4,364,225 controls from 29 countries, identified 697 associations at 635 loci (293 novel) and 308 high-confidence genes, with enrichment in the postsynaptic density, excitatory, inhibitory and medium spiny neurons, amygdala neurons and antidepressant targets; h²SNP in Europeans was 8.4% on the liability scale (assuming a lifetime risk of 15%), with an estimated polygenicity of 6% [24]. Effects were homogeneous across phenotype definitions at 569 of 570 loci, which supports combining clinical and self-reported diagnoses [24].

2.4. Subtypes and genetic correlations between disorders

BD-I and BD-II are genetically close but not identical: the genetic correlation was 0.85 in 2021 and 0.88 in 2025, both significantly below 1, and the h²SNP of BD-I is about twice that of BD-II [17, 25]. BD-I correlates more with schizophrenia (rg 0.66) than BD-II (0.54), and BD-II more with major depression (0.66) than BD-I (0.34); in the 2019 study, the correlation of BD-I with schizophrenia was attributed to psychosis [17, 50]. Overall, BD correlates with schizophrenia (rg 0.68), with major depression (0.44) and, to a lesser extent, with anorexia, attention-deficit/hyperactivity disorder (ADHD) and autism spectrum disorder (ASD) (about 0.2), as well as positive correlations with smoking and alcohol consumption [17]. Seventeen of the 64 loci of 2021 had already been implicated in schizophrenia and seven in depression [17]. The joint analysis of eight disorders identified three groups of interrelated disorders, 109 pleiotropic loci (23 with effects on four or more disorders and 11 with antagonistic effects) and increased brain expression of these genes from the second prenatal trimester onwards [52]. A pleiotropy meta-analysis across schizophrenia, BD and MDD identified 553 variants, with enrichment in cadherin, Wnt and group III metabotropic glutamate receptor pathways and in expression profiles of the second trimester and the first year of life [57].

2.5. Copy number variants

Large rare CNV play a smaller role in BD than in schizophrenia. In 788 trios, de novo CNV were more frequent in BD than in controls (OR 4.8; 95% CI 1.4–16.0), especially with onset before the age of 18 (OR 6.3; 1.7–22.6) [53]. In a British sample of 2591 patients with BD, compared with 6882 patients with schizophrenia and 8842 controls and combined with earlier data, there was evidence of a contribution from three schizophrenia-associated loci: 1q21.1 duplication (P = 0.022), 3q29 deletion (P = 0.03) and 16p11.2 duplication (P = 2.3 × 10⁻⁴, the only one that survives correction for multiple testing); very large CNV, especially deletions above 1 Mb, contribute less in BD than in schizophrenia [54]. In the international consortium, with 6353 cases and 8656 controls, CNV burden did not differ between BD and controls when BD was treated as a single entity, but it was increased in schizoaffective disorder, bipolar type, compared with controls, BD-I and BD-II; within BD-I, psychosis was associated with the schizophrenia polygenic score but not with CNV burden [41].

In depression, the UK Biobank made it possible to study 407,074 participants of European ancestry, after excluding ASD, intellectual disability (ID), ADHD, schizophrenia and BD. The set of 53 neurodevelopmental CNV was associated with self-reported depression (OR 1.34; 95% CI 1.19–1.49), an association attenuated but maintained after adjustment for physical health, education, social deprivation, smoking and alcohol (OR 1.26; 1.11–1.43). Three individual CNV survived Bonferroni correction: 1q21.1 duplication, duplication of the Prader-Willi region (15q11.2-q13) and 16p11.2 duplication. After excluding carriers of these CNV, the overall burden of rare CNV was not associated with depression [58]. The corresponding syndromes include the 16p11.2, 1q21.1, 3q29, 15q13.3 and 22q11.2 CNV and the 15q11-q13 region.

2.6. Rare single nucleotide variants

BipEx exome sequencing, with 13,933 patients and 14,422 controls, found an excess of ultra-rare protein-truncating variants (PTV) in genes under strong evolutionary constraint, in both main subtypes, and enrichment in genes implicated in schizophrenia by the SCHEMA consortium (24,248 cases) and in CHD8 targets; BD GWAS genes were not enriched for PTV. In combination with SCHEMA, AKAP11 emerged as a definitive risk gene (OR 7.06; P = 2.83 × 10⁻⁹); the AKAP-11 protein interacts with GSK3B, a proposed target of lithium [18]. The BipEx expansion released as a preprint in September 2026 (64,435 cases and 168,101 controls from 22 countries) identified 13 genes with exome-wide significance and 20 more with a false discovery rate below 5%, with odds ratios of 2.33 to 8.68, possibly inflated by the "winner's curse"; the strongest were RB1CC1 (OR 5.82), AKAP11 (OR 7.0) and ATP2B2 (OR 4.1), and several (RB1CC1, HERC1, CUL1, SP4, MAGI2, AKAP11, KDM5B) had already been implicated in schizophrenia or neurodevelopment [55]. These results have not yet been peer reviewed and do not, by themselves, justify including these genes in diagnostic panels.

In under-represented populations, genome sequencing of 1598 African Americans with BD and 3295 with schizophrenia, with 17,463 controls, produced only suggestive signals, which illustrates the lack of power outside European samples [59]. In Portugal, the Portuguese Island Collection, with multiplex families from the archipelagos, was studied by genome sequencing: in 41 families with BD (83 cases), 28 with schizophrenia (53 cases) and 62 controls, both affected and unaffected relatives had a greater burden of rare deleterious variants than controls, with no difference between affected and unaffected, a result compatible with a shared familial genetic liability; gene-set associations were only suggestive (preprint of May 2026) [60].

2.7. Monogenic and chromosomal forms and associations of major effect

There is no Mendelian form of BD or MDD with a gene identified and confirmed in several families. There are, however, situations in which a genetic change of major effect markedly increases the risk of mood disorder, and situations in which genes known for Mendelian diseases also appear in psychiatric association lists, which requires careful distinction of the level of evidence.

22q11.2 deletion syndrome is the most studied example of a chromosomal cause of psychiatric disease. GeneReviews (May 2025 revision) reports that about 60% of adults have a psychiatric disorder, with schizophrenia in about 25% and frequent anxiety and depressive disorders, and proposes psychiatric screening before the age of 10 [15]. The Oxford Desk Reference reports behavioural, psychological or psychiatric problems in about 93% of patients, psychiatric disorders including BD and schizophrenia in up to 18% of adults and at least one psychotic episode in 6.5%; the deletion accounts for less than 2% of schizophrenia cases in the population [61]. The estimates diverge according to the population studied and the case definition. Smith's Recognizable Patterns of Human Malformation describes BD or other mood disorders in 11q deletion (Jacobsen syndrome), 18q deletion, 17q21.31 microduplication (reciprocal of Koolen-de Vries syndrome) and Bardet-Biedl syndrome [70–73].

Darier disease, caused by variants in ATP2A2 (sarco/endoplasmic reticulum calcium pump), cosegregated with BD in some families; a Swedish population study confirmed a 4.3-fold higher risk of BD in patients and 1.6-fold in their first-degree relatives, and a 2.3-fold higher risk of schizophrenia in patients [62]. OMIM locates CACNA1C at 12p13.33 and associates it with Brugada syndrome type 3, Timothy syndrome and long QT [74]; the GeneGuidance record consulted wrongly gives 10q21.2, the cytoband of ANK3 [68].

2.8. Pathophysiology suggested by genetics

Genetics does not establish a single mechanism, but it points to convergent processes. In BD, loci are concentrated in genes for ion channels (including CACNB2 and KCNB1), neurotransmitter transporters and synaptic components, expressed in neurons of the prefrontal cortex and hippocampus, with involvement of GABAergic interneurons and medium spiny neurons; enrichment in targets of calcium channel blockers and antiepileptics and the interaction of AKAP-11 with GSK3B link genetics to pharmacological classes used in treatment [17, 18, 25]. The 2019 study also found enrichment in pathways regulating insulin secretion and endocannabinoid signalling [50]. In MDD, the associated genes are enriched in the postsynaptic density and receptor clustering, with involvement of amygdala neurons, and antidepressant targets are over-represented [24]. Expression of these genes in the prenatal period and in childhood suggests a neurodevelopmental component shared with other disorders [52, 55, 57].

2.9. Non-genetic factors and interaction

Heritability is not the complement of a negligible environmental effect. In MDD, individual environment explains about 63% of the variance in liability [22]; the 2023 Canadian guidelines list as non-modifiable risk factors a family history of mood disorder, childhood maltreatment and female sex, and as dynamic factors chronic medical illnesses, substance use disorders, insomnia, periods of hormonal change (puberty, pregnancy, postpartum, perimenopause) and stressful life events [2]. Advanced paternal age is associated with ASD, schizophrenia, BD and developmental delay, with an increase in small de novo chromosomal abnormalities and de novo variants of paternal origin [75].

2.10. Relevant pharmacogenetic genes

Genes with formal pharmacogenetic recommendations in psychiatry are not disease genes: their variants alter pharmacokinetics or the risk of hypersensitivity [29, 30]. CYP2D6 (OMIM 124030; 22q13.2, reference transcript NM_000106.6) has more than 155 haplotypes defined by the PharmVar consortium, with normal function (*1, *2), decreased function (*9, *10, *41) and no function (*3 to *6) alleles, as well as structural variation: gene deletion (*5), duplications and multiplications (xN) and rearrangements with the CYP2D7 pseudogene [30, 76]. CYP2C19 (OMIM 124020; 10q23.33, NM_000769.4) has more than 35 alleles, with no function (*2, *3), decreased function (*9) and increased function (*17); CYP2B6 (OMIM 123930; NM_000767.5) has more than 35 alleles, with decreased function (*6, *9), no function (*18) and increased function (*4) [30]. Star alleles are haplotypes, that is, combinations of variants; the most frequent defining variants, normalised in VariantValidator with the MANE Select transcripts, are shown in the following table [77].

The c.681G>A variant of CYP2C19*2 is synonymous at the level of the predicted protein; the loss of function described in the pharmacogenetic literature results from aberrant splicing, an aspect that the normalisation tool does not assess. For CYP2D6, the phenotype derives from the sum of the activity values of the two alleles (activity score): poor metaboliser with 0, intermediate from 0.25 to 1, normal from 1.25 to 2.25 and ultrarapid above 2.25 [30]. The HLA-B (OMIM 142830) and HLA-A (OMIM 142800) genes belong to the class I major histocompatibility complex, are inherited codominantly and have more than 4000 and 3000 alleles, respectively; the result is reported as positive (one or two copies of the allele) or negative, with no intermediate genotype, and the nomenclature is that of the World Health Organization, without HGVS [29, 79, 80]. The SLC6A4 (OMIM 182138; serotonin transporter; 5-HTTLPR polymorphism) and HTR2A (OMIM 182135; 5-HT2A receptor) genes have been studied extensively, but the evidence is mixed and insufficient for recommendations [30, 81, 82].

3. Clinical Spectrum and Manifestations by System

3.1. Bipolar disorder: poles of the spectrum

The DSM-5-based classification organises bipolar disorders according to the severity of mood elevation. In BD-I there are manic episodes with increased self-esteem, decreased need for sleep, pressured speech, racing thoughts, distractibility, psychomotor agitation and risk-taking behaviour, with significant functional impairment, possible psychotic symptoms and, at times, need for hospital admission. In BD-II, hypomanic episodes are qualitatively similar, but without the duration or severity that would cause marked impairment, hospitalisation or psychosis, and they coexist with major depressive episodes. At the other pole, cyclothymia corresponds to chronic hypomanic and depressive symptoms that do not meet episode criteria. DSM-5 also includes other specified and unspecified bipolar disorder, substance- or medication-induced bipolar disorder and bipolar disorder due to another medical condition [1]. The World Health Organization surveys (World Mental Health, WMH) showed that the severity of manic and depressive symptoms and suicidal behaviour increase monotonically from the subthreshold form to BD-I, whereas role impairment is similar between subtypes; depressive episodes were more severe than manic ones, with severe impairment reported by about 74.0% of people with depression and 50.9% of people with mania [26].

