Neuroanatomical dimensions in major depression: external validation and links with cognition, adverse life events, self-harm, metabolomics and genetics

Neuroanatomical dimensions in major depression: external validation and links with cognition, adverse life events, self-harm, metabolomics and genetics | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Neuroanatomical dimensions in major depression: external validation and links with cognition, adverse life events, self-harm, metabolomics and genetics Wenyi Xiao, Rachel D. Woodham, Yuhan Cui, Junaho Wen, Mathilde Antoniades, and 46 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5975731/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 15 Nov, 2025 Read the published version in Communications Medicine → Version 1 posted You are reading this latest preprint version Abstract Major depressive disorder (MDD) is a leading cause of disability worldwide, yet its diagnosis relies on clinical symptoms alone. Using machine learning applied to deeply phenotyped, medication-free participants with MDD, we identified two neuroanatomical dimensions. Dimension 2 (D2), compared to Dimension 1 (D1), was characterized by reductions in grey and white matter and was associated with limited treatment response to both antidepressant and placebo medications. Validation in UK Biobank general population cohort (n = 37,235) confirmed that D2 is characterized by reduced grey and white matter, alongside widespread cognitive impairments, adverse events in both adulthood and childhood, increased self-harm and suicide attempts, a pro-atherogenic lipid profile, and genetic associations with neurodegenerative traits. These findings suggest that D1 and D2 reflect distinct neurobiological mechanisms underlying MDD, with important implications for and treatment outcomes. External validation was demonstrated in a general population-based cohort that delineated mechanisms underlying heterogeneity in MDD, identifying potential biomarkers that could aid in personalising treatment approaches for this debilitating disorder. Health sciences/Medical research/Biomarkers/Diagnostic markers Health sciences/Medical research/Genetics research Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Major depressive disorder (MDD) is a mental health illness that is common, is a leading cause of disability worldwide, and the main precursor to suicide. 1 MDD is currently a clinical diagnosis, characterized by a persistent low mood and/or a diminished ability to experience usual feelings of pleasure, which is associated with a set of self-reported symptoms and clinical presentations, such as disrupted sleep, changes in appetite, low energy, feelings of guilt, and potential thoughts of death or suicide. 2 MDD is heterogeneous, clinical outcomes are variable, and there are no reliable predictors of treatment response at the individual level. 3 , 4 Approaches based on symptom profiles, comorbid disorders, or putative aetiology have yielded limited biological distinctions. MDD is considered a disorder at present, rather than a disease with a distinct neuropathological mechanisms. There are no neurobiological markers that can aid in the identification of MDD. Furthermore, it is likely that MDD is not a single disease but, instead, comprises multiple neurobiological mechanisms. 5 By applying advances in artificial intelligence to neuroimaging data, it has been possible to identify MDD from healthy participants at the individual level 6 , 7 and further to delineate subtypes within MDD with larger samples consisting of hundreds of participants. 8 However, regional structure and activity in MDD are altered by treatment, including medications, psychotherapy, and neuromodulation, and regional activity can vary with changes in depressive state, such as from a current depressive episode to a remitted state, in which there are few or no symptoms. 9 In treatment resistant depression, a clinical term referring to MDD characterized by continued symptoms despite at least two courses of treatment, the neural correlates almost certainly reflect the effects of multiple treatments in addition to disease-related effects. 10 We recently sought to identify neural patterns at the individual level that characterize first-episode and recurrent MDD in participants who were medication-free and in a current depressive episode. 5 Based on structural MRI data, we found two dimensions that showed distinct treatment responses to antidepressant medication. 11 Dimension 1 (D1) was characterized by preserved grey matter (GM) and white matter (WM) volumes and showed a positive clinical response to selective serotonin reuptake inhibitor (SSRI) antidepressant medication but not to placebo medication. In contrast, Dimension 2 (D2) was characterized by reduced GM and WM volumes and demonstrated limited clinical response to either SSRI or placebo medication. In the present study, we sought to externally validate these dimensions in the general population as well as within a subset of the general population with current depressive symptoms. We used the UK Biobank, a large-scale population dataset of over 500,000 participants with extensive neuroimaging, behavioural, and genetic data, 12 , 13 to test the generalizability of these dimensions and to explore their associations with cognitive functioning, affective expressions, and genetic profiles. In the external validation analyses, we examined associations previously reported with cognitive functioning, 14 , 15 depressive symptomatology, 16 – 18 anxiety-related symptoms, 18 neuroticism-related traits, 19 history of self-harm and suicide attempts, adverse life events and lifestyle behaviours, 20 metabolic measures, 14 , 21 and genome-wide association studies (GWAS). 8 We sought to integrate validation of the neurobiological dimensions in an independent cohort from the general population with multidimensional biopsychosocial factors. Results MRI volumetric measures In UKB general population (n = 37,235), external validation revealed the following classifications: D1 (n = 6,931), D2 (n = 10,262), combined D1 and D2 (n = 2,931), neither D1 or D2 (n = 12,009), margins of D1 or D2 (n = 7,102). The largest structural MRI volumes were observed in D1, followed by combined D1 and D2, neither D1 or D2 classification, and D2 (Table 1 , Figs. 1 – 2 , Supplementary materials). Table 1 Demographic information for UKB general population and individuals with current depressive symptoms for total sample and dimensions Number Female% Age General population Total sample 37,235 (100%) 53.0 64.14 ± 7.50 Dimension 1 6,931 (18.6%) 50.4 64.21 ± 7.35 Dimension 2 10,262 (27.6%) 51.9 63.95 ± 7.68 Combined D1 and D2 931 (2.5%) 44.8 64.42 ± 7.54 Neither D1 nor D2 12,009 (32.3%) 54.6 64.10 ± 7.52 Margin 7,102 (19.1%) 55.6 64.38 ± 7.51 Current depressive symptoms Total sample 1,454 (100%) 59.7 61.71 ± 7.56 Dimension 1 264 (18.2%) 58.0 61.84 ± 7.84 Dimension 2 442 (30.4%) 58.8 61.52 ± 7.56 Combined D1 and D2 30 (2.1%) 56.7 60.55 ± 7.44 Neither D1 nor D2 453 (31.2%) 61.6 61.79 ± 7.13 Margin 265 (18.2%) 60.0 61.90 ± 8.06 Number of participants in the sample is presented with percentages in parentheses indicating proportions relative to the total sample. The mean age of participants is presented in years with ± standard deviation. General population, all participants in the UK biobank with MRI data; current depressive symptoms indicate participants from general population sample who have depression and are currently depressed at the time of their MRI imaging visit. In UKB individuals in a current major depressive episode (MDE) (n = 1,454), external validation revealed the following classifications: D1 (n = 264), D2 (n = 442), combined D1 and D2 (n = 30), neither D1 or D2 (n = 453), margins of D1 or D2 (n = 265) (Table 1 , Fig. 1 , Supplementary materials). Cognitive functioning In UKB general population, D2 showed significantly impaired performance as compared to D1 in all measures of cognitive functioning: fluid intelligence ( β =−0.26, p = 5.73E-45), executive function, taking longer to complete TMT-A numeric path ( β = 0.16, p = 2.65E-13) and TMT-B alphanumeric path ( β = 0.12, p = 5.01E-08), working memory as measured by backward digit span task ( β =−0.18, p = 7.76E-17), processing speed as measured by symbol-digit substitution ( β =−0.14, p = 4.26E-11), nonverbal reasoning as measured by matrix pattern completion ( β =−0.25, p = 8.33E-33), and verbal memory as measured by paired associate learning ( β =−0.058, p = 0.012). In UKB MDE cohort, D2 similarly showed significantly impaired performance as compared to D1 in: fluid intelligence ( β =−0.25, p = 0.004), working memory as measured by backward digit span task ( β =−0.23, p = 0.027), processing speed as measured by symbol-digit substitution ( β =−0.19, p = 0.049), and nonverbal reasoning measured by matrix pattern completion test ( β =−0.23, p = 0.033). There were trends towards significant differences in TMT-A ( β = 0.21, p = 0.051) and TMT-B ( β = 0.20, p = 0.053) in which D2 showed impaired performance relative to D1. There was no significant difference between groups in paired associate learning task ( β =-1.56 p = 0.853). Depressive symptoms In UKB general population, in items related to core depressive symptoms, D2 participants reported higher rates of 'prolonged loss of interest in normal activities' (D1: 37.04%, D2: 40.07%; χ²(1) = 10.87, p = 0.00098, Cramér’s V = 0.030) and 'prolonged feelings of sadness or depression' (D1: 52.98%, D2: 55.53%; χ²(1) = 7.35, p = 0.0067, Cramér’s V = 0.025). In associated items, D2 individuals were more likely to report 'thoughts of death during the worst depression' (D1: 48.44%, D2: 51.56%; χ²(1) = 7.07, p = 0.0078, Cramér’s V = 0.0336) and ‘weight change during the worst episode of depression’ D1: 56.71%, D2: 60.90%; χ²(1) = 9.55, p = 0.0038, Cramér’s V = 0.025). In the PHQ-9 measures, D2 individuals reported greater prevalence of 'recent poor appetite or overeating' (D1: 16.46%, D2: 18.28%; χ²(1) = 6.43, p = 0.011, Cramér’s V = 0.023) There were no significant differences between dimensions in other symptoms, in 'feelings of worthlessness', 'difficulty concentrating', 'did your sleep change', 'feelings of tiredness during the worst period of depression', or in PHQ-9 item symptoms. In UKB MDE cohort, D2 participants reported higher endorsement of 'thoughts of death during the worst period of depression' (D1: 63.2%, D2: 77.2%; χ²(1) = 6.14, p = 0.013, Cramér’s V = 0.145), and within PHQ-9 items, D2 participants more frequently reported ‘recent changes in speed/amount of moving or speaking” (D1: 18.1%, D2: 27.8%; χ²(1) = 3.86, p = 0.049, Cramér’s V = 0.103). There were no significant differences in the remaining depressive symptom items. Anxiety-related symptoms In UKB general population, D2 individuals were more likely to report 'recent inability to stop or control worrying' (D1: 21.32%, D2: 23.93%; χ²(1) = 4.17, p = 0.041, Cramér’s V = 0.018). There were no significant differences in the remaining anxiety symptoms. In UKB MDE cohort, there were no significant differences between D1 and D2 in anxiety symptoms. Neuroticism-related symptoms In UKB general population, D2 individuals were more likely to report 'feeling tense/highly strung' (D1: 11.33%, D2: 12.74%; χ²(1) = 7.34, p = 0.0067, Cramér’s V = 0.021), 'nervous feelings' (D1: 16.54%, D2: 19.21%; χ²(1) = 19.00, p = 1.31E-05, Cramér’s V = 0.034), and 'suffering from nerves' (D1: 13.83%, D2: 16.06%; χ²(1) = 15.25, p = 9.42E-05, Cramér’s V = 0.030). There were no significant differences in any other items. In UKB MDE cohort, D2 participants more frequently reported ‘guilty feelings' (D1: 53.6%, D2: 64.9%; χ²(1) = 6.34, p = 0.012, Cramér’s V = 0.109). There were no significant differences in other items. Self-harm and suicide attempts In UKB general population, D2 individuals showed a significantly greater endorsement of a history of suicide attempts (55.20%) compared to D1 (43.23%) (χ²(1) = 7.68, p = 0.006, Cramér’s V = 0.113) as well as a history of self-harm (4.61%) compared to D1 (3.80%) (χ²(1) = 4.41, p = 0.036, Cramér’s V = 0.019) (Fig. 3 ). In UKB MDE cohort, D2 individuals similarly showed a significantly greater endorsement of a history of suicide attempts in (73.8%) as compared to D1 (45.8%) (χ²(1) = 4.03, p = 0.045, Cramér’s V = 0.247), but there were no significant differences in history of self-harm. Adverse life events and lifestyle behaviours In UKB general population, in items related to adverse life events, there was significantly greater endorsement in D2 compared to D1 in the following items: "experiencing physical violence by a partner or ex-partner" as an adult (D1 9.46%; D2 11.40%, χ²(1) = 12.36, p = 0.0004, Cramér’s V = 0.031), and "being stopped from seeing friends or family by a partner or ex-partner" as an adult (D1 7.38%, D2 8.37%; χ²(1) = 4.11, p = 0.043, Cramér’s V = 0.018); and in childhood events: "being physically abused by a family member" (D1 17.69%, D2 19.73%; χ²(1) = 8.39, p = 0.004, Cramér’s V = 0.025); and "feeling hated by a family member" (D1 15.64%, D2 18.09%; χ²(1) = 13.12, p = 0.0003, Cramér’s V = 0.032) (Fig. 3 ). In UKB MDE cohort, there was significantly greater endorsement in D2 as compared to D1 in the following items: "experiencing physical violence by a partner or ex-partner" as an adult (D1 13.4%; D2 25.7%, χ²(1) = 7.84, p = 0.005, Cramér’s V = 0.142), and "being stopped from seeing friends or family by a partner or ex-partner" as an adult (D1 13.6%, D2 23.0%; χ²(1) = 4.78, p = 0.029, Cramér’s V = 0.112), and "sexual intercourse by partner or ex-partner without consent" as an adult (D1 9.0%, D2 16.5%; χ²(1) = 3.92, p = 0.048, Cramér’s V = 0.101); and in childhood events: "being physically abused by a family member" (D1 21.7%, D2 38.5%; χ²(1) = 11.52, p = 0.0007, Cramér’s V = 0.172), and "feeling hated by a family member" (D1 32.1%, D2 44.3%; χ²(1) = 5.32, p = 0.021, Cramér’s V = 0.118) (Supplementary Fig. 2). In lifestyle behaviours, in UKB general population, there was greater endorsement or previous or current "smoking status" in D2 (38.92%) compared to D1 (37.09%; χ²(1) = 5.78, p = 0.016, Cramér’s V = 0.018) and in “salt added to food” in D2 (43.31%) compared to (D1 43.09%; χ²(1) = 12.97, p = 0.0003, Cramér’s V = 0.028), while there was greater endorsement of "alcohol usually taken with meals" in D1 (74.21%) compared to D2 (72.43%) (χ²(1) = 3.97, p = 0.046, Cramér’s V = 0.020). In UKB MDE cohort, there were no significant differences between D1 and D2. Metabolomics In UKB general population, 39 (57.4%) out of the 68 metabolites tested showed a significant difference between D1 and D2 (FDR p < 0.05). D1 showed significantly increased levels compared to D2 in 16 metabolites. These included 2 measures of intermediate density lipoprotein (IDL) particles, 9 measures of large to very large high-density lipoprotein (HDL) particles, forced expiratory volume and mean cell volume. Among fatty acids there was increased levels three fatty acids in their ratio to total fatty acids, including polyunsaturated fatty acids, omega-6 fatty acids and linoleic acid. D2 showed significantly increased levels compared to D1 in 23 metabolites. These include 13 large to extremely large, very low-density lipoprotein (VLDL) particles, systolic blood pressure, glycated haemoglobin, white blood cell count, C-reactive protein, and pyruvate. Among fatty acids there were significantly higher levels of total monounsaturated fatty acids (MUFA) and its ratio to total fatty acids (Fig. 4 .). In UKB MDE cohort, only forced expiratory volume (FEV1) was significantly higher in D1 compared to D2 (Supplementary Fig. 3). All measures are presented in Supplementary materials. Physical Measures In UKB general population, D2 showed significantly higher levels compared to D1 in: leg fat percentage (right) ( β = 0.052, p = 6.70E-08, η² = 0.052), leg fat percentage (left) (β = 0.046, p = 3.15E-07, η² = 0.046), body fat percentage ( β = 0.050, p = 0.000655, η² = 0.050), arm fat percentage (left) ( β = 0.059, p = 7.40E-05, η² = 0.059), and arm fat percentage (right) ( β = 0.065, p = 1.19E-05, η² = 0.065). In contrast, D1 showed significantly higher levels compared to D2 in: hand grip strength (right) ( β = −0.125, p = 2.01E-24, η² = −0.125), hand grip strength (left) ( β = −0.115, p = 3.19E-21, η² = −0.115), trunk fat-free mass ( β = −0.207, p < 1E-10, η² = −0.207), and trunk predicted mass ( β = −0.204, p < 1E-10, η² = −0.204) (Fig. 5 ). In UKB MDE cohort, D2 similarly showed significantly higher levels compared to D1 in: arm fat percentage (right) ( β = 0.161, p = 0.028, η² = 0.161), and arm fat percentage (left) ( β = 0.152, p = 0.037, η² = 0.152), while D1 exhibited significantly higher levels compared to D2 in: trunk fat-free mass ( β = −0.121, p = 0.028, η² = −0.121) and trunk predicted mass ( β = −0.120, p = 0.028, η² = −0.120) (Supplementary Fig. 4). Genetics GWAS analysis identified 7 genomic loci significantly associated with D1 and 7 genomic loci significantly associated with D2 (significance level P-value < 5x10 − 8 ). 