3.2. Age at onset and course

BD frequently becomes manifest in late adolescence and early adulthood, with a mean age at onset of 25 years. In BD-I, three onset subgroups have been described: early (mean 17.2 years, 41.7% of patients), intermediate (23.9 years, 24.7%) and late (32.2 years, 33.6%); the mean was 20 years in a United States sample and 29 years in a European sample, with a different distribution between subgroups. Early onset is associated with longer treatment delay, more severe depression and more anxiety and substance use [1]. Patients spend about half of their lives with syndromal or subsyndromal symptoms, mainly depressive, and are unable to maintain work function about 30% or more of the time; the Global Burden of Disease attributed to BD 9.9 million years lived with disability, the 16th cause worldwide, and the 6th cause of disability-adjusted life years between the ages of 10 and 24 [1]. About 90–95% of older adults with BD had their first episode before the age of 50; late onset is often related to neurological or physical comorbidity [1].

3.3. Major depressive disorder

In the WMH surveys of 18 countries, the retrospective median age at onset of a major depressive episode was 25.7 years in high-income countries and 24.0 in low- and middle-income countries, with a female:male ratio of about 2:1 and functional impairment associated with temporal proximity of the episode [27]. Anhedonia is reported by up to 70% of patients with MDD and is associated with a lower probability of remission with selective serotonin reuptake inhibitors (SSRI) [2]. Remission with the first antidepressant is around 30% and about half of adults with MDD have an inadequate response to initial pharmacotherapy [30, 83]. Among recurrence factors, the 2023 Canadian guidelines consider residual symptoms and childhood maltreatment to have robust evidence, and severity, chronicity, comorbidities, number of previous episodes, poor social support and persistent stressful events to have weaker evidence [2].

3.4. Comorbidity and mortality

Three quarters of people with bipolar spectrum disorder in the WMH surveys met criteria for at least one other disorder, mainly anxiety disorders (in particular panic attacks) [26]. In the Swedish registry, the risk of other psychiatric disorders was 9.7 to 22.9 times higher in patients with BD and 1.7 to 2.8 times higher in their siblings [20]. About 6–7% of patients identified with BD die by suicide. Suicide attempts are associated with female sex, early onset, depressive polarity of the first and current episode, anxiety, substance use, borderline personality disorder, a family history of suicide in a first-degree relative and previous attempts; death by suicide is associated with male sex and a family history of suicide in a first-degree relative. About 14% of suicides occur during a hospital admission and 26% in the six weeks after discharge [1].

3.5. Mood disorder in syndromic forms

In chromosomal and syndromic forms, mood disorder arises within a broader phenotype. In 22q11.2 deletion syndrome, behavioural changes may begin early (disinhibition and impulsivity or shyness and withdrawal, attention deficit, anxiety, perseveration and social difficulties), ASD occurs in about 20% of children and adult psychiatric disease includes anxiety, depression, BD and schizophrenia [15, 61]. In the case described in section 10, a 20q11.21 microduplication segregated in a family in which ASD and BD coexisted in several members [84]. Patients with Darier disease have an increased risk of BD and schizophrenia [62]. Inborn errors of metabolism may present for years with psychiatric symptoms alone, before any organic sign, and treatment is more effective at that stage than after irreversible neurological damage [40].

4. Clinical Recognition and Warning Signs

4.1. Suspecting bipolarity in a person with depression

Diagnoses of BD are often made up to 10 years after symptom onset, after one to four alternative diagnoses; the most frequent error is MDD, and schizophrenia or other psychoses appear as the initial diagnosis in up to 30% of patients [1]. The guidelines of the Canadian Network for Mood and Anxiety Treatments (CANMAT) with the International Society for Bipolar Disorders (ISBD) consider as suggestive of bipolarity hypersomnia or increased daytime napping, hyperphagia, leaden paralysis, psychomotor retardation, psychotic symptoms or pathological guilt, mood lability, first depression before the age of 25, five or more previous episodes, postpartum depression or puerperal psychosis, antidepressant-induced switching and a positive family history of BD [1]. The United States label for citalopram requires screening for a personal or family history of BD, mania or hypomania before starting any antidepressant [85].

Source: table 6 of the 2018 CANMAT/ISBD guidelines, adapted [1].

4.2. Children and adolescents of parents with bipolar disorder

About 28% of children and young people with MDD progress to BD; family history, early onset and psychotic symptoms are the most powerful predictors [1]. In the offspring of parents with BD, the Canadian guidelines recommend caution with antidepressants and stimulants, which may precipitate mania or hypomania, informing the family and close monitoring for switching [1]. NICE recommends not diagnosing BD in children or young people on the basis of depression with a family history of BD, but maintaining follow-up [32]. In the postpartum period, childbirth may trigger the first episode of hypomania or mania in women with MDD, so antidepressants are used with caution, especially with a family history of BD [1].

4.3. When to think of a specific genetic cause

The Genetic Testing Statement of the International Society of Psychiatric Genetics (ISPG) indicates that major adult psychiatric disorders are only rarely explained by pathogenic CNV or single-gene variants, but that these changes are more frequent when there is a history of altered neurodevelopment, such as ID in childhood [31]. Most of the committee considered that CNV analysis may be of value in atypical presentations or those associated with ID, ASD, learning disorders or certain medical syndromes [31]. Clinical warning signs are: dysmorphic features, conotruncal congenital heart disease, palatal anomalies, immunodeficiency or learning difficulties (22q11.2 deletion syndrome); ID, ASD or epilepsy in the person or relatives; familial keratosis follicularis (Darier disease); and families with many affected members over several generations, a situation in which the Oxford Desk Reference recommends considering an autosomal dominant disorder with incomplete penetrance [15, 16, 62].

4.4. Warning signs of an inborn error of metabolism

Saudubray organises inborn errors of metabolism with psychiatric presentation into three groups [40]. The first includes acute and recurrent attacks of confusion and behavioural change, usually with physical signs (gastrointestinal, headache, dysautonomia, pyramidal signs, altered consciousness): urea cycle disorders, homocysteine remethylation defects, porphyrias and mitochondrial diseases. In an unexplained acute psychiatric presentation, plasma ammonia, homocysteine and lactate should be measured. The second group includes isolated psychiatric signs in adolescence or adulthood in a previously asymptomatic patient: hyperhomocysteinaemias (remethylation and cystathionine β-synthase deficiency) and lipid disorders (metachromatic leukodystrophy, GM2 gangliosidosis, Niemann-Pick disease type C, adrenoleukodystrophy and cerebrotendinous xanthomatosis). In this group, catatonia, visual hallucinations, fluctuating symptoms, confusion, resistance to or worsening with treatment and cognitive decline are atypical clues. The third group includes mild ID since childhood with behavioural or personality disturbance without a defined psychiatric syndrome [40].

4.5. Pharmacogenetic warning signs

Severe or unusual adverse effects at low doses and lack of response to therapeutic doses are the situations in which the Canadian guidelines accept pharmacogenetic testing or plasma level monitoring [2]. Ancestry from South-East Asia, China (Han), Hong Kong, Thailand, Malaysia, the Philippines or India is a warning sign before carbamazepine or oxcarbazepine, owing to the frequency of HLA-B*15:02 [29, 86]. Any skin rash or mucosal ulcer during the first weeks of carbamazepine or lamotrigine requires urgent medical assessment [1].

4.6. Newborn screening

There is no newborn screening for BD or MDD.

5. Diagnosis: Criteria, Tests and Algorithms

5.1. The diagnosis is clinical

The diagnosis of BD and MDD is clinical, according to DSM-5 or International Classification of Diseases criteria, and no genetic test confirms or excludes it [1, 31]. The Canadian guidelines recommend strict adherence to criteria, a careful psychiatric history that includes first-degree relatives and collateral information; self-report questionnaires, such as the Mood Disorders Questionnaire, serve only to flag those who need more detailed assessment, because they have low sensitivity and specificity, especially in community samples or those with much comorbidity [1]. NICE includes family history in the assessment of suspected BD [32]. The baseline assessment before starting pharmacotherapy in BD includes full blood count, fasting glucose and lipid profile, platelets, electrolytes and calcium, liver enzymes, bilirubin, prothrombin time and activated partial thromboplastin time, urinalysis, urine toxicology, creatinine and estimated glomerular filtration rate, 24-hour creatinine clearance if there is renal disease, thyrotropin, electrocardiogram over the age of 40 or if indicated, pregnancy test and prolactin [1]. Before treating a manic episode, secondary causes should be excluded: drugs, medicines (stimulants, corticosteroids, levodopa, antidepressants) and medical or neurological diseases (head injury, brain tumours such as frontal meningiomas, multiple sclerosis, stroke, Cushing disease, hyperthyroidism) [1].

5.2. Genetic assessment: when and what

Genetic assessment has different objectives depending on the context. The following table summarises the tests, indications and limitations according to the sources consulted.

5.3. Practical algorithm in clinic

5.4. Polygenic scores: what they measure and why they are not used

In the 2021 GWAS, the BD PRS explained on average 4.57% of phenotypic variance on the liability scale in European cohorts, with a mean area under the curve (AUC) of 65%; people in the top decile had an odds ratio of 3.5 (95% CI 1.7–7.3) relative to the middle decile, and the variance explained fell to 2.3% and 1.9% in East Asian samples and to 1.2% and 0.4% in African American samples [17]. In 2025, the PRS explained 9.0% of variance in European cohorts; the top quintile had an OR of 7.06 relative to the middle quintile and the median AUC was 0.70, values that the authors considered insufficient for diagnostic prediction in the general population [25]. In depression, the score explained 5.8% of liability variance in Europeans, with an AUC of 0.625, and was not associated with depression in the largest study of African ancestry [24]. In the UK Biobank, different PRS for MDD identified different people as high risk (top 5%): only 9.5% to 47.0% of high-risk individuals with one score were so with another, and the specificity of all scores was between 50.7% and 56.4% [87]. The ISPG classifies PRS as research analyses, not recommended for identifying high-risk people or for aiding diagnosis, and stresses their strong dependence on ancestry [31]. Research continues: in a Pittsburgh cohort of 227 offspring of parents with BD followed for more than 12 years, 38 developed BD, and adding the PRS to a clinical risk calculator raised the AUC to 0.81 (loss of 0.07 without the PRS), especially in offspring who already had anxiety or depression and whose parent had early onset; this result has not been validated for clinical use [91]. Calibration methods that convert the score into a probability of disease, such as the PGC-MDD BPC, are under development [92].

5.5. Pharmacogenetic tests: interpretation and limitations

The CYP2D6, CYP2C19 or CYP2B6 report gives the diplotype (combination of the two star alleles, for example CYP2D6*1/*4) and the predicted phenotype. The CPIC assignment table includes, for example: CYP2D6 ultrarapid metaboliser with *1/*1xN; normal with *1/*1 or *1/*41; intermediate with *4/*41, *10/*10 or *1/*4; poor with *4/*4 or *5/*5; for CYP2C19, ultrarapid *17/*17, rapid *1/*17, normal *1/*1, intermediate *1/*2 or *2/*17, likely intermediate *1/*9, likely poor *2/*9 and poor *2/*2, *2/*3 or *3/*3; and indeterminate when there are alleles of uncertain function [30]. Technical limitations affect the result: CYP2D6 lies next to the homologous pseudogenes CYP2D7 and CYP2D8, has a deletion (*5), duplications of up to 13 tandem copies and hybrid alleles with CYP2D7, which require specific methods; a CYP2C19 panel without *17 cannot identify rapid or ultrarapid metabolisers; and the Association for Molecular Pathology recommendations define the minimum and extended alleles to be tested [93, 94]. CPIC adds that rare or novel untested variants (for example, the CYP2C:TG haplotype in CYP2C18) may make the real phenotype differ from the predicted one, that different laboratories test different sets of variants, that the phenotype also depends on age, diet, comorbidities, smoking, pregnancy and concomitant medication (phenoconversion), and that after liver or haematopoietic cell transplantation the genotyped tissue may not represent the liver enzymes [30]. None of these variants is a clinically useful predictor of common disease or a cause of Mendelian disease [30].

For HLA, the result is positive when there are one or two copies of the allele [29]. In the Han Chinese population, HLA-B*15:02 has a sensitivity of 98%, specificity of 97%, positive predictive value of 7.7% and negative predictive value of 100% for carbamazepine-induced Stevens-Johnson syndrome and toxic epidermal necrolysis (SJS/TEN); for oxcarbazepine, the positive predictive value is 0.73% [29]. Other alleles of the B75 serotype (HLA-B*15:08, *15:11, *15:21) are also associated with SJS/TEN, and a laboratory with high-resolution typing can identify them [29]. A negative result does not remove the need for clinical vigilance, and a positive result does not prove that a rash was caused by the drug [86, 95]. In one study, 4 of 12 patients considered European with SJS/TEN carried HLA-B*15:02 and had South-East Asian ancestry, which is why CPIC recommends considering the allele regardless of self-reported ethnicity [29].

5.6. Content of a genomic test report

The report of a microarray or sequencing test should state the gene or region, the reference genome version (for example GRCh37 or GRCh38), the coordinates or the variant in HGVS nomenclature with the transcript and version, the zygosity, the classification according to the American College of Medical Genetics and Genomics (ACMG) criteria and the segregation evidence; the case in section 10 illustrates a CNV described in GRCh37 coordinates and classified according to ACMG guidance and the European cytogenetics guidelines [84]. A CNV of uncertain significance that segregates with the disorder in a family does not, by itself, demonstrate causality, and a known risk CNV may be present in healthy relatives [31, 84]. Genome tests may generate secondary findings; the ISPG supports their disclosure when actionable, accepting that disclosure of non-urgent findings may be postponed during an active psychotic phase until stabilisation, and respecting the right not to know [31]. The ISPG does not recommend direct-to-consumer genetic tests for medical purposes in psychiatric patients, relatives or healthy people concerned about their risk [31].

5.7. Prenatal and preimplantation diagnosis

For polygenic BD and MDD there is no prenatal or preimplantation genetic diagnosis. These options arise only when a pathogenic CNV or another change of major effect has been identified in the family, in which case the counselling for the corresponding syndrome applies; the ISPG expressly excludes prenatal testing from the scope of its statement [31].

6. Differential Diagnosis

6.1. Psychiatric and medical differential

The CANMAT/ISBD guidelines list as the main differentials of BD MDD or persistent depressive disorder (manic or hypomanic episodes sought and absent), bipolar disorder due to another medical condition, substance- or medication-induced disorder, cyclothymia, psychotic disorders (schizophrenia, schizoaffective disorder, delusional disorder) and borderline, narcissistic and antisocial personality disorders, which may coexist with BD [1]. ADHD, substance use disorder and borderline personality disorder overlap with the symptoms of hypomania and mania, which explains overdiagnosis in some circumstances [1]. In MDD, undiagnosed BD is one of the first hypotheses to review when response to treatment is poor [2].

6.2. Genetic differential

The genetic differential looks for situations in which the mood disorder is the manifestation of a disease with an identifiable cause, with its own implications for treatment, surveillance and counselling. The table summarises the main ones, with the test that distinguishes them.

6.3. Differentials with the greatest practical impact

22q11.2 deletion syndrome deserves priority because it is frequent among chromosomal causes of psychiatric disease, has treatable manifestations in other systems and a 50% risk of transmission to the offspring of each affected person [15, 61]. Wilson disease is a treatable differential: GeneReviews includes depression, BD or bipolar spectrum, personality changes and psychosis among the psychiatric manifestations, which may arise in isolation, and the diagnosis rests on low serum caeruloplasmin and copper, increased urinary copper and identification of biallelic variants [99]. In Niemann-Pick disease type C, psychiatric manifestations or early dementia may dominate the initial presentation, and vertical supranuclear saccade palsy is a sign to be actively sought [100]. In chronic psychiatric symptoms with atypical signs, Saudubray recommends brain MRI, ophthalmological examination, abdominal ultrasound, electromyography and a set of plasma biomarkers (ammonia, lactate, homocysteine, copper and caeruloplasmin, cholestanol, very long-chain fatty acids, lyso-sphingomyelin and hexosaminidase A) [40]. In Huntington disease, subtle motor, cognitive or psychiatric symptoms arise in a prodromal phase that precedes the diagnosis of manifest disease [105].

The distinction between a common variant association and a Mendelian disease is also part of the differential. Polymorphisms in ANK3 or CACNA1C associated with BD should not be reported as diagnoses or as relevant personal risk; rare pathogenic variants in these same genes cause, respectively, a neurodevelopmental disorder and Timothy syndrome or long QT type 8, with completely different investigation and treatment [48, 68, 74]. Likewise, a DISC1 variant or CNV should not be interpreted as a cause of psychiatric disease [45, 64].

7. Treatments and Risk Management Strategies

7.1. Principles

Genetic knowledge has not yet changed the choice of treatment for BD or MDD on the basis of aetiology: there is no gene-directed therapy, and clinical guidelines organise treatment by phase of illness, efficacy, tolerability and patient preference [1, 2]. Genetics intervenes at three specific points: in pharmacogenetics, which has formal recommendations for antidepressants and for carbamazepine and oxcarbazepine; in identifying treatable or syndromic causes that require their own management; and in information about familial risk, which shapes the surveillance of offspring and pregnant women [1, 29, 30, 40]. The pharmacological summary that follows is intended to frame these decisions and does not replace psychiatric guidelines.

The 2018 CANMAT/ISBD guidelines recommend as first line in acute mania lithium, quetiapine, divalproex, asenapine, aripiprazole, paliperidone, risperidone and cariprazine, as monotherapy or in combination; in bipolar I depression, quetiapine, lurasidone with lithium or divalproex, lithium, lamotrigine, lurasidone or adjunctive lamotrigine; and in maintenance, lithium, quetiapine, divalproex, lamotrigine, asenapine and aripiprazole [1]. In people with depression and a high risk of BD, especially with a strong family history, they recommend monitoring for the emergence of manic or mixed symptoms and, based on clinical experience, considering the recommendations for bipolar depression rather than antidepressant monotherapy [1]. In MDD, the 2023 Canadian guidelines adopt a personalised approach with shared decision-making, including psychological, pharmacological, lifestyle, digital health and neuromodulation treatments, and recommend continuing the antidepressant for 6 to 12 months after remission and for 2 years or more when there are risk factors for recurrence [2].

7.2. Lithium

Lithium is the reference treatment for BD maintenance, prevents manic and depressive episodes and has an anti-suicidal effect [1]. Response is associated with an episodic course with remission, a mania-depression-euthymia pattern, low comorbidity and a family history of BD, especially lithium-responsive BD; in a study cited by the guidelines, patients with a responder relative had a 67% probability of responding, compared with 35% without a responder relative [1]. The lithium dose is determined by the serum concentration, and the sources differ on the target.

For lithium, mEq/L and mmol/L are equivalent. The Portuguese SmPC for modified-release lithium carbonate (Priadel® 400 mg) indicates titrations of 200 to 400 mg every 3 to 5 days from 200 to 400 mg, total daily doses of 800 to 1800 mg for serum levels of 0.8 to 1.0 mmol/L and, in prophylaxis, 600 to 1200 mg/day for 0.4 to 0.8 mmol/L, with lower doses in older people and in people weighing less than 50 kg; it is contraindicated in severe renal impairment and not recommended in children [107]. The reference for the therapeutic target is the guideline; the SmPC accepts a wider maintenance range (0.5–1.0 mmol/L) and an upper limit of 1.5 mmol/L, above the NICE maximum of 1.0 mmol/L [32, 107]. NICE adds not to start lithium in primary care outside shared-care arrangements, to avoid over-the-counter non-steroidal anti-inflammatory drugs, to watch for signs of neurotoxicity even at therapeutic levels and to withdraw gradually over at least 4 weeks, preferably up to 3 months [32].

The genetics of lithium response is a research field. The ConLiGen consortium identified, in a GWAS of 2563 patients, a locus on chromosome 21 with two long non-coding RNA genes (rs79663003, rs78015114, rs74795342 and rs75222709); in a prospective study of 73 patients on monotherapy for up to 2 years, carriers of the response-associated alleles had a lower relapse rate (hazard ratio 3.8; 95% CI 1.1–13.0) [108]. In a study of 2586 patients, a low schizophrenia polygenic score was associated with better lithium response (OR 3.46 in the first decile relative to the tenth), with signals in HLA and inflammatory cytokine pathways [109]. Pathway-specific scores explained up to 3.71% of the variance in response and multi-trait scores up to 9.11%, with an AUC of 68.2% [110, 111]. A family with lithium-responsive BD showed linkage to a chromosome 19q region without known GWAS loci [69]. None of these markers has been validated to guide prescribing.

7.3. Carbamazepine, oxcarbazepine and HLA

Carbamazepine is approved in the United States for epilepsy, trigeminal neuralgia and BD; the Equetro® formulation is indicated for acute manic or mixed episodes of bipolar I disorder, with a starting dose of 200 mg twice daily and increases of 200 mg/day; doses above 1600 mg/day have not been studied in mania and the label defines no other maximum dose for this indication [29, 112]. In Portugal, Tegretol® (Novartis) is indicated in acute mania and in the maintenance treatment of bipolar affective disorders, with a dose range of about 400 to 1600 mg/day, usually 400 to 600 mg/day in 2 to 3 doses, rapid increase in acute mania and small increases in maintenance [86]. The Canadian guidelines recommend monitoring serum carbamazepine every 6 to 12 months, mainly for toxicity and adherence, and sodium at least annually, and educating patients about rashes and SJS/TEN [1].

HLA-B*15:02 is strongly associated with SJS/TEN due to carbamazepine and oxcarbazepine, and HLA-A*31:01 with maculopapular exanthema, drug reaction with eosinophilia and systemic symptoms (DRESS) and SJS/TEN due to carbamazepine; in OMIM, these associations appear as susceptibility to severe cutaneous adverse reaction (608579) [29, 113]. SJS involves up to 10% of the body surface and TEN more than 30%, with mortality below 5% and above 30%, respectively; the latency is about 4 to 28 days with continuous intake, and almost all reactions arise in the first 3 months [29]. The United States label estimates 1 to 6 cases of SJS/TEN per 10,000 new users in countries with mainly Caucasian populations and a risk about 10 times higher in some Asian countries [88].

Source: CPIC, 2017 update [29]. For oxcarbazepine, CPIC recommends not using it in HLA-B*15:02 carriers without previous exposure (strong) and makes no recommendation for HLA-A*31:01 owing to lack of evidence [29]. If only one of the alleles has been tested, the other is assumed to be negative. The recommendations apply regardless of region of origin and to adults and children. The other aromatic anticonvulsants (eslicarbazepine, lamotrigine, phenytoin, fosphenytoin, phenobarbital) have weak evidence of association with HLA-B*15:02, but require caution as alternatives; a screening policy that replaces carbamazepine with phenytoin does not reduce the overall rate of SJS/TEN, and after a severe reaction to one aromatic anticonvulsant the others should be avoided [29].