3 genomic loci associated with D1 and 3 associated with D2 were new as their top lead SNP had not associated with any clinical traits in the National Human Genome Research Institute and European Bioinformatics Institute (NHGRI-EBI) GWAS Catalog. 22 GWAS analysis identified 10 independent significant SNPs associated with D1 and 9 associated with D2 (Fig. 6 ). Five independent SNPs associated with D1 had previous reports of positive associations in NHGRI-EBI GWAS Catalog, indicating that carrying the risk allele has been linked with increased depressed affect, 23 schizophrenia, 24 brain volume in dorsolateral prefrontal, posterolateral temporal and superior temporal regions, 25 right hippocampal subfield CA3 (head) volume, 26 body fat percentage, 27 and acne. 28 Three independent significant SNPs associated with D1 had previous reports of negative associations with brain age (i.e. difference between predicted and chronological age), 29 brain volume in orbitofrontal, superior parietal, and superior temporal regions, precuneus, and cortical surface regions, 25 as well as body mass index (Huang et al., 2023). Four independent significant SNPs associated with D1 showed no previous associations in GWAS Catalog. Two independent SNPs associated with D2 had shown positive associations in NHGRI-EBI GWAS Catalog with vertex-wise sulcal depth, 30 and whole-body fat free mass (UKB data field 23101), 31 and non-glioblastoma glioma. 32 SNP rs9823492 has been associated with paired helical filament tau (PHF-tau), which is a biomarker for Alzheimer’s disease, 33 and SNP rs4843547 has been associated with white matter microstructure measures. 34 Five independent significant SNPs associated with D2 showed no previous associations GWAS Catalog. MAGMA-based tissue expression analysis for D1 revealed two significant associations: in cerebellum (p = 0.00030772) and cerebellar hemisphere (p = 0.00034503). Research Domain Criteria (RDoC) associations In MDD participants from COORDINATE-MDD consortium, 11 individual item level baseline data were available in 426 participants from MADRS rating scale (CANBIND and Remedi, N = 130 (D1 n = 46, D2 n = 84)) or the HAMD rating scale (Oxford and EMBARC, N = 296 (D1 n = 148, D2 n = 148)). There were no significant differences between D1 and D2 in age or sex (D1: n = 194 (118 female), mean age 36.02 ± 12.87 years; D2: n = 232 (160 female), mean age 35.79 ± 12.57 years). In RDoC core depression domain, D1 showed significantly greater scores compared to D2 (D1 mean = 0.594, D2 mean = 0.556, p = 0.037), while in sensorimotor systems domain, D2 showed significantly greater scores compared to D1 (D1 mean = 0.233, D2 mean = 0.315, p = 0.005). Negative valance systems showed a trend towards significance with D2 exhibiting greater scores than D1 (D1 mean = 0.395, D2 mean = 0.423, p = 0.052) (Supplementary Materials). There were no significant differences in positive valance systems, cognitive systems, arousal/regulatory systems, anxiety phenotype or neurovegetative symptoms of melancholia phenotype. Sensitivity analyses Repeating the external validation in two UKB subgroups with overlapping ages from the COORDINATE-MDD sample: (1) having a lower age of 56 years and an upper age of 65 years (n = 6523), and (2) having an age range of ten years of the youngest UKB participant (ages 45 to 55 years) (n = 3112), D1 and D2 individuals were identified in both subgroups and showed a similar pattern of results in cognitive functioning, self-harm, adverse life events history, as well as metabolomics (Supplementary Tables 30 and 31). Discussion Neuroimaging-based dimensions derived from first episode and recurrent MDD showed robust external validation in a large UKB general population cohort. D1 was characterised by relatively preserved grey and white matter volumes, while D2 showed reduced grey and white matter volumes that were associated with a pattern of cognitive impairments, increased adverse life events in adulthood and childhood, increased self-harm and suicide attempts, a pro-atherogenic lipid profile, higher body fat and systemic inflammation compared to D1. These patterns were observed both in a large cohort from the general population and in individuals who had core depressive symptoms at the time of their brain scan in UK Biobank. A pattern of cognitive impairments was evident in D2 relative to D1 in a wide range of domains from executive function, working memory, and processing speed to nonverbal reasoning. Brain volume has been associated with general cognitive functioning in which cognitive decline is linked with reduced grey matter volume. 35 , 36 In the present analysis, we observed similar pattern of cognitive deficits in individuals who had endorsed current depressive symptoms at the time of their scan who were classified as D2 compared to D1. The findings suggest that the widespread subtle deficits in grey and white matter in D2 underlie the wide-ranging cognitive deficits observed in D2 compared to D1, which is observable in the general population, suggesting that dimension assignment is a trait feature as well as a state feature in individuals from the general population who had endorsed experiencing a current depressive episode. 11 In the general population, D2 had a significantly greater history of self-harm and suicide attempts than did D1. In addition, trauma-related experiences, such as childhood abuse and exposure to interpersonal violence, were more prevalent among D2 than among D1 individuals. Cumulative grey matter reductions have been consistently observed in trauma-exposed individuals, suggesting a potential contribution to the regional deficits in the development of MDD in D2 individuals. 37 With respect to metabolomics, D2 had a pro-atherogenic lipid profile, higher body fat, and impaired glucose metabolism, reflecting poorer metabolic health, while D1 had a healthier lipid metabolism. D2 was associated with higher levels of very low-density lipoproteins (VLDL), chylomicrons, triglycerides, and cholesterol as compared to D1. D2 also exhibited significantly higher body fat percentages, including leg and arm fat, as well as overall body fat. This was accompanied by reduced muscle mass, as evidenced by lower trunk fat-free mass and trunk predicted mass, and diminished hand grip strength on both sides as compared to D1. This profile is commonly associated with increased cardiovascular risk and metabolic dysregulation, potentially linked to metabolic syndrome, insulin resistance, and obesity. 38 High triglycerides have been associated with an increased risk of depression, anxiety, and stress-related disorders. 39 Furthermore, D2’s higher pyruvate levels, a key glycolysis metabolite, suggest disrupted glucose metabolism, consistent with observations that elevated glucose levels are associated with increased risk of psychiatric disorders 39 and Lv et al. 40 identified a relationship between increased body fat and higher risks of depression. Higher body fat percentages correlate with greater hormonal stress vulnerability (e.g., elevated cortisol levels), which in turn impairs cognitive performance under stress, 41 and is in itself associated with reduced regional brain volumes. 42 , 43 In contrast, D1 displayed a healthier lipid metabolism profile, with higher levels of high-density lipoproteins (HDL) and intermediate-density lipoproteins (IDL). Elevated HDL levels, particularly in large and very large HDL particles, are indicative of better cardiovascular protection and reduced systemic inflammation, which have been found to be protective against psychiatric disorders. 39 Consistent with the metabolic profile observed in D2, Lamers et al. 44 had identified an "immuno-metabolic depression" (IMD) profile which is characterized by increased inflammatory markers and metabolic dysregulation, such as dyslipidemia and higher body fat, and reduced physical fitness. The IMD profile was identified from the epidemiological population-based Netherlands Study of Depression and Anxiety (NESDA) and has been replicated in treatment-naïve adults with MDD. 45 Furthermore, longitudinal findings over six years observed a long-term persistence of high symptom burden in the IMD profile. 44 Our findings suggest that the limited effectiveness of SSRI antidepressant as well as placebo medications might be observed early in the course of illness in D2 dimension individual. 11 This highlights the need for additional treatment options to mitigate long term morbidity and improve clinical outcomes. GWAS analyses revealed distinct genomic loci and independent significant SNPs that were uniquely associated with either D1 or D2 scores. For D1, a positive association was found with SNP rs6782581 , which has a known negative association with BMI. 46 Grey matter volume is linked to BMI, with increased BMI (indicating being overweight or obesity) associated with reductions in grey matter, while normal BMI is linked with preserved grey matter. 42 , 43 However, rs6782581 has also been associated with a slight increase in body fat percentage (0.01%), 27 which appears contradictory as this is also associated with reduced grey matter volume. Indicators of general obesity though, such as BMI or body-fat percentage, are less informative at predicting brain volumes than indicators of central obesity, such as increased waist-to-hip ratio (WHR), with increases associated with reduced grey matter volume. 47 , 48 Supplementary analysis indicated significantly larger WHR in D2 compared to D1 in the UKBB general population. For D2, GWAS identified a positive association with SNP rs1926034 with previous studies linking this SNP to increased whole-body-fat free mass (all mass in the body that is not fat). 31 D2 is characterized by reduced grey matter volume compared to D1, 11 and Pflanz et al. 48 found an that higher whole-body fat-free mass was associated with lower grey matter volume and total brain volume. The findings suggest that variations in genetic associations and body composition metrics contribute to the distinct profiles of D1 and D2. D1 score demonstrated a significant negative association with SNP rs199505 , which has been linked to an increased likelihood of experiencing a depressed affect. 23 Within the RDoC framework, D1 was associated with significantly higher core depression scores and lower sensorimotor system scores compared to D2. This pattern highlights the persistence of core symptoms in the D1 profile, potentially linking its neurobiological features to distinct genetic and phenotypic domains. D2 score showed a negative association with rs12076373 , a variant linked to vertex-wide sulcal depth. 30 Increased sulcal depth reflects enhanced cortical folding, which has been positively associated with cognitive ability and cortical volume. 49 Conversely, reduced cortical folding has been observed in MDD. 50 Cortical folding is closely tied with genetic processes 30 and is predominantly completed before birth, 51 with sulcal patterns present at birth serving as predictors of neurobehavioral outcomes. 52 The negative association with D2 suggests a predisposition to reduced cortical folding, aligning with the reduced cortical volumes observed in D2 individuals. While genes direct brain development, environmental experiences shape the emerging structures and their relationships with each other. The greatest increase in brain weight occurs from birth to 3 years and there is a further five fold increase from 3 to 18 years. 53 However, childhood maltreatment has a pronounced impact on brain development, from grey matter deficits and cortical thinning in anterior cingulate/paracingulate and middle frontal regions 54 to widespread microstructural white matter reductions, particularly in fornix, corpus callosum and optic radiations as well as altered brain activity in the default mode and central executive network and increased responsivity in the amygdala and anterior cingulate to socioaffective cues. 55 , 56 Childhood maltreatment refers to physical, sexual or emotional abuse or neglect and is associated with deleterious effects on cognitive and emotional functioning and increased risk of mental health disorders. 57 Structural deficits and functional alterations linked with childhood maltreatment are a potential vulnerability which might then interacts with stressful life events in adulthood leading to the development of MDD. 58 Our findings suggest a potential pathophysiological mechanism for D2 MDD whereby polymorphisms associated with reduced cortical folding reflect a predisposition to reduced cortical volumes that is exacerbated by childhood maltreatment, leading to cognitive impairments and an immuno-metabolic form of MDD in adulthood which is associated with increased rates of self-harm as observed in D2 individuals in UKB general population,. Limitations of the present study include the binary coding of variables and reliance on self-reported experiences that may have introduced potential biases, including recall inaccuracies or social desirability effects. Additionally, in the demographic characteristics of the UK Biobank cohort, participants tend to have relatively high socioeconomic status and limited ethnic diversity, which does not fully reflect the general population. 59 The UK Biobank cohort was an older age group (mean 64 years) than our COORDINATE-MDD cohort (mean 38 years) which may have introduced a domain shift (Fu et al, 2024). However, sensitivity analysis stratified by age within the UK Biobank cohort showed comparable findings in subgroups with overlapping age ranges (Supplementary Materials). Although D1 and D2 dimensions were validated in the UK Biobank cohort, about 30% of UK Biobank participants had not been classified to either dimension, suggesting the possibility of additional dimensions. Furthermore, while functional connectivity measures have been associated with subtype classifications in MDD, the present analysis was limited to structural neuroimaging, which reduces our ability to fully capture the dynamic and network-level alterations in MDD. Integrating structural and functional measures has the potential to increase the precision in delineation of the biomarkers that comprise MDD. Two neuroanatomical dimensions were identified in medication-free individuals with first episode and recurrent MDD during a current depressive episode, in which D2 was characterized by widespread subtle deficits and showed limited treatment response to SSRI or placebo medication compared to D1. The present findings demonstrated strong generalizability in an independent general population cohort. D2 showed further associations with poorer cognitive functioning, greater exposure to adverse life events, including childhood trauma, an increased history of self-harm and suicide attempts, a pro-atherogenic lipid profile, and genetic links to white matter microstructure and neurodegenerative traits. The findings suggest that D1 and D2 represent biologically meaningful dimensions underlying MDD, associated with distinct neurobiological mechanisms and treatment responses. This highlights the shared and distinct neurobiological mechanisms of these dimensions that are currently categorised under the same clinical diagnosis. External validation demonstrates their potential as biomarkers to elucidate the heterogeneity within the current clinical MDD diagnosis that can aid in predicting treatment response and guiding treatment approaches. Methods We adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines, as detailed in the Supplementary Materials. Participants In the COORDINATE-MDD consortium, raw MRI data were shared from international samples (n = 1,384) in medication-free first episode and recurrent MDD (n = 685), all in a current depressive episode of at least moderate severity, and healthy controls (n = 699). Prospective longitudinal data on treatment response were available for a MDD subset (n = 359). Treatments were either SSRI antidepressant medication (escitalopram, citalopram, sertraline) or placebo (COORDINATE-MDD consortium). 5 , 11 The UK Biobank cohort represents a general population with recruitment taking place from 2006 to 2010 in 22 assessment centres in England, Wales and Scotland, with an age range of 40–69 years and total sample size of 500,000 participants. 60 All participants provided informed written consent, and ethical approval for the study was granted by the National Research Ethics Service Committee North West–Haydock (reference 11/NW/0382). Participants provided sociodemographic, cognitive, and medical information through questionnaires and physical assessments. A subset of participants completed magnetic resonance imaging (MRI), beginning in 2014 (UK Biobank Brain Imaging Documentation; http://www.ukbiobank.ac.uk ). Imaging was conducted on a 3T Siemens Skyra scanner with a T1-weighted MPRAGE protocol, resolution of 1 x 1 x 1 mm and a time to echo (TE) of 2000 ms. 13 UK Biobank MRI data in the present study were acquired from 2014 to 2019 (n = 37,235 (19,736 women (53.0%)); mean age 64.14 years (SD = 7.50)). Within the UKB general population cohort, a subset of individuals with current depressive symptoms indicating a major depressive episode (MDE) was identified using the following two UKB data fields: item 2050 ("frequency of depressed mood over the past two weeks") and item 2060 ("frequency of unenthusiasm or disinterest over the past two weeks"), with endorsement of "More than half the days" or "Nearly every day" for either criterion. Exclusion criteria included comorbid psychiatric disorders (e.g., schizophrenia, bipolar disorder, psychotic symptoms), neurological (e.g., Parkinson’s disease, epilepsy), and medical (e.g., diabetes, hypertension) disorders from ICD-10 diagnostic codes (UKB data field: 41202) (Supplementary Materials). Image preprocessing and harmonization All raw T1-weighted MRI data were first manually assessed for head motion, image artifacts or restricted field-of-view for quality assurance. Images were corrected for magnetic field inhomogeneity, and a multi-atlas segmentation (MUSE) was applied to obtain brain volumes in 145 anatomical regions of interest (ROI). 61 , 62 ROI volumes were linearly harmonized to correct for sex and deep learning-based intracranial volume (DLICV), an ICV estimation method that uses deep learning algorithms to harmonize variability in volumetric measurements across MRI scan, and non-linearly harmonized to correct for age-wise differences. 63 Harmonization was performed independently in the test dataset (UK Biobank) to prevent information leakage. For voxel-based analyses, T1-weighted images were transformed to the Regional Analysis of Volumes Examined in Normalized Space (RAVENS) maps for grey and white matter, which preserves total tissue volumes following normalization to ensure that regional differences are retained following registration and warping, enhancing their precision and reliability in assessing regional volumes. 64 , 65 Generalization with external validation in UK Biobank The semi-supervised Heterogeneity through Discriminative Analysis (HYDRA) model had been implemented to identify the optimal k- dimensional space for delineating the neuroanatomical heterogeneity in individuals with first episode and recurrent MDD, all of whom were unmedicated and experiencing a current depressive episode. 5 , 11 Determination of the optimal number of clusters ( k ) was based on the highest Adjusted Rand Index (ARI), a metric used to evaluate clustering solutions by quantifying the agreement between partitions while remaining relatively insensitive to the variations in the number of clusters. Based on ARI, k = 2 was computed as the optimal number of clusters, identifying two neuroanatomical dimensions which captured the heterogeneity in MDD neuroanatomy. The final HYDRA model was then applied to compute the MDD dimension scores in an independent external cohort derived from UKB. This angle ranges from 0° to 180°. As the dimensionality of the two axes increases, there is a greater likelihood for the angle to fall closer to 90°. 145 brain ROIs were used as training features in the HYDRA models, making each dimension axis 145-dimensional. For external validation, the HYDRA model was applied to the UK Biobank general population. Dimension membership (D1, D2) and expression scores (E1, E2) of the k = 2 dimensions for each participant were derived: 1) Dimension 1 (D1) was designated from E1 > = 0.3 and E2 = 0.3 and E1 = 0.3 and E2 > = 0.3; and 4) neither D1 or D2 was E1 <= -0.3 and E2 <= -0.3 (Fig. 1 ). To evaluate differences in brain volumetric measures in dimension membership, a post-hoc analysis of MUSE features was conducted using one-way ANOVA for groups: 1) D1; 2) D2; 3) combined D1 and D2; and 4) neither D1 or D2, with False Discovery Rate (FDR) correction to account for multiple comparisons with significance set at an FDR-adjusted threshold of p < 0.05. For each MUSE feature, the group means, and proportion of variance explained by the group differences was calculated. Associations with cognitive functioning, depressive and anxiety symptoms, neuroticism-related traits, adverse life events, self-harm and suicide, lifestyle items and metabolomics were assessed for the dimensions (D1, D2). For each item, variables were coded as binary (1 = "Yes," 0 = "No"), with missing data or responses excluded from analysis. Full descriptions of the variables and their classifications are presented in Supplementary Materials. Chi-squared tests were applied to identify significant differences in response patterns between groups. Standardised residuals and effect sizes (Cramér’s V) were computed to specify response patterns driving significant differences. A Cramér’s V of less than 0.2 indicates a weak effect, between 0.2 and 0.6 indicates a moderate effect, and greater than 0.6 indicates a strong effect. Cognitive functioning Cognitive functioning was assessed in the following seven domains: 14 , 66 executive function (trail-making test (TMT) A and B); fluid intelligence (verbal and numerical reasoning); working memory (backward digit span); verbal memory (paired associate learning); complex processing speed (symbol-digit substitution test) and nonverbal reasoning (matrix pattern completion). Poorer cognitive performance is indicated by higher scores in TMT, indicating increased time for completion, and lower scores in all other tests (Full descriptions are presented in Supplementary Materials). Structured analysis was applied to examine differences in cognitive functioning between D1 and D2. For each cognitive test, a general linear model adjusted for age, sex, and dimension membership (D1, D2) was assessed. False Discovery Rate (FDR) correction was applied, setting significance at an FDR-adjusted threshold of p < 0.05. Outliers were addressed by calculating z-scores based on middle 80% of data values, thereby establishing robust mean and standard deviation estimates. Extreme outliers having z-score exceeding 5 were excluded, and z-scores were recalculated for accuracy. All dependent variables were standardized. The resulting beta coefficients represent the standardized strength of associations. Positive beta coefficients indicate higher scores for D2 relative to D1, while negative coefficients indicate lower scores for D2 relative to D1. Partial eta squared (η²) measure effect size for the variance explained by specific factors in ANOVA. Depressive and anxiety symptoms, neuroticism-related traits Depressive symptom items were selected based on Howard et al.: 17 endorsement of at least one core symptom experienced for two weeks or more, namely persistent sadness or loss of interest. Additional six depressive symptoms reflecting functional impairments during the worst period of depression were included: tiredness, changes in sleep patterns, difficulty concentrating, feelings of worthlessness, thoughts of death, and weight changes. Following Thorp et al., 18 depressive symptoms were also selected from endorsement of the following items based on Patient Health Questionnaire-9 (PHQ-9): 67 recent lack of interest or pleasure in doing things, recent poor appetite or overeating, recent trouble concentrating on things, recent feelings of depression, recent feelings of tiredness or low energy, recent feelings of inadequacy, recent changes in speed or amount of moving or speaking, trouble falling or staying asleep or sleeping too much, and recent thoughts of suicide or self-harm. Anxiety-related symptoms were selected based on Thorp et al. 18 in the seven items: feeling nervous or anxious, feeling of foreboding, easy annoyance or irritability, restlessness, trouble relaxing, worrying too much about different things, inability to stop worrying over the last two weeks. Items are based on Generalised Anxiety Disorder-7 scale (GAD-7). 68 Neuroticism-related traits were based on endorsement of twelve items following Okbay et al., 19 which had been assessed from the 12-item Eysenck Personality Inventory framework: 69 mood swings, miserableness, irritability, sensitivity or hurt feelings, fed-up feelings, nervous feelings, worrier or anxious feelings, tense or highly strung, worrying too long after embarrassment, suffering from nerves, loneliness or isolation, and guilt feelings. Adverse life events, self-harm and suicide, lifestyle items Adverse life events were based on data fields assessing five items reflecting experiences of violence related to physical or sexual as an adult and three items related to childhood experiences of physical abuse, sexual molestation and feeling hated by a family member as a child. Self-harm and suicide attempts were assessed in individual items. Lifestyle items which had shown a significant association with suicide attempts in Zhang et al. 20 were selected: ever smoked, current smoking, alcohol use, age of first sexual intercourse, number of sexual partners, and sleep disturbances. Metabolomics A total of 68 measures were examined, reflecting respiratory health, cardiovascular health, metabolic processes, lipid profiles, blood cell counts, lipid metabolism, and inflammation, which have been associated with depression, 21 biological age 70 and metabolic syndrome. 14 Each measure was standardized, allowing beta coefficients to reflect standardized effect sizes. Positive beta coefficients indicate higher scores for D2 relative to D1, while negative coefficients indicate lower scores for D2 relative to D1. Full description of all measures is presented in Supplementary Materials. Genome-wide association studies (GWAS) analysis Imputed genetic data were downloaded from UK Biobank ( https://www.ukbiobank.ac.uk/enable-your-research/about-our-data/genetic-data ) in July 2021. 71 Genotyped and imputed single nucleotide polymorphisms (SNP) (or single nucleotide variations (SNV)) were pre-processed following a quality check protocol. We extracted and excluded participants with mismatched genetically identified sex and self-reported sex; chromosome aneuploidy; and related individuals (2nd-degree related individuals) via King software relationship inference. 72 Duplicate variants, variants with minor allele frequency less than 1%, variants with missing rate higher than 3%, variants which failed the Hardy-Weinberg test at threshold \(\:p<1\times\:{10}^{-10}\) . Participants with more than 3% missing genotypes were excluded. To adjust for population stratification, the first 40 principal components (PC) were derived using PLINK 2 (v2.0.0). 73 30,376 samples and 6,288,959 variants for GWAS analysis in UKB participants with European ancestry were filtered. We performed linear regression for D1 and D2 dimensional scores respectively on autosome variants. Age at imaging scan, sex, ICV and first 40 PC components were included as covariates. After calculating the association analysis via PLINK2, Functional Mapping and Annotation (FUMA) ( https://fuma.ctglab.nl/tutorial#snp2gene ) identified the significant independent SNPs having genome-wide significant threshold with \(\:p\le\:5\times\:{10}^{-8}\) and are independent of each other at \(\:{r}^{2}<0.6\) . Each candidate SNP was queried in the GWAS Catalog to check for any published associations with previous GWAS studies. Research Domain Criteria The Research Domain Criteria (RDoC) framework aims to understand mental illness according to domains which exemplify types of neurobiological functioning based on biological underpinnings, 74 consisting of six major functional domains: negative valence systems (NVS), positive valence systems (PVS), cognitive systems (CS), arousal / regulatory systems (ARS), sensorimotor systems (SS), and social processes (SP). Citrome et al. 75 identified HAMD and MADRS items which align with five RDoC domains (NVS, PVS, CS, ARS, SS), and Ahmed et al. 76 defined three MDD phenotypes (core depression (CD), anxiety (ANX), and neurovegetative symptoms of melancholia) based on HAMD items to represent RDoC constructs, calculating a threshold for scoring across items to determine if an individual would be classified as positive or negative for each phenotype. We applied these items, 75 , 76 to examine RDoC domain phenotype in the present D1 and D2 classification. 