Regulatory sources and guidelines converge on the essentials and diverge on the extent of screening. The United States label for Tegretol® and Equetro® requires testing for HLA-B*1502 before treatment in people with ancestry in genetically at-risk populations and not treating carriers unless the benefit clearly outweighs the risk [88, 112]. The CANMAT/ISBD guidelines recommend genotyping in Han Chinese and other Asian patients before starting carbamazepine [1]. The Portuguese SmPC for Tegretol® recommends testing in Han and Thai ethnicities whenever possible, accepts genotyping in other Asian populations, does not recommend testing in low-prevalence populations or in patients already on treatment and, for HLA-A*3101, considers the data insufficient to recommend systematic testing, but suggests considering it in at-risk populations (Japanese, Caucasians, Indigenous peoples of the Americas, Hispanics, southern India and Arab descent) [86]. The ISPG recommends testing HLA-A and HLA-B before carbamazepine and oxcarbazepine, in line with regulators and expert groups, and reports that HLA-B*15:02 testing is mandatory under national regulations in Taiwan and Hong Kong and free of charge in Thailand [31]. Prospective HLA-B*15:02 screening reduced the incidence of SJS/TEN in a Chinese population [29].

7.4. Antidepressants and pharmacogenetics: CPIC recommendations

The 2023 CPIC guideline covers SSRI (citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline), serotonin and noradrenaline reuptake inhibitors (SNRI: desvenlafaxine, duloxetine, levomilnacipran, milnacipran, venlafaxine) and the serotonergic modulators vilazodone and vortioxetine [30]. CYP2D6 extensively metabolises fluvoxamine, paroxetine, venlafaxine (to an active metabolite, O-desmethylvenlafaxine) and vortioxetine; CYP2C19 extensively metabolises citalopram and escitalopram and is the main pathway for sertraline, in which CYP2B6 also influences exposure [30]. The recommendations assume that the genetic result is available before prescribing and always take into account interactions, age and renal and hepatic function [30].

Source: tables 2 to 4 of CPIC 2023 [30]. In normal metabolisers, the recommendation is always to start with the usual dose (strong); in indeterminate phenotypes there is no recommendation. For fluoxetine and duloxetine there is no recommendation (level C), and desvenlafaxine, levomilnacipran, milnacipran and vilazodone do not depend to a relevant extent on these enzymes [30]. When CYP2C19 and CYP2B6 are both available, the combination of two non-normal phenotypes may justify a change even when each gene alone would not: for example, CYP2C19 poor metaboliser and CYP2B6 poor metaboliser — choose an antidepressant not mainly metabolised by these enzymes (optional) [30]. The SLC6A4 and HTR2A genes have no recommendation, because the evidence is mixed and insufficient for clinical validity and utility [30]. The recommendations for citalopram, escitalopram and sertraline are considered applicable to paediatrics, because all three enzymes reach adult levels in early childhood; for the other drugs, extrapolation to adolescents requires monitoring [30].

For tricyclics (depression dose), the 2016 CPIC guideline recommends, for CYP2D6 ultrarapid or poor metabolisers, avoiding the tricyclic or, if necessary, titrating to a higher dose (ultrarapid) or reducing by 50% (poor), always with therapeutic drug monitoring; for intermediate metabolisers, consider a 25% reduction. In CYP2C19 ultrarapid, rapid or poor metabolisers, avoid tertiary amines (amitriptyline, clomipramine, doxepin, imipramine, trimipramine) and prefer secondary amines (nortriptyline, desipramine), or reduce by 50% in poor metabolisers with monitoring [114]. At low doses for neuropathic pain (for example 25 mg/day in adults), no adjustment is recommended in poor or intermediate metabolisers, with monitoring for adverse effects [114].

7.5. Divergences between guidelines and labels

The Dutch Pharmacogenetics Working Group (DPWG) recommends, in CYP2C19 poor metabolisers, not exceeding 50% of the usual maximum dose of escitalopram and citalopram and 37.5% of that of sertraline; in intermediate metabolisers, 75% for escitalopram and about 70% (65–75%) for citalopram; in ultrarapid metabolisers, avoiding escitalopram, with no action for citalopram; and, for CYP2D6, avoiding paroxetine in ultrarapid metabolisers, with no action in poor and intermediate metabolisers [115]. The DPWG classifies pre-emptive genotyping for these drugs as "potentially beneficial", to be decided case by case [115]. Divergences from CPIC include paroxetine in poor metabolisers (50% reduction in CPIC, no action in DPWG), citalopram in ultrarapid metabolisers (alternative, strong, in CPIC; no action in DPWG) and the way the sertraline reduction is expressed (maintenance dose 50% lower in CPIC; ceiling of 37.5% of the maximum dose in DPWG) [30, 115].

Labels also differ. In the United States, Celexa® (citalopram) is started at 20 mg/day, up to 40 mg/day, and has a maximum dose of 20 mg/day in CYP2C19 poor metabolisers, in people over 60 years of age, in hepatic impairment and with CYP2C19 inhibitors, owing to the risk of QT interval prolongation [85]. In Portugal, the SmPC of a generic citalopram recommends in poor metabolisers a starting dose of 10 mg/day in the first two weeks and a maximum of 20 mg/day, with halving of the dose in older people over 65 years [116]. For vortioxetine, the United States label for Trintellix® (starting dose 10 mg, up to 20 mg) limits the dose to 10 mg/day in CYP2D6 poor metabolisers and halves the dose with strong CYP2D6 inhibitors [117]; the European SmPC for Brintellix® (first authorisation on 18 December 2013; 10 mg/day in adults under 65, between 5 and 20 mg; starting dose of 5 mg/day from the age of 65, with caution above 10 mg/day) reports plasma concentrations about twice as high in poor metabolisers, but sets no specific limit, suggesting only that a lower dose be considered if strong CYP3A4 or CYP2C9 inhibitors are co-administered [118]. The reference dose is that of the guideline (CPIC), and the labels serve as comparison.

7.6. Clinical utility of pharmacogenetic tests: trials

The two trials cannot be compared directly: they differ in population, test and outcome. PRIME Care showed a clear effect on prescribing and a small, non-persistent effect on remission; GUIDED was negative on the primary outcome, and the positive results were limited to secondary outcomes [83, 119]. The 2023 Canadian guidelines cite meta-analyses with absolute differences of 4.6 to 7.2 percentage points in response rate, trials at risk of bias owing to lack of blinding, and conclude that they do not recommend routine use of pharmacogenetic tests, because the benefit is modest and inconsistent and the result delays treatment; they accept testing in the presence of severe or unusual adverse effects at low doses or lack of response to therapeutic doses, and recommend interpreting existing results according to CPIC [2]. CPIC describes the results of proprietary combinatorial tests as positive but modest and outside its evaluation process, because they are opaque [30]. The ISPG estimates, from randomised double-blind trials, that about 1 in 12 patients tested benefits, considers the evidence for widespread use inconclusive and recommends integrating results when they already exist, with greater likely utility after inadequate response or an adverse reaction to an antidepressant or antipsychotic [31]. In a New Zealand series of 52 patients with adverse reactions to or inefficacy of antidepressants, CYP2D6 and CYP2C19 phenotypes were considered explanatory ("actionable") in 38% of drug-response pairs with CPIC A or B evidence (48% in adverse reactions and 21% in inefficacy) [120].

7.7. Pregnancy, lactation and the postpartum period

Decisions on psychotropic drugs in pregnancy take into account lithium embryopathy and fetal valproate syndrome. In a United States Medicaid study of 1,325,563 pregnancies, cardiac malformations occurred in 2.41% of those exposed to lithium in the first trimester and in 1.15% of those not exposed (adjusted relative risk 1.65; 95% CI 1.02–2.68), with a dose relationship (relative risk 3.22 above 900 mg/day) and right ventricular outflow tract obstruction, including Ebstein anomaly, in 0.60% compared with 0.18% [121]. The Canadian guidelines recommend preconception counselling for all women of childbearing age at least 3 months before pregnancy, fetal ultrasound if there is first-trimester lithium exposure, avoiding divalproex (neural tube defects up to 5%, other anomalies, neurodevelopmental delay and a mean loss of 9 IQ points) and folic acid 5 mg/day before conception and during pregnancy, with no proof that it protects against anticonvulsant-associated spina bifida [1]. The Oxford Desk Reference indicates that anticonvulsant treatment in pregnancy, for epilepsy or mood disorder, is associated with a 2- to 3-fold higher overall risk of malformation, with 10.7% for valproate and 16.8% for polytherapy in a meta-analysis of women with epilepsy, and that lamotrigine and levetiracetam appear to have a lower teratogenic effect [122]. For SSRI, there is no confirmed embryopathy; studies on cardiac malformations are discordant, with debate over paroxetine and right outflow tract defects and fluoxetine and ventricular septal defect [123]. In a Nordic cohort of 1,633,877 births, SSRI use was not associated, after adjustment, with stillbirth, neonatal or post-neonatal death [124]. NICE recommends discussing the risks and benefits of lithium and preconception planning with women of reproductive age [106].

7.8. Agents and circumstances to avoid

In people with BD or at high risk of BD, antidepressant monotherapy carries a risk of switching; in the offspring of parents with BD, antidepressants and stimulants require caution [1]. Non-steroidal anti-inflammatory drugs should be avoided in patients on lithium [32]. Carbamazepine and oxcarbazepine should not be started in HLA-B*15:02 carriers [29]. Paroxetine and fluoxetine are potent CYP2D6 inhibitors and may convert a normal or intermediate metaboliser into a functionally poor phenotype (phenoconversion), which also applies to other CYP2D6 substrates [30]. Valproate follows the warnings of the British regulatory agency (MHRA) in people under 55 years of age, women and girls and men, and should not be started in primary care for BD [32]. In 22q11.2 deletion syndrome, GeneReviews reports that caffeine may worsen anxiety [15].

7.9. Developments in the last 12 months and level of evidence

A Europe PMC search for randomised pharmacogenomic trials in depression published in 2025 and 2026 found no new results, which does not exclude their existence. None of the developments changes, as of the reference date, the clinical recommendations on genetic testing in BD or MDD.

8. Reproductive Options and Family Counselling

8.1. Principles of empirical risk

In multifactorial conditions, counselling is based on empirical risks, obtained from family studies and registries, and not on figures derived from mathematical models [16]. The Oxford Desk Reference sets out the general principles: the risk is highest in close relatives and falls rapidly with distance of relationship; it is greater when the proband has severe disease; it increases when there are two or more close relatives affected; and, when several relatives are affected, an autosomal dominant disorder with incomplete penetrance should be considered [16]. In Falconer's model, for a disease with a population incidence of 1/1000, the expected risk in first-degree relatives approaches the square root of the incidence (about 3%), falls to about 1% in the second degree and approaches the population risk in the third degree [16]. These values illustrate the model and do not replace the specific data for BD and MDD.

8.2. Empirical risks in bipolar disorder

The Danish data refer to parents with psychiatric admission and underestimate cases treated only as outpatients; registry values are average population risks, and the risk of a specific family depends on age at onset, severity, number of affected members and subtype [16, 28]. The transmitted risk is not limited to the parent's disorder: offspring of parents with BD have an increased risk of schizophrenia, and offspring of parents with schizophrenia of mood disorders [19, 89].

8.3. Empirical risks in major depressive disorder

In the meta-analysis by Sullivan and colleagues, the odds ratio of MDD in first-degree relatives of a patient was 2.84 (95% CI 2.31–3.49), and recurrence of episodes in the proband was the best predictor of familial aggregation [22]. Relatives of patients with MDD showed no significant increase in the risk of schizophrenia in the meta-analysis of high-risk studies (RR 1.71; 95% CI 0.19–15.16) [89]. The 2023 Canadian guidelines consider that a family history of mood disorder substantially increases the risk of MDD [2].