11 Total scores for all items within each phenotype were transformed into a percentage score of the total possible score to enable standardised comparison between scores assessed using MADRS and HAMD. Additionally, we identified corresponding items on the MADRS scale to measure the three phenotypes that were measured using HAMD items by Ahmed et al. 76 and created comparative thresholds to determine positive or negative phenotype classification (Supplementary Table 10). Phenotype scores between D1 and D2 groups were analysed by general linear modelling ANOVA. Proportion of participants who were phenotype positive or negative at baseline was analysed by Chi-square test. P-values were corrected using FDR. Sensitivity analyses We repeated the external validation and phenotype associations in UKB subgroups which overlapped with the COORDINATE-MDD sample age range: 5 , 11 (1) having a lower age of 56 years and an upper age of 65 years (n = 6523), and (2) having an age range of ten years of the youngest UKB participant (ages 45 to 55 years) (n = 3112) (Supplementary Tables 30 and 31). Declarations Data availability UK Biobank data is accessible only to approved researchers through the UK Biobank Research Analysis Platform (UKB-RAP) upon application and approval by the UK Biobank Access Management Committee. CAN-BIND data are available from https://www.braincode.ca/; EMBARC data are available from https://nda.nih.gov/edit_collection.html?id=2199; original data are available from individual co-authors, and the derived data are available on reasonable request to corresponding authors C.H.Y.F. and C.D. Code availability The MUSE algorithm for image segmentation is available at https://www.nitrc.org/projects/cbica_muse. The HYDRA algorithm is available at https://github.com/evarol/HYDRA. The MIDAS algorithm is available at https://github.com/evarol/MIDAS. The following R packages were used: WebPower 0.8.6 (https://cran.r-project.org/web/packages/WebPower/WebPower.pdf), effectsize 0.8.2, and ggplot2 3.4.0. (https://cran.r-project.org/web/packages/ggplot2/ggplot2.pdf). Competing Interests Statement S.R.A. has consulted for Indoc Research Canada. B.W.D. has received research support from Boehringer-Ingelheim, Compass, Pathways, NIMH, Otsuka, and Uson and honoraria for consulting from Aya Biosciences, Myriad Neuroscience, Otsuka, Sophren Therapeutics, Cerebral Therapeutics, Sage. C.H.Y.F. has received grant funding from Brain and Behavior (NARSAD), Eli Lilly and Co. Milken Institute, Flow Neuroscience, MRC, NIMH, Rosetrees Trust, and Wellcome Trust and is Section Editor of Brain Research Bulletin. C.J.H. serves as a consultant for P1vital, Lundbeck, Servier, and Compass Pathways. She holds grant income from Zogenix and J&J. S.H.K. has received funding for consulting or speaking engagements from Abbvie, Boehringer-Ingelheim, Janssen, Lundbeck, Lundbeck Institute, Merck, Otsuka Pfizer, Sunovion, and Servier. He has received research support from Abbott, Brain Canada, CIHR (Canadian Institutes of Health Research), Janssen, Lundbeck, Ontario Brain Institute, Otsuka, Pfizer, and SPOR (Canada’s Strategy for Patient-Oriented Research). He has stock/stock options in Field Trip Health. G.M.K. has served as a speaker for Angelini, Abbvie, Cybin, and H. Lundbeck and as an advisor for Sanos, Onsero, Pangea Botanica, Gilgamesh, and Seaport Therapeutics. H.S.M. has received grant funding from NIH, grant support from Wellcome Leap and Hope for Depression Research Foundation, and consulting and IP licenses fees from Abbott Labs. A.M.M. has received research support from Eli Lilly, Janssen, and The Sackler Trust. A.M.M. has also received speaker fees from Illumina and Janssen. W.E.C. serves on the National Advisory Board of the George West Mental health Foundation, as a board member of Hugarheill ehf (an Icelandic company dedicated to the prevention of depression), and on the scientific advisory boards of AIM for Mental Health and the Anxiety and Depression Association of America; he is supported by the Mary and John Brock Foundation, the Pitts Foundation, and the Fuqua family foundations, and he receives book royalties from John Wiley. H.S. has received grant funding from NIH. S.C.S. received research support from Brain Canada, CIHR (Canadian Institutes of Health Research), Ontario Brain Institute, and CFI (Canadian Foundation for Innovation). S.C.S. is a founder and shareholder of ADMdx, Inc. D.T. has received grant funding from NIH. I.H.G. has received grant funding from NIH. M.H.T. received research support from NIH, PCORI, and AFSP, is a consultant for Alkermes Inc., Alto Neuroscience Inc, Axsome Therapeutics, Boegringer Ingelheim, GH Research, GreenLight VitalSign6 Inc, Heading Health, Inc., Janssen Pharmaceutical, Legion Health, Merck Sharp & Dohme Corp., Mind Medicine Inc., Navitor, Neurocrine Biosciences Inc., Noema Pharma AG, Orexo US Inc., Otsuka Canada Pharmaceutical Inc, Otsuka Pharmaceutical Development & Commercialization, Inc. (MDD section adviser), SAGE Therapeutics, Signant Health, and Takeda Pharmaceuticals Inc and receives editorial compensation from Oxford University Press. A.H.Y. reports the following conflicts of interest: paid lectures and advisory boards for the following companies: Astrazenaca, Eli Lilly, Lundbeck, Sunovion, Servier, Livanova, Janssen, Allegan, Bionomics, Sumitomo Dainippon Pharma, COMPASS, Sage, Novartis; consultant to Johnson & Johnson and to Livanova; received honoraria for attending advisory boards and presenting talks at meetings organized by LivaNova; principal investigator in the Restore-Life VNS registry study funded by LivaNova; UK chief investigator for Novartis MDD study MIJ821A12201; principal investigator on ESKETINTRD3004: ‘An Open-label, Long-term, Safety and Efficacy Study of Intranasal Esketamine in Treatment-resistant Depression’; principal investigator on ‘The Effects of Psilocybin on Cognitive Function in Healthy Participants’; principal investigator on ‘The Safety and Efficacy of Psilocybin in Participants with Treatment-Resistant Depression (P-TRD)’; no shareholdings in pharmaceutical companies; deputy editor of BJPsych Open. Grant funding (past and present): NIMH (USA); CIHR (Canada); NARSAD (USA); Stanley Medical Research Institute (USA); MRC (UK); Wellcome Trust (UK); Royal College of Physicians (Edin); BMA (UK); UBC-VGH Foundation (Canada); WEDC (Canada); CCS Depression Research Fund (Canada); MSFHR (Canada); NIHR (UK); Janssen (UK). R.Z. is a private psychiatrist service provider at The London Depression Institute and co-investigator on a Livanova-funded observational study of vagus nerve stimulation for depression. R.Z. has received honoraria for talks at medical symposia sponsored by Lundbeck as well as Janssen. He has collaborated with EMIS PLC and advises Depsee Ltd. He is affiliated with the D’Or Institute of Research and Education, Rio de Janeiro, and advises the Scients Institute, USA. Authors A. Singh, A. Stolicyn, B.R.G., B.N.F., C.C.F., C.-G.Y., C.D., D.A., D.S., D.W., G.E., H.C.W., I.M.A., I.S., J.F.W.D., J.Q., J.S., J.A.G., K.H., K.S.C., K.Q., M.P.P., M.A., M.D.S., M.G., Q.G., R.E., R.D.W., S.H., S.R., T.A.V., T.C., V.G.F., Y.C., Y.F., and W.X declare that they have no competing interests. Author Contributions C.H.Y.F. and C.D. led the project. C.H.Y.F., G.E., W.X., and C.D. were responsible for the study concept and the design of the study . B.W.D., C.C.F., S.H.K., M.H.T., T.A.V., M.P.P., D.A., I.M.A., B.R.G., H.S.M., W.E.C., Q.G., M.D.S., I.H.G., A.M.M., A. Stolicyn, and H.C.W. contributed to the data acquisition. W.X., R.D.W., Y.C., M.A., and D.S. conducted the data analysis and created the figures. J.W., H.S., A. Singh, C.H.Y.F., and C.D. supervised the statistical analysis. C.H.Y.F. and C.D. interpreted the data. J.W., Y.F., D.A., S.H., A.M.M., R.Z., H.S.M., and C.C.F. provided crucial advice for the study. W.X, R.D.W., Y.C. and C.H.Y.F. wrote the manuscript. S.R.A., T.C., K.S.C., B.N.F., V.G.F., M.G., B.R.G., K.H., K.Q., S.R., M.D.S., J.S., J.A.G., A. Stolicyn, I.S., S.C.S., D.T., T.A.V., D.W., I.M.A., W.E.C., J.F.W.D., B.W.D., R.E., Q.G., I.H.G., C.J.H., S.H.K., G.M.K., H.S.M., M.P.P., J.Q., M.H.T., H.C.W., C.-G.Y., and A.H.Y. made substantial contributions to the manuscript and provided critical comments. References James, S. L. et al. Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. The Lancet 392 , 1789–1858 (2018). https://doi.org/10.1016/S0140-6736(18)32279-7 American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, (DSM-5®) . (American Psychiatric Association, 2013). Fu, C. H. Y., Fan, Y. & Davatzikos, C. Addressing heterogeneity (and homogeneity) in treatment mechanisms in depression and the potential to develop diagnostic and predictive biomarkers. NeuroImage Clin. 24 , 101997 (2019). https://doi.org/10.1016/j.nicl.2019.101997 Fu, C. H. Y. & Costafreda, S. G. Neuroimaging-based biomarkers in psychiatry: Clinical opportunities of a paradigm shift. Can. J. Psychiatry 58 , 499–508 (2013). https://doi.org/10.1177/070674371305800904 Fu, C. H. Y. et al. AI-based dimensional neuroimaging system for characterizing heterogeneity in brain structure and function in major depressive disorder: COORDINATE-MDD consortium design and rationale. BMC Psychiatry 23 , 59 (2023). https://doi.org/10.1186/s12888-022-04509-7 Costafreda, S. G., Chu, C., Ashburner, J. & Fu, C. H. Y. Prognostic and diagnostic potential of the structural neuroanatomy of depression. PLoS ONE 4 , e6353 (2009). https://doi.org/10.1371/journal.pone.0006353 Sankar, A. et al. Diagnostic potential of structural neuroimaging for depression from a multi-ethnic community sample. BJPsych Open 2 , 247–254 (2016). https://doi.org/10.1192/bjpo.bp.115.002493 Wen, J. et al. Characterizing heterogeneity in neuroimaging, cognition, clinical symptoms, and genetics among patients with late-life depression. JAMA Psychiatry 79 , 464–474 (2022). https://doi.org/10.1001/jamapsychiatry.2022.0020 Enneking, V., Leehr, E. J., Dannlowski, U. & Redlich, R. Brain structural effects of treatments for depression and biomarkers of response: a systematic review of neuroimaging studies. Psychol. Med. 50 , 187–209 (2020). https://doi.org/10.1017/S0033291719003660 Drysdale, A. T. et al. Resting-state connectivity biomarkers define neurophysiological subtypes of depression. Nat. Med. 23 , 28–38 (2017). https://doi.org/10.1038/nm.4246 Fu, C. H. Y. et al. Neuroanatomical dimensions in medication-free individuals with major depressive disorder and treatment response to SSRI antidepressant medications or placebo. Nat. Ment. Health 2 , 164–176 (2024). https://doi.org/10.1038/s44220-023-00187-w Miller, K. L. et al. Multimodal population brain imaging in the UK Biobank prospective epidemiological study. Nat. Neurosci. 19 , 1523–1536 (2016). https://doi.org/10.1038/nn.4393 Alfaro-Almagro, F. et al. Image processing and Quality Control for the first 10,000 brain imaging datasets from UK Biobank. NeuroImage 166 , 400–424 (2018). https://doi.org/10.1016/j.neuroimage.2017.10.034 Qureshi, D. et al. Association of Metabolic Syndrome with Neuroimaging and Cognitive Outcomes in the UK Biobank. Diabetes Care 47 , 1415–1423 (2024). https://doi.org/10.2337/dc24-0537 Colombo, F. et al. Multimodal brain-derived subtypes of Major depressive disorder differentiate patients for anergic symptoms, immune-inflammatory markers, history of childhood trauma and treatment-resistance. Eur. Neuropsychopharmacol. 85 , 45–57 (2024). https://doi.org/10.1016/j.euroneuro.2024.05.015 Davis, K. A. S. et al. Mental health in UK Biobank - development, implementation and results from an online questionnaire completed by 157 366 participants: a reanalysis. BJPsych Open 6 , e18 (2020). https://doi.org/10.1192/bjo.2019.100 Howard, D. M. et al. Genetic stratification of depression in UK Biobank. Transl. Psychiatry 10 , 163 (2020). https://doi.org/10.1038/s41398-020-0848-0 Thorp, J. G. et al. Symptom-level modelling unravels the shared genetic architecture of anxiety and depression. Nat. Hum. Behav. 5 , 1432–1442 (2021). https://doi.org/10.1038/s41562-021-01094-9 Okbay, A. et al. Genome-wide association study identifies 74 loci associated with educational attainment. Nature 533 , 539–542 (2016). https://doi.org/10.1038/nature17671 Zhang, B. et al. Identifying behaviour-related and physiological risk factors for suicide attempts in the UK Biobank. Nat. Hum. Behav. 8 , 1784–1797 (2024). https://doi.org/10.1038/s41562-024-01903-x Amin, N. et al. Interplay of Metabolome and Gut Microbiome in Individuals with Major Depressive Disorder vs Control Individuals. JAMA Psychiatry 80 , 597–609 (2023). https://doi.org/10.1001/jamapsychiatry.2023.0685 Sollis, E. et al. The NHGRI-EBI GWAS Catalog: knowledgebase and deposition resource. Nucleic Acids Res. 51 , D977–D985 (2023). https://doi.org/10.1093/nar/gkac1010 Nagel, M. et al. Meta-analysis of genome-wide association studies for neuroticism in 449,484 individuals identifies novel genetic loci and pathways. Nat. Genet. 50 , 920–927 (2018). https://doi.org/10.1038/s41588-018-0151-7 Trubetskoy, V. et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nature 604 , 502–508 (2022). https://doi.org/10.1038/s41586-022-04434-5 Makowski, C. et al. Larger cerebral cortex is genetically correlated with greater frontal area and dorsal thickness. Proc. Natl. Acad. Sci. 120 , e2214834120 (2023). https://doi.org/10.1073/pnas.2214834120 Liu, N. et al. Cross-ancestry genome-wide association meta-analyses of hippocampal and subfield volumes. Nat. Genet. 55 , 1126–1137 (2023). https://doi.org/10.1038/s41588-023-01425-8 Martin, S. et al. Genetic Evidence for Different Adiposity Phenotypes and Their Opposing Influences on Ectopic Fat and Risk of Cardiometabolic Disease. Diabetes 70 , 1843–1856 (2021). https://doi.org/10.2337/db21-0129 Teder-Laving, M. et al. Genome-wide meta-analysis identifies novel loci conferring risk of acne vulgaris. Eur. J. Hum. Genet. 32 , 1136–1143 (2024). https://doi.org/10.1038/s41431-023-01326-8 Kim, J., Lee, J., Nam, K. & Lee, S. Investigation of genetic variants and causal biomarkers associated with brain aging. Sci. Rep. 13 , 1526 (2023). https://doi.org/10.1038/s41598-023-27903-x van der Meer, D. et al. The genetic architecture of human cortical folding. Sci. Adv. 7 , eabj9446 (2021). https://doi.org/10.1126/sciadv.abj9446 Harris, B. H. L. et al. New role of fat-free mass in cancer risk linked with genetic predisposition. Sci. Rep. 14 , 7270 (2024). https://doi.org/10.1038/s41598-024-54291-7 Melin, B. S. et al. Genome-wide association study of glioma subtypes identifies specific differences in genetic susceptibility to glioblastoma and non-glioblastoma tumors. Nat. Genet. 49 , 789–794 (2017). https://doi.org/10.1038/ng.3823 Wang, H. et al. Genome-wide interaction analysis of pathological hallmarks in Alzheimer’s disease. Neurobiol. Aging 93 , 61–68 (2020). https://doi.org/10.1016/j.neurobiolaging.2020.04.025 Zhao, B. et al. Large-scale GWAS reveals genetic architecture of brain white matter microstructure and genetic overlap with cognitive and mental health traits (n = 17,706). Mol. Psychiatry 26 , 3943–3955 (2021). https://doi.org/10.1038/s41380-019-0569-z Cox, S. R., Ritchie, S. J., Fawns-Ritchie, C., Tucker-Drob, E. M. & Deary, I. J. Structural brain imaging correlates of general intelligence in UK Biobank. Intelligence 76 , 101376 (2019). https://doi.org/10.1016/j.intell.2019.101376 Lammer, L. et al. Impact of social isolation on grey matter structure and cognitive functions: A population-based longitudinal neuroimaging study. eLife 12 , e83660 (2023). https://doi.org/10.7554/eLife.83660 Liu, R. T., Kraines, M. A., Massing-Schaffer, M. & Alloy, L. B. Rejection Sensitivity and Depression: Mediation by Stress Generation. Psychiatry 77 , 86–97 (2014). https://doi.org/10.1521/psyc.2014.77.1.86 Vekic, J., Stefanovic, A. & Zeljkovic, A. Obesity and Dyslipidemia: A Review of Current Evidence. Curr. Obes. Rep. 12 , 207–222 (2023). https://doi.org/10.1007/s13679-023-00518-z Chourpiliadis, C. et al. Metabolic Profile and Long-Term Risk of Depression, Anxiety, and Stress-Related Disorders. JAMA Netw. Open 7 , e244525 (2024). https://doi.org/10.1001/jamanetworkopen.2024.4525 Lv, X. et al. Body composition, lifestyle, and depression: a prospective study in the UK biobank. BMC Public Health 24 , 393 (2024). https://doi.org/10.1186/s12889-024-17891-6 Mujica-Parodi, L. R., Renelique, R. & Taylor, M. K. Higher body fat percentage is associated with increased cortisol reactivity and impaired cognitive resilience in response to acute emotional stress. Int. J. Obes. 33 , 157–165 (2009). https://doi.org/10.1038/ijo.2008.218 Cole, J. H. et al. Body mass index, but not FTO genotype or major depressive disorder, influences brain structure. Neuroscience 252 , 109–117 (2013). https://doi.org/10.1016/j.neuroscience.2013.07.015 Hidese, S. et al. Association of body mass index and its classifications with gray matter volume in individuals with a wide range of body mass index group: A whole-brain magnetic resonance imaging study. Front. Hum. Neurosci. 16 , (2022). https://doi.org/10.3389/fnhum.2022.926804 Lamers, F. et al. Depression profilers and immuno-metabolic dysregulation: Longitudinal results from the NESDA study. Brain. Behav. Immun. 88 , 174–183 (2020). https://doi.org/10.1016/j.bbi.2020.04.002 Brydges, C. R. et al. Metabolomic and inflammatory signatures of symptom dimensions in major depression. Brain. Behav. Immun. 102 , 42–52 (2022). https://doi.org/10.1016/j.bbi.2022.02.003 Huang, J. et al. Genomics and phenomics of body mass index reveals a complex disease network. Nat. Commun. 13 , 7973 (2022). https://doi.org/10.1038/s41467-022-35553-2 Hamer, M. & Batty, G. D. Association of body mass index and waist-to-hip ratio with brain structure. Neurology 92 , e594–e600 (2019). https://doi.org/10.1212/WNL.0000000000006879 Pflanz, C.-P. et al. Central obesity is selectively associated with cerebral gray matter atrophy in 15,634 subjects in the UK Biobank. Int. J. Obes. 46 , 1059–1067 (2022). https://doi.org/10.1038/s41366-021-00992-2 Gautam, P., Anstey, K. J., Wen, W., Sachdev, P. S. & Cherbuin, N. Cortical gyrification and its relationships with cortical volume, cortical thickness, and cognitive performance in healthy mid-life adults. Behav. Brain Res. 287 , 331–339 (2015). https://doi.org/10.1016/j.bbr.2015.03.018 Penttilä, J. et al. Cortical folding in patients with bipolar disorder or unipolar depression. J. Psychiatry Neurosci. JPN 34 , 127–135 (2009). Garel, C. et al. Fetal cerebral cortex: normal gestational landmarks identified using prenatal MR imaging. AJNR Am. J. Neuroradiol. 