8.4. Psychiatric genetic counselling

Unlike ID and ASD, there are no clinical practice guidelines for genetic testing in common adult psychiatric disorders, and genetic counselling is not to be confused with genetic testing [90]. The first specialised psychiatric genetic counselling clinic for non-syndromic disorders opened in Vancouver in 2012; in the first year, 85.3% of clients who consented to the use of their data were women, 76% were self-referred and 65.3% had personal experience of mental illness, and the empowerment (Genetic Counseling Outcome Scale, GCOS) and illness self-management self-efficacy scales increased significantly one month after the consultation [126]. In a randomised pilot trial with 120 participants with schizophrenia, BD or schizoaffective disorder, genetic counselling and an educational leaflet increased knowledge, but only counselling improved the accuracy of risk perception; there was no difference in internalised stigma or perceived control [127]. In the records of the same clinic, empowerment increased with a large effect (d = 1.10) in 307 patients, with greater benefit in those with a low baseline score and when the indication was to understand the recurrence risk, and it increased regardless of the time elapsed since symptom onset or diagnosis [128, 129].

The ISPG recommends that genetic tests for diagnostic purposes and genomic tests include informed consent and counselling by a professional with competence in mental health and in genetic interpretation, with a medical genetics referral when a recognised genetic disease is identified or when there are reproductive implications [31]. Counselling includes explaining the multifactorial model, the empirical risks adjusted to the family, the limited significance of common variants and polygenic scores, and the surveillance of the offspring of people with BD (section 4.2) [1, 31].

8.5. When there is a genetic change of major effect

When a pathogenic CNV is identified, such as the 22q11.2 deletion, each child of a carrier has a 50% probability of inheriting it, and counselling follows that of the respective syndrome [15]. The CNV associated with BD and depression (1q21.1 and 16p11.2 duplications, 3q29 deletion, 15q11.2-q13 duplication) have incomplete penetrance and may be present in apparently healthy people, so cascade testing informs about risk but does not predict disease [31, 54, 58]. In the case in section 10, the family received counselling on the recurrence risk and the implications of a duplication of as yet unestablished significance [84].

8.6. Prenatal diagnosis, testing of minors and paternal age

There is no indication for prenatal or preimplantation diagnosis in polygenic BD or MDD; these options apply only to a known familial change of major effect, with the limitations of incomplete penetrance [31]. Polygenic scores should not be used to estimate the risk of children or embryos, because they have no clinical validity and depend on ancestry [31, 87]. The ISPG does not recommend direct-to-consumer genetic tests to assess the risk of psychiatric disorder [31]. Advanced paternal age is associated with a higher frequency of BD and other neurodevelopmental and psychiatric disorders, but, in the absence of a family history, no additional test is recommended for that reason [75]. Pregnancy planning in women with BD includes reviewing medication at least 3 months in advance and postpartum follow-up, a period in which childbirth may trigger a first episode of hypomania or mania [1].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Prevalence of bipolar disorder

The household surveys of the WMH initiative, with 61,392 adults from 11 countries assessed with the CIDI 3.0 structured diagnostic interview, estimated lifetime prevalences of 0.6% for BD-I, 0.4% for BD-II, 1.4% for subthreshold BD and 2.4% for the bipolar spectrum, and 12-month prevalences of 0.4%, 0.3%, 0.8% and 1.5% [26]. Variation between countries was wide: in the United States (National Comorbidity Survey Replication, 2002–2003), the bipolar spectrum reached 4.4% over the lifetime (BD-I 1.0%, BD-II 1.1%); in São Paulo, Brazil, 2.1% (BD-I 0.9%, BD-II 0.2%); in India, 0.1%; in New Zealand, 3.9% [26]. Fewer than half of the people with a bipolar spectrum disorder received mental health treatment, and only 25.2% in low-income countries had contact with mental health services [26]. In Canada, the national survey found lifetime prevalences of 0.87% for BD-I and 0.67% for BD-II [1]. The Brazilian textbook quotes a prevalence of 0.8%, similar to that of schizophrenia, without stating the population or the period [42]. These differences reflect the method, the sample and the setting, and none of these estimates is a validated national prevalence for Portugal; the sources consulted do not include Portuguese prevalence data for BD.

9.2. Prevalence of major depression

In the WMH surveys of 18 countries (89,037 adults), the lifetime prevalence of major depressive episode was 14.6% in high-income countries and 11.1% in low- and middle-income countries, with similar 12-month prevalences (5.5% and 5.9%) and a female:male ratio of about 2:1 [27]. Among European countries, lifetime prevalence ranged from 9.9% in Germany and Italy to 21.0% in France, with 17.9% in the Netherlands, 14.1% in Belgium and 10.6% in Spain; in the United States it was 19.2% (8.3% at 12 months) and in São Paulo 18.4% (10.4% at 12 months) [27]. In Canada, about 12% of people aged 12 years or over reported a lifetime episode of MDD and 4.7% an episode in the past year; in 2020 the World Health Organization estimated that more than 300 million people worldwide had MDD [2]. Data from Portugal are not included in the sources consulted.

9.3. Genetic diversity and equity

Most genetic discovery has been made in populations of European ancestry, and the performance of polygenic scores falls in other ancestries: in BD, the variance explained was about 2% in East Asian samples and lower in African American samples; in depression, the score was not associated with the disorder in the largest study of African ancestry [17, 24]. The 2025 GWAS identified an association specific to the East Asian cohort, and the extended BipEx sequencing found rare-variant signals consistent across ancestries [25, 55]. The ISPG stresses that any future clinical use of polygenic scores will require markers informative for people of diverse ancestry [31].

Founder and isolated populations are useful for identifying rare variants: the GPR156 p.(Glu533Asp) variant was identified in a large Old Order Mennonite family, and the Portuguese Island Collection, of multiplex families from the Portuguese archipelagos, showed an increased burden of rare deleterious variants in affected and unaffected individuals [60, 63]. In Scotland, the t(1;11) translocation in a single family founded the research on DISC1 [44].

9.4. Pharmacogenetic allele frequencies by population

The Portuguese SmPC draws attention to the difference between allele frequency and carrier frequency: since each person has two chromosomes, the percentage of patients carrying at least one copy is almost twice the allele frequency [86]. The frequency of HLA-A*31:01 in South American groups, including Brazil, is relevant to the prescription of carbamazepine in admixed populations, in which self-reported ethnicity is a poor indicator of ancestry; the CPIC gives the example of patients regarded as European who carried HLA-B*15:02 through South-East Asian ancestry [29, 86].

9.5. Special populations

In children and adolescents, the diagnosis of BD requires caution, depression with a family history of BD justifies follow-up without a premature diagnosis, and the CPIC pharmacogenetic recommendations for citalopram, escitalopram and sertraline apply to paediatrics [30, 32]. In pregnancy and the postpartum period, the recommendations of section 7.7 apply. In older people, lithium level targets are lower and the US label for citalopram limits the dose to 20 mg/day above 60 years [1, 85, 106]. In people with ID, ASD or other neurodevelopmental conditions, pathogenic CNVs are more frequent, which justifies genetic assessment [31].

9.6. European networks and patient associations

There is no European Reference Network (ERN) for common psychiatric disorders; chromosomal syndromes and rare diseases with mood disturbance fall within ERN ITHACA, dedicated to rare malformation syndromes and neurodevelopmental disorders [33]. In Portugal, ADEB — Associação de Apoio a Doentes Depressivos e Bipolares supports patients and families [34]; in Brazil, ABRATA, founded in 1999, supports people with mood disorders and their families [35]. In the United States, the Depression and Bipolar Support Alliance (DBSA) and, in the United Kingdom, Bipolar UK have similar roles [36, 37]; at European level, GAMIAN-Europe represents patients and EUFAMI the families of people with mental illness [38, 39]. The International Society for Bipolar Disorders co-edits the CANMAT/ISBD guidelines, and the CPIC guidelines are available on the ClinPGx platform [1, 130].

10. Illustrative Clinical Case

10.1. Family with bipolar disorder, autism spectrum disorder and 20q11.21 microduplication

Simoncini and colleagues described, in 2023, a family followed at the Psychiatric Clinic of the University of Pisa, Italy [84]. The proband (II-4), a 24-year-old woman, was assessed during a major depressive episode in the context of BD-I diagnosed with the SCID-5 structured interview. In childhood she had been diagnosed with specific language disorder, hearing loss, an abnormality of the vocal apparatus, ASD and apraxia, and she needed a support teacher. In adolescence she was bullied, with emotional lability, crying spells, anxiety and episodes of behavioural dyscontrol, which worsened at 19 years after her parents' divorce. She began psychiatric treatment at 22 years for mood swings, with benefit from oxcarbazepine up to 300 mg/day; in the following year, worsening depression, mood swings, dyscontrol and suicidal ideation led to her first admission, during which the diagnosis of BD-I with comorbid ASD was established. The scales confirmed significant autistic symptoms (AdAS Spectrum 100/160, cut-off below 70; Autism Spectrum Quotient 40/50, cut-off below 26; ADI-R and ADOS-2 positive) [84].

The family history showed 30-year-old monozygotic twin brothers with ASD and ID since childhood; a 33-year-old sister with BD-II and autistic traits, who only sought treatment at 30 years for depression with anxiety; a 55-year-old father with BD-II, a major depressive episode with a suicide attempt after the separation and autistic traits; and a 50-year-old mother with no mental disorder [84]. Array comparative genomic hybridisation (array-CGH, 60,000-probe platform) identified a duplication of about 350 kb at 20q11.21 (GRCh37, chr20:30,712,111–31,062,545) in the father, the sister, both twins and the proband, absent in the mother, never described in healthy controls and with no overlapping duplications in the DECIPHER and ClinVar databases. The region contains eight genes (TM9SF4, TSPY26P, PLAGL2, POFUT1, MIR1825, KIF3B, ASXL1 and NOL4L), and the authors highlighted TM9SF4, listed in the SFARI database of autism genes. The family received genetic counselling about the recurrence risk and the implications of the duplication [84]. The pedigree is shown in figure 1 (https://pmc.ncbi.nlm.nih.gov/articles/PMC10702349/figure/fig1/) and the array-CGH profile in figure 2 (https://pmc.ncbi.nlm.nih.gov/articles/PMC10702349/figure/fig2/) of the original article.

10.2. Reading the case in the light of the evidence

The case illustrates the ISPG recommendations: microarray assessment was performed in a family with BD and ASD or ID in several members, that is, in a presentation with a neurodevelopmental component, and not in an adult with isolated BD [31]. The duplication segregates with the phenotype in five relatives, but the authors themselves acknowledge that they cannot exclude other genetic changes undetected by the platform and that pathogenicity requires more patients and functional studies; the article does not state a final ACMG classification, and the CNV corresponds, from what is described, to a variant of uncertain significance with familial segregation [84]. 20q11.21 is not among the recurrent CNVs associated with BD in case-control studies (1q21.1 and 16p11.2 duplications, 3q29 deletion) [54]. The proband's treatment regimen, with five psychotropic drugs including low-dose lithium sulfate, is an individual decision and not a model; the guidelines recommend choosing according to the phase of illness and monitoring the lithium level [1, 32]. Oxcarbazepine was used with no mention of HLA genotyping; in a patient of European ancestry, the frequency of HLA-B*15:02 is below 1%, and the CPIC makes no recommendation for HLA-A*31:01 with this drug [29].

10.3. Complementary pharmacogenetic series

In the New Zealand UDRUGS series, two CYP2D6 intermediate metabolisers (B18 and B46) had venlafaxine inefficacy, attributed to reduced formation of the active metabolite O-desmethylvenlafaxine, and one CYP2C19 rapid metaboliser (B4) had nightmares and exhaustion with amitriptyline, attributed to increased conversion to nortriptyline [120]. Across the series, the CYP2D6 and CYP2C19 phenotypes were considered explanatory in 38% of drug-response pairs [120]. These interpretations are retrospective and do not demonstrate causality; the CPIC makes no recommendation for action with venlafaxine in CYP2D6 intermediate metabolisers [30].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

BD is a chronic, recurrent illness: patients spend about half of their lives with symptoms, mainly depressive, and 6–7% of identified patients die by suicide [1]. Severe impairment affects about three quarters of people during depressive episodes and half during manic ones [26]. In MDD, remission with the first antidepressant is around 30%, and residual symptoms and childhood maltreatment are the most robust recurrence factors [2, 83]. Up to 30% of patients with BD do not respond to lithium [109]. Genetic knowledge does not yet provide validated prognostic markers: polygenic scores and markers of lithium response remain research tools [31, 108, 111].