22 , 184–189 (2001). Dubois, J. et al. Primary cortical folding in the human newborn: an early marker of later functional development. Brain 131 , 2028–2041 (2008). https://doi.org/10.1093/brain/awn137 Hedman, A. M., van Haren, N. E. M., Schnack, H. G., Kahn, R. S. & Hulshoff Pol, H. E. Human brain changes across the life span: a review of 56 longitudinal magnetic resonance imaging studies. Hum. Brain Mapp. 33 , 1987–2002 (2012). Yang, W. et al. The effects of childhood maltreatment on cortical thickness and gray matter volume: a coordinate-based meta-analysis. Psychol. Med. 53 , 1681–1699 (2023). https://doi.org/10.1017/S0033291723000661 Ireton, R., Hughes, A. & Klabunde, M. A Functional Magnetic Resonance Imaging Meta-Analysis of Childhood Trauma. Biol. Psychiatry Cogn. Neurosci. Neuroimaging 9 , 561–570 (2024). https://doi.org/10.1016/j.bpsc.2024.01.009 Lim, L., Howells, H., Radua, J. & Rubia, K. Aberrant structural connectivity in childhood maltreatment: A meta-analysis. Neurosci. Biobehav. Rev. 116 , 406–414 (2020). https://doi.org/10.1016/j.neubiorev.2020.07.004 Green, J. G. et al. Childhood Adversities and Adult Psychiatric Disorders in the National Comorbidity Survey Replication I: Associations With First Onset of DSM-IV Disorders. Arch. Gen. Psychiatry 67 , 113–123 (2010). https://doi.org/10.1001/archgenpsychiatry.2009.186 Ringwald, K. G. et al. Interaction of recent stressful life events and childhood abuse on orbitofrontal grey matter volume in adults with depression. J. Affect. Disord. 312 , 122–127 (2022). https://doi.org/10.1016/j.jad.2022.06.050 Keyes, K. M. & Westreich, D. UK Biobank, big data, and the consequences of non-representativeness. The Lancet 393 , 1297 (2019). https://doi.org/10.1016/S0140-6736(18)33067-8 Sudlow, C. et al. UK Biobank: An Open Access Resource for Identifying the Causes of a Wide Range of Complex Diseases of Middle and Old Age. PLOS Med. 12 , e1001779 (2015). https://doi.org/10.1371/journal.pmed.1001779 Doshi, J. et al. MUSE: MUlti-atlas region Segmentation utilizing Ensembles of registration algorithms and parameters, and locally optimal atlas selection. NeuroImage 127 , 186–195 (2016). https://doi.org/10.1016/j.neuroimage.2015.11.073 Varol, E., Sotiras, A. & Davatzikos, C. HYDRA: Revealing heterogeneity of imaging and genetic patterns through a multiple max-margin discriminative analysis framework. NeuroImage 145 , 346–364 (2017). https://doi.org/10.1016/j.neuroimage.2016.02.041 Fortin, J.-P. et al. Harmonization of cortical thickness measurements across scanners and sites. NeuroImage 167 , 104–120 (2018). https://doi.org/10.1016/j.neuroimage.2017.11.024 Davatzikos, C., Genc, A., Xu, D. & Resnick, S. M. Voxel-based morphometry using the RAVENS maps: methods and validation using simulated longitudinal atrophy. NeuroImage 14 , 1361–1369 (2001). https://doi.org/10.1006/nimg.2001.0937 Srinivasan, D. et al. A comparison of Freesurfer and multi-atlas MUSE for brain anatomy segmentation: Findings about size and age bias, and inter-scanner stability in multi-site aging studies. NeuroImage 223 , 117248 (2020). https://doi.org/10.1016/j.neuroimage.2020.117248 Fawns-Ritchie, C. & Deary, I. J. Reliability and validity of the UK Biobank cognitive tests. PLoS ONE 15 , e0231627 (2020). https://doi.org/10.1371/journal.pone.0231627 Kroenke, K., Spitzer, R. & Williams, W. The PHQ-9: Validity of a brief depression severity measure. J. Gen. Intern. Med. 16 , 606–616 (2001). Spitzer, R. L., Kroenke, K., Williams, J. B. W. & Löwe, B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch. Intern. Med. 166 , 1092–1097 (2006). Eysenck, H. J. & Eysenck, S. B. G. Manual of the Eysenck Personality Questionnaire (Junior and Adult) . (Hodder and Stoughton Educational, 1975). Mao, R. et al. Association of biological age acceleration with cardiac morphology, function, and incident heart failure: insights from UK Biobank participants. Eur. Heart J. - Cardiovasc. Imaging 25 , 1315–1323 (2024). https://doi.org/10.1093/ehjci/jeae126 Bycroft, C. et al. The UK Biobank resource with deep phenotyping and genomic data. Nature 562 , 203–209 (2018). https://doi.org/10.1038/s41586-018-0579-z Manichaikul, A. et al. Robust relationship inference in genome-wide association studies. Bioinformatics 26 , 2867–2873 (2010). https://doi.org/10.1093/bioinformatics/btq559 Purcell, S. et al. PLINK: A Tool Set for Whole-Genome Association and Population-Based Linkage Analyses. Am. J. Hum. Genet. 81 , 559–575 (2007). Insel, T. et al. Research Domain Criteria (RDoC): Toward a new classification framework for research on mental disorders. Am. J. Psychiatry 167 , 748–751 (2010). https://doi.org/10.1176/appi.ajp.2010.09091379 Citrome, L. et al. Making sense of the matrix: A qualitative assessment and commentary on connecting psychiatric symptom scale items to the Research Domain Criteria (RDoC). Innov. Clin. Neurosci. 19 , 26–32 (2022). Ahmed, A. T. et al. Mapping depression rating scale phenotypes onto research domain criteria (RDoC) to inform biological research in mood disorders. J. Affect. Disord. 238 , 1–7 (2018). https://doi.org/10.1016/j.jad.2018.05.005 Additional Declarations Yes there is potential Competing Interest. S.R.A. has consulted for Indoc Research Canada. B.W.D. has received research support from Boehringer-Ingelheim, Compass, Pathways, NIMH, Otsuka, and Uson and honoraria for consulting from Aya Biosciences, Myriad Neuroscience, Otsuka, Sophren Therapeutics, Cerebral Therapeutics, Sage. C.H.Y.F. has received grant funding from Brain and Behavior (NARSAD), Eli Lilly and Co. Milken Institute, Flow Neuroscience, MRC, NIMH, Rosetrees Trust, and Wellcome Trust and is Section Editor of Brain Research Bulletin. C.J.H. serves as a consultant for P1vital, Lundbeck, Servier, and Compass Pathways. She holds grant income from Zogenix and J&J. S.H.K. has received funding for consulting or speaking engagements from Abbvie, Boehringer-Ingelheim, Janssen, Lundbeck, Lundbeck Institute, Merck, Otsuka Pfizer, Sunovion, and Servier. He has received research support from Abbott, Brain Canada, CIHR (Canadian Institutes of Health Research), Janssen, Lundbeck, Ontario Brain Institute, Otsuka, Pfizer, and SPOR (Canada’s Strategy for Patient-Oriented Research). He has stock/stock options in Field Trip Health. G.M.K. has served as a speaker for Angelini, Abbvie, Cybin, and H. Lundbeck and as an advisor for Sanos, Onsero, Pangea Botanica, Gilgamesh, and Seaport Therapeutics. H.S.M. has received grant funding from NIH, grant support from Wellcome Leap and Hope for Depression Research Foundation, and consulting and IP licenses fees from Abbott Labs. A.M.M. has received research support from Eli Lilly, Janssen, and The Sackler Trust. A.M.M. has also received speaker fees from Illumina and Janssen. W.E.C. serves on the National Advisory Board of the George West Mental health Foundation, as a board member of Hugarheill ehf (an Icelandic company dedicated to the prevention of depression), and on the scientific advisory boards of AIM for Mental Health and the Anxiety and Depression Association of America; he is supported by the Mary and John Brock Foundation, the Pitts Foundation, and the Fuqua family foundations, and he receives book royalties from John Wiley. H.S. has received grant funding from NIH. S.C.S. received research support from Brain Canada, CIHR (Canadian Institutes of Health Research), Ontario Brain Institute, and CFI (Canadian Foundation for Innovation). S.C.S. is a founder and shareholder of ADMdx, Inc. D.T. has received grant funding from NIH. I.H.G. has received grant funding from NIH. M.H.T. received research support from NIH, PCORI, and AFSP, is a consultant for Alkermes Inc., Alto Neuroscience Inc, Axsome Therapeutics, Boegringer Ingelheim, GH Research, GreenLight VitalSign6 Inc, Heading Health, Inc., Janssen Pharmaceutical, Legion Health, Merck Sharp & Dohme Corp., Mind Medicine Inc., Navitor, Neurocrine Biosciences Inc., Noema Pharma AG, Orexo US Inc., Otsuka Canada Pharmaceutical Inc, Otsuka Pharmaceutical Development & Commercialization, Inc. (MDD section adviser), SAGE Therapeutics, Signant Health, and Takeda Pharmaceuticals Inc and receives editorial compensation from Oxford University Press. A.H.Y. reports the following conflicts of interest: paid lectures and advisory boards for the following companies: Astrazenaca, Eli Lilly, Lundbeck, Sunovion, Servier, Livanova, Janssen, Allegan, Bionomics, Sumitomo Dainippon Pharma, COMPASS, Sage, Novartis; consultant to Johnson & Johnson and to Livanova; received honoraria for attending advisory boards and presenting talks at meetings organized by LivaNova; principal investigator in the Restore-Life VNS registry study funded by LivaNova; UK chief investigator for Novartis MDD study MIJ821A12201; principal investigator on ESKETINTRD3004: ‘An Open-label, Long-term, Safety and Efficacy Study of Intranasal Esketamine in Treatment-resistant Depression’; principal investigator on ‘The Effects of Psilocybin on Cognitive Function in Healthy Participants’; principal investigator on ‘The Safety and Efficacy of Psilocybin in Participants with Treatment-Resistant Depression (P-TRD)’; no shareholdings in pharmaceutical companies; deputy editor of BJPsych Open. Grant funding (past and present): NIMH (USA); CIHR (Canada); NARSAD (USA); Stanley Medical Research Institute (USA); MRC (UK); Wellcome Trust (UK); Royal College of Physicians (Edin); BMA (UK); UBC-VGH Foundation (Canada); WEDC (Canada); CCS Depression Research Fund (Canada); MSFHR (Canada); NIHR (UK); Janssen (UK). R.Z. is a private psychiatrist service provider at The London Depression Institute and co-investigator on a Livanova-funded observational study of vagus nerve stimulation for depression. R.Z. has received honoraria for talks at medical symposia sponsored by Lundbeck as well as Janssen. He has collaborated with EMIS PLC and advises Depsee Ltd. He is affiliated with the D’Or Institute of Research and Education, Rio de Janeiro, and advises the Scients Institute, USA. Authors A. Singh, A. Stolicyn, B.R.G., B.N.F., C.C.F., C.-G.Y., C.D., D.A., D.S., D.W., G.E., H.C.W., I.M.A., I.S., J.F.W.D., J.Q., J.S., J.A.G., K.H., K.S.C., K.Q., M.P.P., M.A., M.D.S., M.G., Q.G., R.E., R.D.W., S.H., S.R., T.A.V., T.C., V.G.F., Y.C., Y.F., and W.X declare that they have no competing interests. Supplementary Files reportingsummary.pdf Reporting Summary Supplementarytables14to31.xlsx Supplementary Tables 14 to 31 Supplementarymaterials.pdf Supplementary materials Cite Share Download PDF Status: Published Journal Publication published 15 Nov, 2025 Read the published version in Communications Medicine → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5975731","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":415429177,"identity":"87a499bc-c4e7-43fa-a4dc-6f0b3b81f4ca","order_by":0,"name":"Wenyi Xiao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvUlEQVRIiWNgGAWjYJCCAwwMNgwM7BCOAbFa0hgYmEnRAgSHSdBicIDH8MCPivOJ/cwMjB9+MBw2JkILW8LBnjO3E2c2MzBL9jAcNiOoxewA84EDvG23EzcA3SYNdKENEVoYGw7+bTuXuP8wA/NvIrUwHzjM23YgcQMzAxvIFsIOsz/MlnBY5kyy8YzDjG2WPQbphL0v2d5j/PFNhZ1sf3vz4Rs/KqwNGwjqYYazGBtIiMhRMApGwSgYBXgBAKjrOZEG2my4AAAAAElFTkSuQmCC","orcid":"","institution":"University of East London","correspondingAuthor":true,"prefix":"","firstName":"Wenyi","middleName":"","lastName":"Xiao","suffix":""},{"id":415429176,"identity":"1bdcc231-a2b9-4a14-b1d0-09ac0f945aaf","order_by":1,"name":"Rachel D. Woodham","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAyUlEQVRIiWNgGAWjYNACtgMMDOwNDAwPIDxitfAAcQJpWiQSiNRi3sBj+OBD2R153ZlvDB8kMNjJM0ikJeDVInOAx9hwxrlnhttu5xgbJDAkGzZIpB3Aq0WCgXebNG/bYUagFjOg25gTGCTSG4jSYr/t5hnzHwkM9cRrSdx2g8cM6P3DQC2EHMbM/xnol8PJ286kFUskGBw3bON5loBfC3tbIjDEDttuO35444cPFdXy/OxpBni1MDCj8AyIjMhRMApGwSgYBfgBANWcQepry7MQAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-5581-9449","institution":"University of East London","correspondingAuthor":true,"prefix":"","firstName":"Rachel","middleName":"D.","lastName":"Woodham","suffix":""},{"id":415429178,"identity":"fc978882-3799-488a-b00c-f376b65771c3","order_by":2,"name":"Yuhan Cui","email":"","orcid":"","institution":"UPENN","correspondingAuthor":false,"prefix":"","firstName":"Yuhan","middleName":"","lastName":"Cui","suffix":""},{"id":415429179,"identity":"ade1d966-7e98-401c-9e38-86db218a3b71","order_by":3,"name":"Junaho Wen","email":"","orcid":"","institution":"University of Pennsylvania","correspondingAuthor":false,"prefix":"","firstName":"Junaho","middleName":"","lastName":"Wen","suffix":""},{"id":415429180,"identity":"751d0d4d-9db0-45a2-b7d6-ba1ab95caffb","order_by":4,"name":"Mathilde Antoniades","email":"","orcid":"https://orcid.org/0000-0002-8241-835X","institution":"UPENN","correspondingAuthor":false,"prefix":"","firstName":"Mathilde","middleName":"","lastName":"Antoniades","suffix":""},{"id":415429183,"identity":"0806eaa4-275f-4f73-8016-2dd4354668f9","order_by":5,"name":"Dhivya Srinivasan","email":"","orcid":"","institution":"Center for Biomedical Image Computing and Analytics (CBICA)","correspondingAuthor":false,"prefix":"","firstName":"Dhivya","middleName":"","lastName":"Srinivasan","suffix":""},{"id":415429185,"identity":"03473a2c-c12d-41c0-a644-9231c6c413dc","order_by":6,"name":"Yong Fan","email":"","orcid":"https://orcid.org/0000-0001-9869-4685","institution":"University of Pennsylvania","correspondingAuthor":false,"prefix":"","firstName":"Yong","middleName":"","lastName":"Fan","suffix":""},{"id":415429187,"identity":"c5d04f7d-9345-46d4-bdc7-205df7c32a7a","order_by":7,"name":"Guray Erus","email":"","orcid":"https://orcid.org/0000-0001-6633-4861","institution":"UPenn","correspondingAuthor":false,"prefix":"","firstName":"Guray","middleName":"","lastName":"Erus","suffix":""},{"id":415429189,"identity":"91d7b9ab-1448-4b2c-b8b5-f670d3072036","order_by":8,"name":"Jose Garcia","email":"","orcid":"","institution":"University of Pennsylvania","correspondingAuthor":false,"prefix":"","firstName":"Jose","middleName":"","lastName":"Garcia","suffix":""},{"id":415429191,"identity":"b36aca94-1ff1-42bd-9718-dbbfd30605e9","order_by":9,"name":"Stephen Arnott","email":"","orcid":"https://orcid.org/0000-0002-0616-7771","institution":"Rotman Research Institute","correspondingAuthor":false,"prefix":"","firstName":"Stephen","middleName":"","lastName":"Arnott","suffix":""},{"id":415429192,"identity":"cfa3f469-ea93-4bd8-b4b5-d5e7a99d508c","order_by":10,"name":"Taolin Chen","email":"","orcid":"","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Taolin","middleName":"","lastName":"Chen","suffix":""},{"id":415429193,"identity":"b27dcf9f-efe5-4cf7-b6ba-54befd0d259d","order_by":11,"name":"Ki Sueng Choi","email":"","orcid":"https://orcid.org/0000-0002-6446-0677","institution":"Icahn School of Medicine at Mount Sinai","correspondingAuthor":false,"prefix":"","firstName":"Ki","middleName":"Sueng","lastName":"Choi","suffix":""},{"id":415429195,"identity":"fbbf50f6-3828-483d-b693-ee3ae4da0a81","order_by":12,"name":"Cherise Chin-Fatt","email":"","orcid":"https://orcid.org/0000-0001-5080-6799","institution":"University of Texas Southwestern Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Cherise","middleName":"","lastName":"Chin-Fatt","suffix":""},{"id":415429197,"identity":"e9e9666f-b89e-4bc1-bd88-d187666b5fcd","order_by":13,"name":"Benicio Frey","email":"","orcid":"https://orcid.org/0000-0001-8267-943X","institution":"McMaster University","correspondingAuthor":false,"prefix":"","firstName":"Benicio","middleName":"","lastName":"Frey","suffix":""},{"id":415429198,"identity":"9d867a83-7732-46c8-b716-583a2c6552cf","order_by":14,"name":"Vibe Frokjaer","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Vibe","middleName":"","lastName":"Frokjaer","suffix":""},{"id":415429199,"identity":"773313c9-7306-48b5-bd79-4dd7caef05ef","order_by":15,"name":"Melanie Ganz","email":"","orcid":"https://orcid.org/0000-0002-9120-8098","institution":"Rigshospitalet","correspondingAuthor":false,"prefix":"","firstName":"Melanie","middleName":"","lastName":"Ganz","suffix":""},{"id":415429202,"identity":"92ae4ea8-2bbb-4667-92f4-be8a36c31b63","order_by":16,"name":"Beata