11.2. Surveillance table

The genotype result stays with the patient throughout life, which is why a typing error has lasting consequences; CPIC recommends repeating HLA typing when a South-East Asian patient with a negative result for HLA-B*15:02, and without another allele of the B75 serotype, develops carbamazepine-induced SJS/TEN [29]. The sources consulted do not describe specific protocols for transition from paediatric to adult services in these disorders.

12. Summary and Final Considerations

Bipolar disorder and major depressive disorder are mood disorders with heritabilities of about 60% and 37%, respectively, and a genetic architecture dominated by thousands of common variants of small effect. The genome-wide association studies of the Psychiatric Genomics Consortium identified 298 loci in bipolar disorder and 697 associations in depression, implicating synaptic genes, ion channels and specific neuronal populations, and showed that bipolar disorder type I lies closer to schizophrenia and type II closer to depression. Rare variation makes a minor contribution: recurrent CNV (1q21.1 and 16p11.2 duplications, 3q29 deletion) and truncating variants in constrained genes, such as AKAP11, with a recent expansion to 13 genes still in preprint form.

In medical genetics practice, no test diagnoses these disorders. Recurrence risks are empirical: about 4% bipolar disorder in the offspring of a parent admitted with the illness and about 25% when both parents are affected, against 0.5% in the Danish population. Psychiatric genetic counselling improves knowledge, risk perception and empowerment. Polygenic scores have no clinical utility, because of the small variance explained, the low specificity and the dependence on ancestry. Microarray is justified when the mood disorder arises with intellectual disability, autism spectrum disorder, dysmorphic features or other syndromic signs, and metabolic investigation when there are atypical signs or unexplained acute presentations, because diseases such as Wilson disease or Niemann-Pick disease type C are treatable.

Pharmacogenetics is the application with formal recommendations. HLA-B*15:02 should be tested before carbamazepine and oxcarbazepine in at-risk ancestries, and HLA-A*31:01 guides carbamazepine use according to CPIC and the Portuguese SmPC. The CYP2D6, CYP2C19 and CYP2B6 genotypes guide the choice and dose of paroxetine, fluvoxamine, venlafaxine, vortioxetine, citalopram, escitalopram, sertraline and tricyclics when they are available or in the face of unusual adverse effects and lack of response, although the Canadian guidelines do not recommend routine testing and trials show modest benefits. The divergences between CPIC, the DPWG and the United States and European labels should be known, and the guideline is the reference for the decision.