Godlewska","email":"","orcid":"https://orcid.org/0000-0002-5973-3765","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Beata","middleName":"","lastName":"Godlewska","suffix":""},{"id":415429203,"identity":"6e976712-21ce-4676-b9db-dbcd11091d63","order_by":17,"name":"Stefanie Hassel","email":"","orcid":"","institution":"University of Calgary","correspondingAuthor":false,"prefix":"","firstName":"Stefanie","middleName":"","lastName":"Hassel","suffix":""},{"id":415429204,"identity":"0ea019df-9af0-4e01-9307-c6d100d6a9f6","order_by":18,"name":"Keith Ho","email":"","orcid":"","institution":"University Health Network","correspondingAuthor":false,"prefix":"","firstName":"Keith","middleName":"","lastName":"Ho","suffix":""},{"id":415429205,"identity":"0f739016-3115-487d-9663-d2cda102dadd","order_by":19,"name":"Andrew McIntosh","email":"","orcid":"https://orcid.org/0000-0002-0198-4588","institution":"University of Edinburgh","correspondingAuthor":false,"prefix":"","firstName":"Andrew","middleName":"","lastName":"McIntosh","suffix":""},{"id":415429207,"identity":"f1548214-bed2-431b-94de-c16277a11fe2","order_by":20,"name":"Kun Qin","email":"","orcid":"https://orcid.org/0000-0003-1254-6388","institution":"Taihe Hospital, Hubei University of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Kun","middleName":"","lastName":"Qin","suffix":""},{"id":415429209,"identity":"7bc86716-711e-46e9-b9b4-4d150c26c0b2","order_by":21,"name":"Susan Rotzinger","email":"","orcid":"","institution":"University Health Network","correspondingAuthor":false,"prefix":"","firstName":"Susan","middleName":"","lastName":"Rotzinger","suffix":""},{"id":415429210,"identity":"e4aca7c4-dcf0-4120-8312-df21958742cc","order_by":22,"name":"Matthew Sacchet","email":"","orcid":"","institution":"Center for Depression, Anxiety, and Stress Research, McLean Hospital, Harvard Medical School, Belmont, USA","correspondingAuthor":false,"prefix":"","firstName":"Matthew","middleName":"","lastName":"Sacchet","suffix":""},{"id":415429213,"identity":"3a694a53-97ed-49d3-9b65-1ce55c9d5200","order_by":23,"name":"Jonathan Savitz","email":"","orcid":"https://orcid.org/0000-0001-8143-182X","institution":"Laureate Institute for Brain Research","correspondingAuthor":false,"prefix":"","firstName":"Jonathan","middleName":"","lastName":"Savitz","suffix":""},{"id":415429214,"identity":"f427cf18-c8d6-4b06-9436-5643bf8fd21f","order_by":24,"name":"Haochang Shou","email":"","orcid":"https://orcid.org/0000-0002-3043-047X","institution":"University of Pennsylvania","correspondingAuthor":false,"prefix":"","firstName":"Haochang","middleName":"","lastName":"Shou","suffix":""},{"id":415429215,"identity":"4ea88d8b-e661-435b-89b9-07bdd6c16375","order_by":25,"name":"Ashish Singh","email":"","orcid":"https://orcid.org/0000-0003-4486-3636","institution":"University of Pennsylvania","correspondingAuthor":false,"prefix":"","firstName":"Ashish","middleName":"","lastName":"Singh","suffix":""},{"id":415429216,"identity":"af773603-9910-4ed1-8e0a-21d6e797ba9e","order_by":26,"name":"Aleks Stolicyn","email":"","orcid":"https://orcid.org/0000-0002-1147-7539","institution":"University of Edinburgh","correspondingAuthor":false,"prefix":"","firstName":"Aleks","middleName":"","lastName":"Stolicyn","suffix":""},{"id":415429217,"identity":"bc4e6e1e-91f2-43b7-91ec-5b15b2939cd1","order_by":27,"name":"Irina Strigo","email":"","orcid":"https://orcid.org/0000-0002-8799-716X","institution":"San Francisco VA Health Care System/UCSF","correspondingAuthor":false,"prefix":"","firstName":"Irina","middleName":"","lastName":"Strigo","suffix":""},{"id":415429218,"identity":"25148085-65aa-4f94-9b1d-2ab172d29d71","order_by":28,"name":"Stephen Strother","email":"","orcid":"https://orcid.org/0000-0002-3198-217X","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Stephen","middleName":"","lastName":"Strother","suffix":""},{"id":415429221,"identity":"3585859e-c711-4b19-980b-6b19540eaa7c","order_by":29,"name":"Duygu Tosun","email":"","orcid":"https://orcid.org/0000-0001-8644-7724","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Duygu","middleName":"","lastName":"Tosun","suffix":""},{"id":415429223,"identity":"2d0e0292-a4f6-4bd4-b008-148549d9b31b","order_by":30,"name":"Dongtao Wei","email":"","orcid":"","institution":"Southwest University","correspondingAuthor":false,"prefix":"","firstName":"Dongtao","middleName":"","lastName":"Wei","suffix":""},{"id":415429225,"identity":"7a7664a8-05c9-4135-8d22-8dbb5ee19b8d","order_by":31,"name":"Roland Zahn","email":"","orcid":"https://orcid.org/0000-0002-8447-1453","institution":"King's College London","correspondingAuthor":false,"prefix":"","firstName":"Roland","middleName":"","lastName":"Zahn","suffix":""},{"id":415429227,"identity":"e5694c09-f6ed-4596-a511-ccdd082a3988","order_by":32,"name":"Ian Anderson","email":"","orcid":"","institution":"University of Manchester","correspondingAuthor":false,"prefix":"","firstName":"Ian","middleName":"","lastName":"Anderson","suffix":""},{"id":415429228,"identity":"bf8dfec3-ad86-4950-b748-69243ef84ace","order_by":33,"name":"Edward Craighead","email":"","orcid":"https://orcid.org/0000-0001-9957-0027","institution":"Emory College of Arts and Sciences","correspondingAuthor":false,"prefix":"","firstName":"Edward","middleName":"","lastName":"Craighead","suffix":""},{"id":415429229,"identity":"fc41ca86-b6fd-4517-99e6-9ef6906bde72","order_by":34,"name":"Bill Deakin","email":"","orcid":"https://orcid.org/0000-0002-2750-962X","institution":"University of Manchester","correspondingAuthor":false,"prefix":"","firstName":"Bill","middleName":"","lastName":"Deakin","suffix":""},{"id":415429230,"identity":"a9a6acc0-0fdc-4b31-87e3-26adeb6c8cfd","order_by":35,"name":"Boadie Dunlop","email":"","orcid":"https://orcid.org/0000-0002-4653-0483","institution":"Emory University School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Boadie","middleName":"","lastName":"Dunlop","suffix":""},{"id":415429231,"identity":"4d4c9ce2-a864-48dd-8d6c-ed9bd6184444","order_by":36,"name":"Rebecca Elliott","email":"","orcid":"https://orcid.org/0000-0002-7602-010X","institution":"University of Manchester","correspondingAuthor":false,"prefix":"","firstName":"Rebecca","middleName":"","lastName":"Elliott","suffix":""},{"id":415429232,"identity":"e031b15e-9b56-4fe6-bc9c-fe92d1fbd4f5","order_by":37,"name":"Qiyong Gong","email":"","orcid":"https://orcid.org/0000-0002-5912-4871","institution":"West China Hospital of Sichuan University","correspondingAuthor":false,"prefix":"","firstName":"Qiyong","middleName":"","lastName":"Gong","suffix":""},{"id":415429233,"identity":"d4542fc1-c678-4007-a527-89fe893e390c","order_by":38,"name":"Ian Gotlib","email":"","orcid":"","institution":"Stanford University","correspondingAuthor":false,"prefix":"","firstName":"Ian","middleName":"","lastName":"Gotlib","suffix":""},{"id":415429234,"identity":"8ac68f5c-551b-44cf-abf5-71e91a0d6146","order_by":39,"name":"Catherine Harmer","email":"","orcid":"https://orcid.org/0000-0002-1609-8335","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Catherine","middleName":"","lastName":"Harmer","suffix":""},{"id":415429235,"identity":"5830cc06-f2fd-4b02-bf4d-7df49006a69e","order_by":40,"name":"Sidney Kennedy","email":"","orcid":"https://orcid.org/0000-0001-5339-7185","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Sidney","middleName":"","lastName":"Kennedy","suffix":""},{"id":415429236,"identity":"567b962d-0673-4321-8242-5a9b12b848e5","order_by":41,"name":"Gitte Knudsen","email":"","orcid":"","institution":"University Hospital Rigshospitalet","correspondingAuthor":false,"prefix":"","firstName":"Gitte","middleName":"","lastName":"Knudsen","suffix":""},{"id":415429237,"identity":"539ea4dd-134d-43dd-b94b-92c74e1163cf","order_by":42,"name":"Helen Mayberg","email":"","orcid":"https://orcid.org/0000-0002-1672-2716","institution":"Icahn School of Medicine at Mount Sinai","correspondingAuthor":false,"prefix":"","firstName":"Helen","middleName":"","lastName":"Mayberg","suffix":""},{"id":415429238,"identity":"4d47ce82-d160-438c-a1cf-64eb18ce1774","order_by":43,"name":"Martin Paulus","email":"","orcid":"https://orcid.org/0000-0002-0825-3606","institution":"Laureate Institute for Brain Research, Tulsa, USA","correspondingAuthor":false,"prefix":"","firstName":"Martin","middleName":"","lastName":"Paulus","suffix":""},{"id":415429239,"identity":"c2b310b2-4554-4f1a-994e-479afddb390c","order_by":44,"name":"Jiang Qiu","email":"","orcid":"","institution":"Southwest University","correspondingAuthor":false,"prefix":"","firstName":"Jiang","middleName":"","lastName":"Qiu","suffix":""},{"id":415429240,"identity":"055169fa-e912-47da-8997-4ab3b47a3c34","order_by":45,"name":"Madhukar Trivedi","email":"","orcid":"https://orcid.org/0000-0002-2983-1110","institution":"The University of Texas Southwestern Medical Center","correspondingAuthor":false,"prefix":"","firstName":"Madhukar","middleName":"","lastName":"Trivedi","suffix":""},{"id":415429241,"identity":"d461a7af-8781-4b8a-9e36-b91336a55d6f","order_by":46,"name":"Heather Whalley","email":"","orcid":"https://orcid.org/0000-0002-4505-8869","institution":"The University of Edinburgh","correspondingAuthor":false,"prefix":"","firstName":"Heather","middleName":"","lastName":"Whalley","suffix":""},{"id":415429242,"identity":"f02e7e7e-4d56-4256-91cb-a6fa557ec24a","order_by":47,"name":"Chao-Gan Yan","email":"","orcid":"https://orcid.org/0000-0003-3413-5977","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Chao-Gan","middleName":"","lastName":"Yan","suffix":""},{"id":415429243,"identity":"4ffad870-917f-4c90-9216-fcec8d21defe","order_by":48,"name":"Allan Young","email":"","orcid":"","institution":"Department of Psychological Medicine, Centre for Affective Disorders, Institute of Psychiatry, Psychology \u0026 Neuroscience, King's College London, 103 Denmark Hill, London, SE5 8AZ, UK","correspondingAuthor":false,"prefix":"","firstName":"Allan","middleName":"","lastName":"Young","suffix":""},{"id":415429244,"identity":"2640b976-6b8d-4a0f-922c-2afb724a9331","order_by":49,"name":"Christos Davatzikos","email":"","orcid":"https://orcid.org/0000-0002-1025-8561","institution":"UPenn","correspondingAuthor":false,"prefix":"","firstName":"Christos","middleName":"","lastName":"Davatzikos","suffix":""},{"id":415429245,"identity":"2844ae26-92a2-43a5-9f0c-10186b922bbb","order_by":50,"name":"Cynthia H.Y. Fu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAr0lEQVRIiWNgGAWjYBAC/gb+DwwfGBhkINwDRGiROMDDwDiDgYGHeC0GDjwMzDykaWHgYZO2bbPhYWA//ICZ5wxRWviPSee2pfEw8KQZMPPcINaW3LbDQIflAF34gVgtlm3/eRj43xCrxQGohbENGG4SIFuIcZjEYR5my55zyTxsEs8MDs4hxvv87T2MN36U2cnx8yc/fPDmGBFaGJihNBsDcbEyCkbBKBgFo4AYAADTPSjHRjovHgAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0003-4313-3500","institution":"University of East London","correspondingAuthor":true,"prefix":"","firstName":"Cynthia","middleName":"H.Y.","lastName":"Fu","suffix":""}],"badges":[],"createdAt":"2025-02-06 18:15:09","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5975731/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5975731/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s43856-025-01219-5","type":"published","date":"2025-11-15T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":79258294,"identity":"c3bf8156-03b7-4540-ae0b-4cb80f03e5e9","added_by":"auto","created_at":"2025-03-26 09:10:21","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":321328,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eExpression of the two dimensions in the UK Biobank sample.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e Neuroanatomical MUSE features used for training and validation. These features represent brain regions used to classify dimensions in the COORD-MDD cohort and validate findings in the UK Biobank dataset. \u003cstrong\u003e(B)\u003c/strong\u003e Training and validation workflow. (a) The pre-trained HYDRA model was developed using the COORD-MDD cohort, identifying two distinct dimensions: Dimension 1 (D1) and Dimension 2 (D2). (b) External validation was conducted by applying the pre-trained HYDRA model to the UK Biobank dataset. Expression scores (E1 for D1, E2 for D2) were computed based on contributions to the respective dimensions. \u003cstrong\u003e(C)\u003c/strong\u003e Quadrant plots illustrating the application of pre-trained HYDRA model, trained on the MDD population, to the external UK Biobank (UKB) sample. The x-axis (E2) and y-axis (E1) represent the expression scores for each individual on Dimension 2 and Dimension 1, respectively. Dimension membership was determined based on these scores: individuals were assigned to Dimension 1 if E1 was greater than 0.3 and E2 was less than -0.3, to Dimension 2 if E1 was less than -0.3 and E2 was greater than 0.3, to the Mixed category if both E1 and E2 were greater than 0.3, and to None if both E1 and E2 were less than -0.3. All other individuals not meeting these criteria were classified as Margin. Panel A depicts the general UKB population (n = 37,235), while Panel B displays the subset of currently depressed participants (n = 1,454).\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5975731/v1/1ce33c4d3d49ba45b6e5ec74.jpg"},{"id":79256883,"identity":"50d715c8-8273-4cdc-b5f0-40ab7c1da0c4","added_by":"auto","created_at":"2025-03-26 09:02:21","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":493996,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBrain Volumetric Differences Across Dimension Membership Groups using MUSE features. \u003c/strong\u003eNeuroanatomical MUSE features comparing D1 and D2 groups show higher volumetric values in D1.\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5975731/v1/04e63c181a88812fe7dc13c3.jpg"},{"id":79258295,"identity":"01d75b7c-19f4-4b0d-8229-4e89a71352b8","added_by":"auto","created_at":"2025-03-26 09:10:21","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":44227,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eUK Biobank general population significant differences in Adverse Life Events and Self Harm variables between D1 and D2. \u003c/strong\u003eParticipants in the D2 group said “yes” to having self-harmed, attempted suicide and to having experienced adverse life events in significantly higher proportions than participants in the D1 group. Only participants who responded “yes” to having previously self-harmed were asked “Have you harmed yourself with the intention to end your life?”. The percentage of participants who said yes is displayed for the following samples: ever-self harmed (D1 n=4847, D2 n=6979); ever attempted suicide (D1 n=229, D2 n=375); Physically abused by family as a child (D1 n=5392, D2 n=7583); Felt hated by family member as a child (D1 n=5383, D2 n=7580); Physical violence by partner or ex-partner as an adult (D1 n=5383, D2 n=7579); Stopped from seeing friends or family by partner or ex-partner as an adult (D1 n=5025, D2 n=7576). Significance was determined using two-sided Chi-Square test. * = p\u0026lt;0.05, ** = p\u0026lt;0.005, *** = p\u0026lt;0.001.\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5975731/v1/167706689900e1224b8953f6.jpg"},{"id":79258310,"identity":"6120c808-2d55-4a6c-8077-7039a9084716","added_by":"auto","created_at":"2025-03-26 09:10:21","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":144446,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eUK Biobank general population metabolomic comparisons between D1 and D2. \u003c/strong\u003eThe x-axis represents the beta value with a negative value indicating that D1 has higher levels than D2 for the metabolite, and a positive beta value indicates that D2 has greater levels than D1. Significant results are presented in blue and non-significant differences are presented in grey. Significant differences between D1 and D2 was determined using two-sided independent samples ANOVA with age and sex included as covariates.\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5975731/v1/6ccd4e3b70512d24b07096f2.jpg"},{"id":79256889,"identity":"5f8c589a-5920-4294-b8a8-2641b3a2c372","added_by":"auto","created_at":"2025-03-26 09:02:21","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":50716,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eUK Biobank general population physical measures between D1 and D2. \u003c/strong\u003eThe x-axis represents the beta value with a negative value indicating that D1 has higher levels than D2 for the physical measure, and a positive beta value indicates that D2 has greater levels than D1. Significant results are presented in blue and non-significant differences are presented in grey. \u0026nbsp;Significant differences between D1 and D2 was determined using two-sided independent samples ANOVA with age and sex included as covariates.\u003c/p\u003e","description":"","filename":"5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5975731/v1/9d9bc8ca70d3d4cb443ef18f.jpg"},{"id":79259231,"identity":"3f2e1b32-e1c5-4364-86dc-588f29445ad3","added_by":"auto","created_at":"2025-03-26 09:18:21","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":297003,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eManhattan plots of Distinct Genetic Profiles in the Genome-Wide Association Study (GWAS) Between Expressions scores 1 and 2.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA) E1 was significantly associated with 10 independent significant SNPs across 7 genomic loci. B) E2 was significantly associated with 9 independent significant SNPs across 7 genomic loci. Independent significant SNPs are SNPs that passed GWAS \u003cem\u003eP \u003c/em\u003evalue threshold (5e-8) (red line) and are independent from each other with r\u003csup\u003e2\u003c/sup\u003e \u0026lt; 0.6. C) QQ plots for E1 analysis. D) QQ plot for E2 analysis.\u003c/p\u003e","description":"","filename":"6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5975731/v1/4e506de845ce7f3cff22451b.jpg"},{"id":97040293,"identity":"08bb1c83-1d1c-49cd-b1f6-e2be5ff6c77e","added_by":"auto","created_at":"2025-11-29 08:10:28","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2788139,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5975731/v1/89d92baf-bfe6-46ab-b168-600928f7ac86.pdf"},{"id":79256881,"identity":"a3c614e6-5868-4dc7-b2fd-94487c09553f","added_by":"auto","created_at":"2025-03-26 09:02:21","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1667606,"visible":true,"origin":"","legend":"Reporting Summary","description":"","filename":"reportingsummary.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5975731/v1/45393e91dadd5cfbd4286df6.pdf"},{"id":79258297,"identity":"3c79b8e6-97ee-4db3-9ced-4bcdc20db9bc","added_by":"auto","created_at":"2025-03-26 09:10:21","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":1710568,"visible":true,"origin":"","legend":"Supplementary Tables 14 to 31","description":"","filename":"Supplementarytables14to31.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5975731/v1/4b4d189ff327da3c161294e7.xlsx"},{"id":79256891,"identity":"f6bdec10-9ca5-4515-9c04-cbcf45c4e391","added_by":"auto","created_at":"2025-03-26 09:02:21","extension":"pdf","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":2255271,"visible":true,"origin":"","legend":"Supplementary materials","description":"","filename":"Supplementarymaterials.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5975731/v1/48da6cc4b7a40f6fb36def5d.pdf"}],"financialInterests":"\u003cb\u003eYes\u003c/b\u003e there is potential Competing Interest.\nS.R.A. has consulted for Indoc Research Canada. \r\nB.W.D. has received research support from Boehringer-Ingelheim, Compass, Pathways, NIMH, Otsuka, and Uson and honoraria for consulting from Aya Biosciences, Myriad Neuroscience, Otsuka, Sophren Therapeutics, Cerebral Therapeutics, Sage. \r\nC.H.Y.F. has received grant funding from Brain and Behavior (NARSAD), Eli Lilly and Co. Milken Institute, Flow Neuroscience, MRC, NIMH, Rosetrees Trust, and Wellcome Trust and is Section Editor of Brain Research Bulletin. \r\nC.J.H. serves as a consultant for P1vital, Lundbeck, Servier, and Compass Pathways. She holds grant income from Zogenix and J\u0026J.\r\nS.H.K. has received funding for consulting or speaking engagements from Abbvie, Boehringer-Ingelheim, Janssen, Lundbeck, Lundbeck Institute, Merck, Otsuka Pfizer, Sunovion, and Servier. He has received research support from Abbott, Brain Canada, CIHR (Canadian Institutes of Health Research), Janssen, Lundbeck, Ontario Brain Institute, Otsuka, Pfizer, and SPOR (Canada’s Strategy for Patient-Oriented Research). He has stock/stock options in Field Trip Health. \r\nG.M.K. has served as a speaker for Angelini, Abbvie, Cybin, and H. Lundbeck and as an advisor for Sanos, Onsero, Pangea Botanica, Gilgamesh, and Seaport Therapeutics.\r\nH.S.M. has received grant funding from NIH, grant support from Wellcome Leap and Hope for Depression Research Foundation, and consulting and IP licenses fees from Abbott Labs. \r\nA.M.M. has received research support from Eli Lilly, Janssen, and The Sackler Trust. A.M.M. has also received speaker fees from Illumina and Janssen. \r\nW.E.C. serves on the National Advisory Board of the George West Mental health Foundation, as a board member of Hugarheill ehf (an Icelandic company dedicated to the prevention of depression), and on the scientific advisory boards of AIM for Mental Health and the Anxiety and Depression Association of America; he is supported by the Mary and John Brock Foundation, the Pitts Foundation, and the Fuqua family foundations, and he receives book royalties from John Wiley. \r\nH.S. has received grant funding from NIH. \r\nS.C.S. received research support from Brain Canada, CIHR (Canadian Institutes of Health Research), Ontario Brain Institute, and CFI (Canadian Foundation for Innovation). S.C.S. is a founder and shareholder of ADMdx, Inc. \r\nD.T. has received grant funding from NIH. \r\nI.H.G. has received grant funding from NIH. \r\nM.H.T. received research support from NIH, PCORI, and AFSP, is a consultant for Alkermes Inc., Alto Neuroscience Inc, Axsome Therapeutics, Boegringer Ingelheim, GH Research, GreenLight VitalSign6 Inc, Heading Health, Inc., Janssen Pharmaceutical, Legion Health, Merck Sharp \u0026 Dohme Corp., Mind Medicine Inc., Navitor, Neurocrine Biosciences Inc., Noema Pharma AG, Orexo US Inc., Otsuka Canada Pharmaceutical Inc, Otsuka Pharmaceutical Development \u0026 Commercialization, Inc. (MDD section adviser), SAGE Therapeutics, Signant Health, and Takeda Pharmaceuticals Inc and receives editorial compensation from Oxford University Press. \r\nA.H.Y. reports the following conflicts of interest: paid lectures and advisory boards for the following companies: Astrazenaca, Eli Lilly, Lundbeck, Sunovion, Servier, Livanova, Janssen, Allegan, Bionomics, Sumitomo Dainippon Pharma, COMPASS, Sage, Novartis; consultant to Johnson \u0026 Johnson and to Livanova; received honoraria for attending advisory boards and presenting talks at meetings organized by LivaNova; principal investigator in the Restore-Life VNS registry study funded by LivaNova; UK chief investigator for Novartis MDD study MIJ821A12201; principal investigator on ESKETINTRD3004: ‘An Open-label, Long-term, Safety and Efficacy Study of Intranasal Esketamine in Treatment-resistant Depression’; principal investigator on ‘The Effects of Psilocybin on Cognitive Function in Healthy Participants’; principal investigator on ‘The Safety and Efficacy of Psilocybin in Participants with Treatment-Resistant Depression (P-TRD)’; no shareholdings in pharmaceutical companies; deputy editor of BJPsych Open. Grant funding (past and present): NIMH (USA); CIHR (Canada); NARSAD (USA); Stanley Medical Research Institute (USA); MRC (UK); Wellcome Trust (UK); Royal College of Physicians (Edin); BMA (UK); UBC-VGH Foundation (Canada); WEDC (Canada); CCS Depression Research Fund (Canada); MSFHR (Canada); NIHR (UK); Janssen (UK). \r\nR.Z. is a private psychiatrist service provider at The London Depression Institute and co-investigator on a Livanova-funded observational study of vagus nerve stimulation for depression. R.Z. has received honoraria for talks at medical symposia sponsored by Lundbeck as well as Janssen. He has collaborated with EMIS PLC and advises Depsee Ltd. He is affiliated with the D’Or Institute of Research and Education, Rio de Janeiro, and advises the Scients Institute, USA.\r\nAuthors A. Singh, A. Stolicyn, B.R.G., B.N.F., C.C.F., C.-G.Y., C.D., D.A., D.S., D.W., G.E., H.C.W., I.M.A., I.S., J.F.W.D., J.Q., J.S., J.A.G., K.H., K.S.C., K.Q., M.P.P., M.A., M.D.S., M.G., Q.G., R.E., R.D.W., S.H., S.R., T.A.V., T.C., V.G.F., Y.C., Y.F., and W.X declare that they have no competing interests.","formattedTitle":"Neuroanatomical dimensions in major depression: external validation and links with cognition, adverse life events, self-harm, metabolomics and genetics","fulltext":[{"header":"Introduction","content":"\u003cp\u003eMajor depressive disorder (MDD) is a mental health illness that is common, is a leading cause of disability worldwide, and the main precursor to suicide.\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e MDD is currently a clinical diagnosis, characterized by a persistent low mood and/or a diminished ability to experience usual feelings of pleasure, which is associated with a set of self-reported symptoms and clinical presentations, such as disrupted sleep, changes in appetite, low energy, feelings of guilt, and potential thoughts of death or suicide.\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e MDD is heterogeneous, clinical outcomes are variable, and there are no reliable predictors of treatment response at the individual level.\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e Approaches based on symptom profiles, comorbid disorders, or putative aetiology have yielded limited biological distinctions. MDD is considered a disorder at present, rather than a disease with a distinct neuropathological mechanisms. There are no neurobiological markers that can aid in the identification of MDD. Furthermore, it is likely that MDD is not a single disease but, instead, comprises multiple neurobiological mechanisms.\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eBy applying advances in artificial intelligence to neuroimaging data, it has been possible to identify MDD from healthy participants at the individual level\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e and further to delineate subtypes within MDD with larger samples consisting of hundreds of participants.\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e However, regional structure and activity in MDD are altered by treatment, including medications, psychotherapy, and neuromodulation, and regional activity can vary with changes in depressive state, such as from a current depressive episode to a remitted state, in which there are few or no symptoms.\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e In treatment resistant depression, a clinical term referring to MDD characterized by continued symptoms despite at least two courses of treatment, the neural correlates almost certainly reflect the effects of multiple treatments in addition to disease-related effects.\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e We recently sought to identify neural patterns at the individual level that characterize first-episode and recurrent MDD in participants who were medication-free and in a current depressive episode.\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e Based on structural MRI data, we found two dimensions that showed distinct treatment responses to antidepressant medication.\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e Dimension 1 (D1) was characterized by preserved grey matter (GM) and white matter (WM) volumes and showed a positive clinical response to selective serotonin reuptake inhibitor (SSRI) antidepressant medication but not to placebo medication. In contrast, Dimension 2 (D2) was characterized by reduced GM and WM volumes and demonstrated limited clinical response to either SSRI or placebo medication.\u003c/p\u003e \u003cp\u003eIn the present study, we sought to externally validate these dimensions in the general population as well as within a subset of the general population with current depressive symptoms. We used the UK Biobank, a large-scale population dataset of over 500,000 participants with extensive neuroimaging, behavioural, and genetic data,\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e,\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e to test the generalizability of these dimensions and to explore their associations with cognitive functioning, affective expressions, and genetic profiles. In the external validation analyses, we examined associations previously reported with cognitive functioning,\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e depressive symptomatology,\u003csup\u003e\u003cspan additionalcitationids=\"CR17\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e anxiety-related symptoms,\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e neuroticism-related traits,\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e history of self-harm and suicide attempts, adverse life events and lifestyle behaviours,\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e metabolic measures,\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e and genome-wide association studies (GWAS).\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e We sought to integrate validation of the neurobiological dimensions in an independent cohort from the general population with multidimensional biopsychosocial factors.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eMRI volumetric measures\u003c/p\u003e \u003cp\u003eIn UKB general population (n\u0026thinsp;=\u0026thinsp;37,235), external validation revealed the following classifications: D1 (n\u0026thinsp;=\u0026thinsp;6,931), D2 (n\u0026thinsp;=\u0026thinsp;10,262), combined D1 and D2 (n\u0026thinsp;=\u0026thinsp;2,931), neither D1 or D2 (n\u0026thinsp;=\u0026thinsp;12,009), margins of D1 or D2 (n\u0026thinsp;=\u0026thinsp;7,102). The largest structural MRI volumes were observed in D1, followed by combined D1 and D2, neither D1 or D2 classification, and D2 (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, Supplementary materials).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDemographic information for UKB general population and individuals with current depressive symptoms for total sample and dimensions\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNumber\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eFemale%\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003eGeneral population\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal sample\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e37,235 (100%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e53.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e64.14\u0026thinsp;\u0026plusmn;\u0026thinsp;7.50\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDimension 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6,931 (18.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e50.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e64.21\u0026thinsp;\u0026plusmn;\u0026thinsp;7.35\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDimension 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e10,262 (27.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e51.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e63.95\u0026thinsp;\u0026plusmn;\u0026thinsp;7.68\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCombined D1 and D2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e931 (2.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e44.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e64.42\u0026thinsp;\u0026plusmn;\u0026thinsp;7.54\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNeither D1 nor D2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12,009 (32.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e54.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e64.10\u0026thinsp;\u0026plusmn;\u0026thinsp;7.52\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMargin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e7,102 (19.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e55.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e64.38\u0026thinsp;\u0026plusmn;\u0026thinsp;7.51\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eCurrent depressive symptoms\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal sample\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1,454 (100%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e59.