13. References

1. Yatham LN, Kennedy SH, Parikh SV, Schaffer A, Bond DJ, Frey BN, et al. Canadian Network for Mood and Anxiety Treatments (CANMAT) and International Society for Bipolar Disorders (ISBD) 2018 guidelines for the management of patients with bipolar disorder. Bipolar Disord. 2018;20(2):97-170. doi:10.1111/bdi.12609. PMID: 29536616
2. Lam RW, Kennedy SH, Adams C, Bahji A, Beaulieu S, Bhat V, et al. Canadian Network for Mood and Anxiety Treatments (CANMAT) 2023 Update on Clinical Guidelines for Management of Major Depressive Disorder in Adults: Réseau canadien pour les traitements de l'humeur et de l'anxiété (CANMAT) 2023 : Mise à jour des lignes directrices cliniques pour la prise en charge du trouble dépressif majeur chez les adultes. Can J Psychiatry. 2024;69(9):641-687. doi:10.1177/07067437241245384. PMID: 38711351
3. Online Mendelian Inheritance in Man (OMIM). Files morbidmap.txt and mimTitles.txt (local copy). Baltimore: Johns Hopkins University. https://omim.org/
4. Online Mendelian Inheritance in Man, OMIM®. MIM Number 125480: MAJOR AFFECTIVE DISORDER 1; MAFD1 [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/125480
5. Online Mendelian Inheritance in Man, OMIM®. MIM Number 309200: MAJOR AFFECTIVE DISORDER 2; MAFD2 [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/309200
6. Online Mendelian Inheritance in Man, OMIM®. MIM Number 609633: MAJOR AFFECTIVE DISORDER 3; MAFD3 [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/609633
7. Online Mendelian Inheritance in Man, OMIM®. MIM Number 611247: MAJOR AFFECTIVE DISORDER 4; MAFD4 [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/611247
8. Online Mendelian Inheritance in Man, OMIM®. MIM Number 611535: MAJOR AFFECTIVE DISORDER 5; MAFD5 [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/611535
9. Online Mendelian Inheritance in Man, OMIM®. MIM Number 611536: MAJOR AFFECTIVE DISORDER 6; MAFD6 [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/611536
10. Online Mendelian Inheritance in Man, OMIM®. MIM Number 612357: MAJOR AFFECTIVE DISORDER 8; MAFD8 [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/612357
11. Online Mendelian Inheritance in Man, OMIM®. MIM Number 612371: MAJOR AFFECTIVE DISORDER 7; MAFD7 [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/612371
12. Online Mendelian Inheritance in Man, OMIM®. MIM Number 612372: MAJOR AFFECTIVE DISORDER 9; MAFD9 [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/612372
13. Online Mendelian Inheritance in Man, OMIM®. MIM Number 608516: MAJOR DEPRESSIVE DISORDER; MDD [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/608516
14. Orphanet. ORPHAcodes API — search by approximate name "bipolar" and "depress" [Internet]. Paris: INSERM; 2026 [accessed 31 Oct 2026]. Available at: https://api.orphacode.org/
15. McDonald-McGinn DM, Hain HS, Emanuel BS, Zackai EH. 22q11.2 Deletion Syndrome. 1999 Sep 23 [updated 2020 Feb 27; revised 2025 May 08]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. NBK1523. https://www.ncbi.nlm.nih.gov/books/NBK1523/
16. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Complex inheritance. p. 14–15.
17. Mullins N, Forstner AJ, O'Connell KS, Coombes B, Coleman JRI, Qiao Z, et al. 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. doi:10.1038/s41588-021-00857-4. PMID: 34002096
18. Palmer DS, Howrigan DP, Chapman SB, Adolfsson R, Bass N, Blackwood D, et al. Exome sequencing in bipolar disorder identifies AKAP11 as a risk gene shared with schizophrenia. Nat Genet. 2022;54(5):541-547. doi:10.1038/s41588-022-01034-x. PMID: 35410376
19. Lichtenstein P, Yip BH, Björk C, Pawitan Y, Cannon TD, Sullivan PF, et al. Common genetic determinants of schizophrenia and bipolar disorder in Swedish families: a population-based study. Lancet. 2009;373(9659):234-239. doi:10.1016/s0140-6736(09)60072-6. PMID: 19150704
20. Song J, Bergen SE, Kuja-Halkola R, Larsson H, Landén M, Lichtenstein P. Bipolar disorder and its relation to major psychiatric disorders: a family-based study in the Swedish population. Bipolar Disord. 2015;17(2):184-193. doi:10.1111/bdi.12242. PMID: 25118125
21. Johansson V, Kuja-Halkola R, Cannon TD, Hultman CM, Hedman AM. A population-based heritability estimate of bipolar disorder - In a Swedish twin sample. Psychiatry Res. 2019;278:180-187. doi:10.1016/j.psychres.2019.06.010. PMID: 31207455
22. Sullivan PF, Neale MC, Kendler KS. Genetic epidemiology of major depression: review and meta-analysis. Am J Psychiatry. 2000;157(10):1552-1562. doi:10.1176/appi.ajp.157.10.1552. PMID: 11007705
23. Kendler KS, Gatz M, Gardner CO, Pedersen NL. A Swedish national twin study of lifetime major depression. Am J Psychiatry. 2006;163(1):109-114. doi:10.1176/appi.ajp.163.1.109. PMID: 16390897
24. Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium. Trans-ancestry genome-wide study of depression identifies 697 associations implicating cell types and pharmacotherapies. Cell. 2025;188(3):640-652.e9. doi:10.1016/j.cell.2024.12.002. PMID: 39814019
25. O'Connell KS, Koromina M, van der Veen T, Boltz T, David FS, Yang JMK, et al. Genomics yields biological and phenotypic insights into bipolar disorder. Nature. 2025;639(8056):968-975. doi:10.1038/s41586-024-08468-9. PMID: 39843750
26. Merikangas KR, Jin R, He JP, Kessler RC, Lee S, Sampson NA, et al. Prevalence and correlates of bipolar spectrum disorder in the world mental health survey initiative. Arch Gen Psychiatry. 2011;68(3):241-251. doi:10.1001/archgenpsychiatry.2011.12. PMID: 21383262
27. Bromet E, Andrade LH, Hwang I, Sampson NA, Alonso J, de Girolamo G, et al. Cross-national epidemiology of DSM-IV major depressive episode. BMC Med. 2011;9:90. doi:10.1186/1741-7015-9-90. PMID: 21791035
28. Gottesman II, Laursen TM, Bertelsen A, Mortensen PB. Severe mental disorders in offspring with 2 psychiatrically ill parents. Arch Gen Psychiatry. 2010;67(3):252-257. doi:10.1001/archgenpsychiatry.2010.1. PMID: 20194825
29. Phillips EJ, Sukasem C, Whirl-Carrillo M, Müller DJ, Dunnenberger HM, Chantratita W, et al. Clinical Pharmacogenetics Implementation Consortium Guideline for HLA Genotype and Use of Carbamazepine and Oxcarbazepine: 2017 Update. Clin Pharmacol Ther. 2018;103(4):574-581. doi:10.1002/cpt.1004. PMID: 29392710
30. Bousman CA, Stevenson JM, Ramsey LB, Sangkuhl K, Hicks JK, Strawn JR, et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) Guideline for CYP2D6, CYP2C19, CYP2B6, SLC6A4, and HTR2A Genotypes and Serotonin Reuptake Inhibitor Antidepressants. Clin Pharmacol Ther. 2023;114(1):51-68. doi:10.1002/cpt.2903. PMID: 37032427
31. International Society of Psychiatric Genetics. Genetic Testing Statement [Internet]. Brentwood (TN): ISPG; approved 2014, revised 2017, updated and approved 2019 Mar 11 [accessed 31 Oct 2026]. Available at: https://ispg.net/genetic-testing-statement/
32. National Institute for Health and Care Excellence. Bipolar disorder: assessment and management. Clinical guideline CG185 [Internet]. London: NICE; 2014 Sep 24 [last updated 2025 Sep 2; accessed 31 Oct 2026]. Available at: https://www.nice.org.uk/guidance/cg185
33. ERN-ITHACA — European Reference Network on Rare Malformation Syndromes, Intellectual and Other Neurodevelopmental Disorders [Internet]. 2026 [accessed 31 Oct 2026]. Available at: https://ern-ithaca.eu
34. ADEB — Associação de Apoio a Doentes Depressivos e Bipolares [Internet]. Lisbon [accessed 31 Oct 2026]. Available at: https://www.adeb.pt
35. ABRATA — Associação Brasileira de Familiares, Amigos e Portadores de Transtornos Afetivos [Internet]. São Paulo [accessed 31 Oct 2026]. Available at: https://www.abrata.org.br
36. Depression and Bipolar Support Alliance [Internet]. Chicago (IL) [accessed 31 Oct 2026]. Available at: https://www.dbsalliance.org
37. Bipolar UK [Internet]. London [accessed 31 Oct 2026]. Available at: https://www.bipolaruk.org
38. GAMIAN-Europe — Global Alliance of Mental Illness Advocacy Networks-Europe [Internet]. Brussels [accessed 31 Oct 2026]. Available at: https://www.gamian.eu
39. EUFAMI — European Federation of Families of People with Mental Illness [Internet]. Leuven [accessed 31 Oct 2026]. Available at: https://eufami.org
40. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 2, Inborn errors of metabolism in adults: a diagnostic approach to neurological and psychiatric presentations, section 2.3.7 Psychiatric disorders (Table 2.9).
41. Charney AW, Stahl EA, Green EK, Chen CY, Moran JL, Chambert K, et al. Contribution of Rare Copy Number Variants to Bipolar Disorder Risk Is Limited to Schizoaffective Cases. Biol Psychiatry. 2019;86(2):110-119. doi:10.1016/j.biopsych.2018.12.009. PMID: 30686506
42. Ribeiro EM, coord. Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. Section 7.7 Adult diseases (Psychiatric disorders: schizophrenia; bipolar disorder). p. 165–167.
43. St Clair D, Blackwood D, Muir W, Carothers A, Walker M, Spowart G, et al. Association within a family of a balanced autosomal translocation with major mental illness. Lancet. 1990;336(8706):13-16. doi:10.1016/0140-6736(90)91520-k. PMID: 1973210
44. Blackwood DH, Fordyce A, Walker MT, St Clair DM, Porteous DJ, Muir WJ. Schizophrenia and affective disorders--cosegregation with a translocation at chromosome 1q42 that directly disrupts brain-expressed genes: clinical and P300 findings in a family. Am J Hum Genet. 2001;69(2):428-433. doi:10.1086/321969. PMID: 11443544
45. Mathieson I, Munafò MR, Flint J. Meta-analysis indicates that common variants at the DISC1 locus are not associated with schizophrenia. Mol Psychiatry. 2012;17(6):634-641. doi:10.1038/mp.2011.41. PMID: 21483435
46. Farrell MS, Werge T, Sklar P, Owen MJ, Ophoff RA, O'Donovan MC, et al. Evaluating historical candidate genes for schizophrenia. Mol Psychiatry. 2015;20(5):555-562. doi:10.1038/mp.2015.16. PMID: 25754081
47. Craddock N, Sklar P. Genetics of bipolar disorder. Lancet. 2013;381(9878):1654-1662. doi:10.1016/s0140-6736(13)60855-7. PMID: 23663951
48. GeneGuidance. ANK3 gene record (clinical curation) [accessed 31 Oct 2026].
49. International Schizophrenia Consortium, Purcell SM, Wray NR, Stone JL, Visscher PM, O'Donovan MC, et al. Common polygenic variation contributes to risk of schizophrenia and bipolar disorder. Nature. 2009;460(7256):748-752. doi:10.1038/nature08185. PMID: 19571811
50. Stahl EA, Breen G, Forstner AJ, McQuillin A, Ripke S, Trubetskoy V, et al. Genome-wide association study identifies 30 loci associated with bipolar disorder. Nat Genet. 2019;51(5):793-803. doi:10.1038/s41588-019-0397-8. PMID: 31043756
51. Howard DM, Adams MJ, Clarke TK, Hafferty JD, Gibson J, Shirali M, et al. 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. doi:10.1038/s41593-018-0326-7. PMID: 30718901
52. Cross-Disorder Group of the Psychiatric Genomics Consortium. Genomic Relationships, Novel Loci, and Pleiotropic Mechanisms across Eight Psychiatric Disorders. Cell. 2019;179(7):1469-1482.e11. doi:10.1016/j.cell.2019.11.020. PMID: 31835028
53. Malhotra D, McCarthy S, Michaelson JJ, Vacic V, Burdick KE, Yoon S, et al. High frequencies of de novo CNVs in bipolar disorder and schizophrenia. Neuron. 2011;72(6):951-963. doi:10.1016/j.neuron.2011.11.007. PMID: 22196331
54. Green EK, Rees E, Walters JT, Smith KG, Forty L, Grozeva D, et al. Copy number variation in bipolar disorder. Mol Psychiatry. 2016;21(1):89-93. doi:10.1038/mp.2014.174. PMID: 25560756
55. Liao C, Ye R, Sealock JM, Boltz T, Ivankovic F, Finucane H, et al. Rare coding variation implicates thirteen genes in bipolar disorder across 232,536 individuals from global populations. medRxiv [preprint]. 2026. doi:10.64898/2026.09.30.26364416
56. Song J, Pasman JA, Johansson V, Kuja-Halkola R, Harder A, Karlsson R, et al. Polygenic Risk Scores and Twin Concordance for Schizophrenia and Bipolar Disorder. JAMA Psychiatry. 2024;81(12):1246-1252. doi:10.1001/jamapsychiatry.2024.2406. PMID: 39196586
57. Friligkou E, Pathak GA, Tylee DS, De Lillo A, Koller D, Cabrera-Mendoza B, et al. Characterizing pleiotropy among bipolar disorder, schizophrenia, and major depression: a genome-wide cross-disorder meta-analysis. Psychol Med. 2025;55:e145. doi:10.1017/s0033291725001217. PMID: 40357923
58. Kendall KM, Rees E, Bracher-Smith M, Legge S, Riglin L, Zammit S, et al. Association of Rare Copy Number Variants With Risk of Depression. JAMA Psychiatry. 2019;76(8):818-825. doi:10.1001/jamapsychiatry.2019.0566. PMID: 30994872
59. Li R, Gagliano Taliun SA, Liao K, Flickinger M, Sobell JL, Genovese G, et al. Whole-genome sequence-based association analysis of African American individuals with bipolar disorder and schizophrenia. HGG Adv. 2026;7(1):100499. doi:10.1016/j.xhgg.2025.100499. PMID: 40886051
60. Kazemi H, Drake J, Bigdeli T, Bacanu S, Nguyen TH, Benke K, et al. Gene-Set Based Rare Variant Association Analysis of Whole Genome Sequencing Data in the Portuguese Island Collection for Schizophrenia and Bipolar Disorder. medRxiv [preprint]. 2026. doi:10.64898/2026.05.28.26354351
61. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. 22q11 deletion syndrome. p. 624–627.
62. Cederlöf M, Bergen SE, Långström N, Larsson H, Boman M, Craddock N, et al. The association between Darier disease, bipolar disorder, and schizophrenia revisited: a population-based family study. Bipolar Disord. 2015;17(3):340-344. doi:10.1111/bdi.12257. PMID: 25213221
63. Miller BR, Gonzaga-Jauregui C, Brigatti KW, de Jong J, Breese RS, Ko SY, et al. A rare variant in GPR156 associated with depression in a Mennonite pedigree causes habenula hyperactivity and stress sensitivity in mice. Proc Natl Acad Sci U S A. 2025;122(16):e2404754122. doi:10.1073/pnas.2404754122. PMID: 40228124
64. GeneGuidance. DISC1 gene record (clinical curation) [accessed 31 Oct 2026].
65. Online Mendelian Inheritance in Man, OMIM®. MIM Number 605210: DISRUPTED IN SCHIZOPHRENIA 1; DISC1 [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/605210
66. Online Mendelian Inheritance in Man, OMIM®. MIM Number 181500: SCHIZOPHRENIA; SCZD [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/181500
67. Online Mendelian Inheritance in Man, OMIM®. MIM Number 600465: ANKYRIN 3; ANK3 [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/600465
68. GeneGuidance. CACNA1C gene record (clinical curation) [accessed 31 Oct 2026].
69. Wong AHC, Paterson AD, Roslin NM, Tetzlaff E, Sedrak AB, Zhou Y, et al. Genetic Linkage for Bipolar Disorder to the Distal Region of Chromosome 19q: A Large Family Whole Genome Sequencing Study. Bipolar Disord. 2026;28(4):e70112. doi:10.1111/bdi.70112. PMID: 42144802
70. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Deletion 11q syndrome (Jacobsen syndrome). p. 50–51.
71. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Deletion 18q syndrome. p. 58–59.
72. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. 17q21 microdeletion syndrome (Koolen-de Vries). p. 92–93.
73. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Bardet-Biedl syndrome. p. 826–827.
74. Online Mendelian Inheritance in Man, OMIM®. MIM Number 114205: CALCIUM CHANNEL, VOLTAGE-DEPENDENT, L TYPE, ALPHA-1C SUBUNIT; CACNA1C [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/114205
75. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Paternal age. p. 772–773.
76. Online Mendelian Inheritance in Man, OMIM®. MIM Number 124030: CYTOCHROME P450, SUBFAMILY IID, POLYPEPTIDE 6; CYP2D6 [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/124030
77. Freeman PJ, Hart RK, Gretton LJ, Brookes AJ, Dalgleish R. VariantValidator: accurate validation, mapping, and formatting of sequence variation descriptions. Hum Mutat. 2018;39(1):61-68. doi:10.1002/humu.23348. Tool at https://variantvalidator.org (accessed 31 Oct 2026)
78. ClinVar. VCV000016889 — CYP2D6*4 [Internet]. Bethesda (MD): National Center for Biotechnology Information [accessed 31 Oct 2026]. Available at: https://www.ncbi.nlm.nih.gov/clinvar/variation/16889/
79. Online Mendelian Inheritance in Man, OMIM®. MIM Number 142830: MAJOR HISTOCOMPATIBILITY COMPLEX, CLASS I, B; HLA-B [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/142830
80. Online Mendelian Inheritance in Man, OMIM®. MIM Number 142800: MAJOR HISTOCOMPATIBILITY COMPLEX, CLASS I, A; HLA-A [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/142800
81. Online Mendelian Inheritance in Man, OMIM®. MIM Number 182138: SOLUTE CARRIER FAMILY 6 (NEUROTRANSMITTER TRANSPORTER, SEROTONIN), MEMBER 4; SLC6A4 [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/182138
82. Online Mendelian Inheritance in Man, OMIM®. MIM Number 182135: 5-HYDROXYTRYPTAMINE RECEPTOR 2A; HTR2A [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/182135
83. Oslin DW, Lynch KG, Shih MC, Ingram EP, Wray LO, Chapman SR, et al. Effect of Pharmacogenomic Testing for Drug-Gene Interactions on Medication Selection and Remission of Symptoms in Major Depressive Disorder: The PRIME Care Randomized Clinical Trial. JAMA. 2022;328(2):151-161. doi:10.1001/jama.2022.9805. PMID: 35819423
84. Simoncini M, Violi M, Valetto A, Bertini V, Cruz-Sanabria F, Massoni L, et al. Case report: Familial case with autism spectrum and bipolar disorder showing a 20q11.21 microduplication including TM9SF4. Front Psychiatry. 2023;14:1240663. doi:10.3389/fpsyt.2023.1240663. PMID: 38076678
85. Allergan, Inc. CELEXA (citalopram) tablets — prescribing information. NDA 020822 [Internet]. Silver Spring (MD): U.S. Food and Drug Administration; revision in force 2023-10-09 [accessed 31 Oct 2026]. Available at: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=4259d9b1-de34-43a4-85a8-41dd214e9177
86. Novartis Farma — Produtos Farmacêuticos, S.A. Tegretol 200 mg e 400 mg comprimidos, Tegretol CR, Tegretol 20 mg/ml suspensão oral — Resumo das Características do Medicamento [Internet]. Lisbon: INFARMED; text revised 2026 Sep 9 [accessed 31 Oct 2026]. Available at: https://extranet.infarmed.pt/INFOMED-fo/download-ficheiro.xhtml?med_id=8335&tipo_doc=rcm
87. Mollon J, Schultz LM, Knowles EEM, Jacquemont S, Glahn DC, Almasy L. Low Stability and Specificity of Polygenic Risk Scores for Major Psychiatric Disorders Limit Their Clinical Utility. Biol Psychiatry. 2025;98(6):476-484. doi:10.1016/j.biopsych.2025.03.006. PMID: 40113122
88. Novartis Pharmaceuticals Corporation. TEGRETOL / TEGRETOL-XR (carbamazepine) — prescribing information. NDA 016608, 018927, 020234 [Internet]. Silver Spring (MD): U.S. Food and Drug Administration; revision in force 2026-08-20 [accessed 31 Oct 2026]. Available at: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=8d409411-aa9f-4f3a-a52c-fbcb0c3ec053
89. Rasic D, Hajek T, Alda M, Uher R. Risk of mental illness in offspring of parents with schizophrenia, bipolar disorder, and major depressive disorder: a meta-analysis of family high-risk studies. Schizophr Bull. 2014;40(1):28-38. doi:10.1093/schbul/sbt114. PMID: 23960245
90. Austin JC. Evidence-Based Genetic Counseling for Psychiatric Disorders: A Road Map. Cold Spring Harb Perspect Med. 2020;10(6):a036608. doi:10.1101/cshperspect.a036608. PMID: 31501264
91. Hafeman DM, Uher R, Merranko J, Zwicker A, Goldstein B, Goldstein TR, et al. Person-level contributions of bipolar polygenic risk score to the prediction of new-onset bipolar disorder in at-risk offspring. J Affect Disord. 2025;368:359-365. doi:10.1016/j.jad.2024.09.107. PMID: 39299598
92. Uffelmann E, Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium, Schizophrenia Working Group of the Psychiatric Genomics Consortium, Price AL, Posthuma D, Peyrot WJ. Estimating disorder probability based on polygenic prediction using the BPC approach. Nat Commun. 2025;16(1):8443. doi:10.1038/s41467-025-62929-x. PMID: 41006251
93. GeneGuidance. CYP2D6 gene record (clinical curation) [accessed 31 Oct 2026].
94. GeneGuidance. CYP2C19 gene record (clinical curation) [accessed 31 Oct 2026].
95. GeneGuidance. HLA-B gene record (clinical curation) [accessed 31 Oct 2026].
96. Mulle JG, Gambello MJ, Russo RS, Murphy MM, Burrell TL, Klaiman C, White S, Saulnier CA, Walker EF, Cubells JF, Shultz S, Li L. 3q29 Recurrent Deletion. 2016 Sep 22 [updated 2021 Jul 01; revised 2017 Oct 19]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. NBK385289. https://www.ncbi.nlm.nih.gov/books/NBK385289/
97. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Gómez–López-Hernández syndrome (cerebello-trigeminal dysplasia). p. 250–251.
98. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. Nail-patella syndrome (hereditary osteo-onychodysplasia). p. 612–615.
99. Weiss KH, Schilsky M. Wilson Disease. 1999 Oct 22 [updated 2023 Jan 12]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. NBK1512. https://www.ncbi.nlm.nih.gov/books/NBK1512/
100. Bremova-Ertl T, Patterson M. Niemann-Pick Disease Type C. 2000 Jan 26 [updated 2025 Nov 20; revised 2019 Aug 29]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. NBK1296. https://www.ncbi.nlm.nih.gov/books/NBK1296/
101. Federico A, Gallus GN. Cerebrotendinous Xanthomatosis. 2003 Jul 16 [updated 2022 Mar 17; revised 2024 Nov 14]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. NBK1409. https://www.ncbi.nlm.nih.gov/books/NBK1409/
102. Raymond GV, Moser AB, Fatemi A. X-Linked Adrenoleukodystrophy. 1999 Mar 26 [updated 2023 Apr 06; revised 2002 Aug 15]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. NBK1315. https://www.ncbi.nlm.nih.gov/books/NBK1315/
103. Sacharow SJ, Levy HL. Homocystinuria due to Cystathionine Beta-Synthase Deficiency. 2004 Jan 15 [updated 2025 Sep 25; revised 2005 Aug 15]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. NBK1524. https://www.ncbi.nlm.nih.gov/books/NBK1524/
104. Simpson KL, MacLeod EL, Kakajiwala A, Gropman AL, Mew NA. Urea Cycle Disorders Overview. 2003 Apr 29 [updated 2025 Jul 03; revised 2015 Apr 09]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. NBK1217. https://www.ncbi.nlm.nih.gov/books/NBK1217/
105. Brás IC, Dawson J, Kay C, Caron NS, Hayden MR. Huntington Disease. 1998 Oct 23 [updated 2026 Feb 12; revised 2025 Aug 30]. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. NBK1305. https://www.ncbi.nlm.nih.gov/books/NBK1305/
106. National Institute for Health and Care Excellence. Depression in adults: treatment and management. NICE guideline NG222 [Internet]. London: NICE; 2022 Jun 29 [accessed 31 Oct 2026]. Available at: https://www.nice.org.uk/guidance/ng222
107. Essential Pharma Limited. Priadel 400 mg comprimidos de libertação modificada (lithium carbonate) — Resumo das Características do Medicamento [Internet]. Lisbon: INFARMED; approved 2022 Nov 15 [accessed 31 Oct 2026]. Available at: https://extranet.infarmed.pt/INFOMED-fo/
108. Hou L, Heilbronner U, Degenhardt F, Adli M, Akiyama K, Akula N, et al. Genetic variants associated with response to lithium treatment in bipolar disorder: a genome-wide association study. Lancet. 2016;387(10023):1085-1093. doi:10.1016/s0140-6736(16)00143-4. PMID: 26806518
109. International Consortium on Lithium Genetics (ConLi+Gen), Amare AT, Schubert KO, Hou L, Clark SR, Papiol S, et al. Association of Polygenic Score for Schizophrenia and HLA Antigen and Inflammation Genes With Response to Lithium in Bipolar Affective Disorder: A Genome-Wide Association Study. JAMA Psychiatry. 2018;75(1):65-74. doi:10.1001/jamapsychiatry.2017.3433. PMID: 29121268
110. Sharew NT, Clark SR, Papiol S, Heilbronner U, Degenhardt F, Fullerton JM, et al. Pathway-Specific Polygenic Scores for Predicting Clinical Lithium Treatment Response in Patients With Bipolar Disorder. Biol Psychiatry Glob Open Sci. 2025;5(5):100558. doi:10.1016/j.bpsgos.2025.100558. PMID: 40821394
111. Sharew NT, Clark SR, Hartmann S, Papiol S, Heilbronner U, Degenhardt F, et al. Using multi-trait polygenic scores to predict lithium responsiveness in patients with bipolar disorder. Br J Psychiatry. 2026:1-12. doi:10.1192/bjp.2026.10776. PMID: 42841250
112. Validus Pharmaceuticals LLC. EQUETRO (carbamazepine) extended-release capsules — prescribing information. NDA 021710 [Internet]. Silver Spring (MD): U.S. Food and Drug Administration; revision in force 2025-08-19 [accessed 31 Oct 2026]. Available at: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=be478f3c-40f6-47cc-8ab9-f420a9372b1c
113. Online Mendelian Inheritance in Man, OMIM®. MIM Number 608579: SEVERE CUTANEOUS ADVERSE REACTION, SUSCEPTIBILITY TO [Internet]. Baltimore (MD): Johns Hopkins University [accessed 31 Oct 2026]. Available at: https://omim.org/entry/608579
114. Hicks JK, Sangkuhl K, Swen JJ, Ellingrod VL, Müller DJ, Shimoda K, et al. Clinical pharmacogenetics implementation consortium guideline (CPIC) for CYP2D6 and CYP2C19 genotypes and dosing of tricyclic antidepressants: 2016 update. Clin Pharmacol Ther. 2017;102(1):37-44. doi:10.1002/cpt.597. PMID: 27997040
115. Brouwer JMJL, Nijenhuis M, Soree B, Guchelaar HJ, Swen JJ, van Schaik RHN, et al. Dutch Pharmacogenetics Working Group (DPWG) guideline for the gene-drug interaction between CYP2C19 and CYP2D6 and SSRIs. Eur J Hum Genet. 2022;30(10):1114-1120. doi:10.1038/s41431-021-01004-7. PMID: 34782755
116. Generis Farmacêutica, S.A. Citalopram Aurovitas 10, 20 e 40 mg comprimidos revestidos por película — Resumo das Características do Medicamento [Internet]. Lisbon: INFARMED; approved 2024 Apr 26 [accessed 31 Oct 2026]. Available at: https://extranet.infarmed.pt/INFOMED-fo/download-ficheiro.xhtml?med_id=589660&tipo_doc=rcm
117. Takeda Pharmaceuticals America. TRINTELLIX (vortioxetine) tablets — prescribing information [Internet]. Silver Spring (MD): U.S. Food and Drug Administration; revision in force 2025-03-05 [accessed 31 Oct 2026]. Available at: https://dailymed.nlm.nih.gov/
118. H. Lundbeck A/S. Brintellix (vortioxetine) — Summary of Product Characteristics (EPAR product information) [Internet]. Amsterdam: European Medicines Agency; first authorisation 2013 Dec 18, renewal 2018 Nov 20 [accessed 31 Oct 2026]. Available at: https://www.ema.europa.eu/en/documents/product-information/brintellix-epar-product-information_en.pdf
119. Greden JF, Parikh SV, Rothschild AJ, Thase ME, Dunlop BW, DeBattista C, et al. Impact of pharmacogenomics on clinical outcomes in major depressive disorder in the GUIDED trial: A large, patient- and rater-blinded, randomized, controlled study. J Psychiatr Res. 2019;111:59-67. doi:10.1016/j.jpsychires.2019.01.003. PMID: 30677646
120. Kee PS, Maggo SDS, Kennedy MA, Chin PKL. The pharmacogenetics of CYP2D6 and CYP2C19 in a case series of antidepressant responses. Front Pharmacol. 2023;14:1080117. doi:10.3389/fphar.2023.1080117. PMID: 36895946
121. Patorno E, Huybrechts KF, Bateman BT, Cohen JM, Desai RJ, Mogun H, et al. Lithium Use in Pregnancy and the Risk of Cardiac Malformations. N Engl J Med. 2017;376(23):2245-2254. doi:10.1056/nejmoa1612222. PMID: 28591541
122. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Fetal anticonvulsant syndrome (FACS). p. 732–735.
123. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Drugs in pregnancy. p. 718–721.
124. Stephansson O, Kieler H, Haglund B, Artama M, Engeland A, Furu K, et al. Selective serotonin reuptake inhibitors during pregnancy and risk of stillbirth and infant mortality. JAMA. 2013;309(1):48-54. doi:10.1001/jama.2012.153812. PMID: 23280224
125. Streit F, McIntosh A, Lu Y. The Genetic Heterogeneity of Depression: A Perspective From the Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium. Biol Psychiatry. 2026;99(12):1068-1069. doi:10.1016/j.biopsych.2026.04.005. PMID: 42209045
126. Inglis A, Koehn D, McGillivray B, Stewart SE, Austin J. Evaluating a unique, specialist psychiatric genetic counseling clinic: uptake and impact. Clin Genet. 2015;87(3):218-224. doi:10.1111/cge.12415. PMID: 24773225
127. Hippman C, Ringrose A, Inglis A, Cheek J, Albert AY, Remick R, et al. A pilot randomized clinical trial evaluating the impact of genetic counseling for serious mental illnesses. J Clin Psychiatry. 2016;77(2):e190-8. doi:10.4088/jcp.14m09710. PMID: 26930535
128. Gerrard S, Inglis A, Morris E, Austin J. Relationships between patient- and session-related variables and outcomes of psychiatric genetic counseling. Eur J Hum Genet. 2020;28(7):907-914. doi:10.1038/s41431-020-0592-1. PMID: 32066934
129. Saxton S, Inglis A, Morris E, Hercher L, Austin J. Outcomes of psychiatric genetic counseling in relation to time since diagnosis and symptom onset. J Genet Couns. 2022;31(5):1148-1154. doi:10.1002/jgc4.1585. PMID: 35484926
130. Clinical Pharmacogenetics Implementation Consortium. CPIC guidelines (ClinPGx) [Internet]. 2026 [accessed 31 Oct 2026]. Available at: https://www.clinpgx.org/cpic/guidelines