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e61.71\u0026thinsp;\u0026plusmn;\u0026thinsp;7.56\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDimension 1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e264 (18.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e58.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e61.84\u0026thinsp;\u0026plusmn;\u0026thinsp;7.84\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDimension 2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e442 (30.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e58.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e61.52\u0026thinsp;\u0026plusmn;\u0026thinsp;7.56\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCombined D1 and D2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e30 (2.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e56.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e60.55\u0026thinsp;\u0026plusmn;\u0026thinsp;7.44\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNeither D1 nor D2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e453 (31.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e61.6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e61.79\u0026thinsp;\u0026plusmn;\u0026thinsp;7.13\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMargin\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e265 (18.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e60.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e61.90\u0026thinsp;\u0026plusmn;\u0026thinsp;8.06\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"4\"\u003eNumber of participants in the sample is presented with percentages in parentheses indicating proportions relative to the total sample. The mean age of participants is presented in years with \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e\u0026plusmn;\u003c/span\u003e\u0026thinsp;standard deviation. General population, all participants in the UK biobank with MRI data; current depressive symptoms indicate participants from general population sample who have depression and are currently depressed at the time of their MRI imaging visit.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn UKB individuals in a current major depressive episode (MDE) (n\u0026thinsp;=\u0026thinsp;1,454), external validation revealed the following classifications: D1 (n\u0026thinsp;=\u0026thinsp;264), D2 (n\u0026thinsp;=\u0026thinsp;442), combined D1 and D2 (n\u0026thinsp;=\u0026thinsp;30), neither D1 or D2 (n\u0026thinsp;=\u0026thinsp;453), margins of D1 or D2 (n\u0026thinsp;=\u0026thinsp;265) (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Supplementary materials).\u003c/p\u003e \u003cp\u003eCognitive functioning\u003c/p\u003e \u003cp\u003eIn UKB general population, D2 showed significantly impaired performance as compared to D1 in all measures of cognitive functioning: fluid intelligence (\u003cem\u003eβ\u003c/em\u003e=\u0026minus;0.26, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;5.73E-45), executive function, taking longer to complete TMT-A numeric path (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.16, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.65E-13) and TMT-B alphanumeric path (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.12, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;5.01E-08), working memory as measured by backward digit span task (\u003cem\u003eβ\u003c/em\u003e=\u0026minus;0.18, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;7.76E-17), processing speed as measured by symbol-digit substitution (\u003cem\u003eβ\u003c/em\u003e=\u0026minus;0.14, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;4.26E-11), nonverbal reasoning as measured by matrix pattern completion (\u003cem\u003eβ\u003c/em\u003e=\u0026minus;0.25, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;8.33E-33), and verbal memory as measured by paired associate learning (\u003cem\u003eβ\u003c/em\u003e=\u0026minus;0.058, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.012).\u003c/p\u003e \u003cp\u003eIn UKB MDE cohort, D2 similarly showed significantly impaired performance as compared to D1 in: fluid intelligence (\u003cem\u003eβ\u003c/em\u003e=\u0026minus;0.25, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.004), working memory as measured by backward digit span task (\u003cem\u003eβ\u003c/em\u003e=\u0026minus;0.23, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.027), processing speed as measured by symbol-digit substitution (\u003cem\u003eβ\u003c/em\u003e=\u0026minus;0.19, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.049), and nonverbal reasoning measured by matrix pattern completion test (\u003cem\u003eβ\u003c/em\u003e=\u0026minus;0.23, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.033). There were trends towards significant differences in TMT-A (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.21, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.051) and TMT-B (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.20, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.053) in which D2 showed impaired performance relative to D1. There was no significant difference between groups in paired associate learning task (\u003cem\u003eβ\u003c/em\u003e=-1.56 \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.853).\u003c/p\u003e \u003cp\u003eDepressive symptoms\u003c/p\u003e \u003cp\u003eIn UKB general population, in items related to core depressive symptoms, D2 participants reported higher rates of 'prolonged loss of interest in normal activities' (D1: 37.04%, D2: 40.07%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;10.87, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.00098, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.030) and 'prolonged feelings of sadness or depression' (D1: 52.98%, D2: 55.53%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;7.35, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0067, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.025).\u003c/p\u003e \u003cp\u003eIn associated items, D2 individuals were more likely to report 'thoughts of death during the worst depression' (D1: 48.44%, D2: 51.56%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;7.07, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0078, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.0336) and \u0026lsquo;weight change during the worst episode of depression\u0026rsquo; D1: 56.71%, D2: 60.90%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;9.55, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0038, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.025). In the PHQ-9 measures, D2 individuals reported greater prevalence of 'recent poor appetite or overeating' (D1: 16.46%, D2: 18.28%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;6.43, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.011, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.023)\u003c/p\u003e \u003cp\u003eThere were no significant differences between dimensions in other symptoms, in 'feelings of worthlessness', 'difficulty concentrating', 'did your sleep change', 'feelings of tiredness during the worst period of depression', or in PHQ-9 item symptoms.\u003c/p\u003e \u003cp\u003eIn UKB MDE cohort, D2 participants reported higher endorsement of 'thoughts of death during the worst period of depression' (D1: 63.2%, D2: 77.2%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;6.14, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.013, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.145), and within PHQ-9 items, D2 participants more frequently reported \u0026lsquo;recent changes in speed/amount of moving or speaking\u0026rdquo; (D1: 18.1%, D2: 27.8%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;3.86, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.049, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.103). There were no significant differences in the remaining depressive symptom items.\u003c/p\u003e \u003cp\u003eAnxiety-related symptoms\u003c/p\u003e \u003cp\u003eIn UKB general population, D2 individuals were more likely to report 'recent inability to stop or control worrying' (D1: 21.32%, D2: 23.93%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;4.17,\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.041, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.018). There were no significant differences in the remaining anxiety symptoms.\u003c/p\u003e \u003cp\u003eIn UKB MDE cohort, there were no significant differences between D1 and D2 in anxiety symptoms.\u003c/p\u003e \u003cp\u003eNeuroticism-related symptoms\u003c/p\u003e \u003cp\u003eIn UKB general population, D2 individuals were more likely to report 'feeling tense/highly strung' (D1: 11.33%, D2: 12.74%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;7.34, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0067, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.021), 'nervous feelings' (D1: 16.54%, D2: 19.21%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;19.00, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.31E-05, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.034), and 'suffering from nerves' (D1: 13.83%, D2: 16.06%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;15.25, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;9.42E-05, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.030). There were no significant differences in any other items.\u003c/p\u003e \u003cp\u003eIn UKB MDE cohort, D2 participants more frequently reported \u0026lsquo;guilty feelings' (D1: 53.6%, D2: 64.9%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;6.34, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.012, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.109). There were no significant differences in other items.\u003c/p\u003e \u003cp\u003eSelf-harm and suicide attempts\u003c/p\u003e \u003cp\u003eIn UKB general population, D2 individuals showed a significantly greater endorsement of a history of suicide attempts (55.20%) compared to D1 (43.23%) (χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;7.68, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.006, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.113) as well as a history of self-harm (4.61%) compared to D1 (3.80%) (χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;4.41, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.036, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.019) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn UKB MDE cohort, D2 individuals similarly showed a significantly greater endorsement of a history of suicide attempts in (73.8%) as compared to D1 (45.8%) (χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;4.03, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.045, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.247), but there were no significant differences in history of self-harm.\u003c/p\u003e \u003cp\u003eAdverse life events and lifestyle behaviours\u003c/p\u003e \u003cp\u003eIn UKB general population, in items related to adverse life events, there was significantly greater endorsement in D2 compared to D1 in the following items: \"experiencing physical violence by a partner or ex-partner\" as an adult (D1 9.46%; D2 11.40%, χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;12.36, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0004, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.031), and \"being stopped from seeing friends or family by a partner or ex-partner\" as an adult (D1 7.38%, D2 8.37%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;4.11, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.043, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.018); and in childhood events: \"being physically abused by a family member\" (D1 17.69%, D2 19.73%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;8.39, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.004, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.025); and \"feeling hated by a family member\" (D1 15.64%, D2 18.09%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;13.12, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0003, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.032) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn UKB MDE cohort, there was significantly greater endorsement in D2 as compared to D1 in the following items: \"experiencing physical violence by a partner or ex-partner\" as an adult (D1 13.4%; D2 25.7%, χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;7.84, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.005, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.142), and \"being stopped from seeing friends or family by a partner or ex-partner\" as an adult (D1 13.6%, D2 23.0%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;4.78, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.029, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.112), and \"sexual intercourse by partner or ex-partner without consent\" as an adult (D1 9.0%, D2 16.5%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;3.92, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.048, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.101); and in childhood events: \"being physically abused by a family member\" (D1 21.7%, D2 38.5%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;11.52, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0007, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.172), and \"feeling hated by a family member\" (D1 32.1%, D2 44.3%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;5.32, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.021, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.118) (Supplementary Fig.\u0026nbsp;2).\u003c/p\u003e \u003cp\u003eIn lifestyle behaviours, in UKB general population, there was greater endorsement or previous or current \"smoking status\" in D2 (38.92%) compared to D1 (37.09%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;5.78, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.016, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.018) and in \u0026ldquo;salt added to food\u0026rdquo; in D2 (43.31%) compared to (D1 43.09%; χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;12.97, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0003, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.028), while there was greater endorsement of \"alcohol usually taken with meals\" in D1 (74.21%) compared to D2 (72.43%) (χ\u0026sup2;(1)\u0026thinsp;=\u0026thinsp;3.97, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.046, Cram\u0026eacute;r\u0026rsquo;s V\u0026thinsp;=\u0026thinsp;0.020). In UKB MDE cohort, there were no significant differences between D1 and D2.\u003c/p\u003e \u003cp\u003eMetabolomics\u003c/p\u003e \u003cp\u003eIn UKB general population, 39 (57.4%) out of the 68 metabolites tested showed a significant difference between D1 and D2 (FDR \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05). D1 showed significantly increased levels compared to D2 in 16 metabolites. These included 2 measures of intermediate density lipoprotein (IDL) particles, 9 measures of large to very large high-density lipoprotein (HDL) particles, forced expiratory volume and mean cell volume. Among fatty acids there was increased levels three fatty acids in their ratio to total fatty acids, including polyunsaturated fatty acids, omega-6 fatty acids and linoleic acid. D2 showed significantly increased levels compared to D1 in 23 metabolites. These include 13 large to extremely large, very low-density lipoprotein (VLDL) particles, systolic blood pressure, glycated haemoglobin, white blood cell count, C-reactive protein, and pyruvate. Among fatty acids there were significantly higher levels of total monounsaturated fatty acids (MUFA) and its ratio to total fatty acids (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e.).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn UKB MDE cohort, only forced expiratory volume (FEV1) was significantly higher in D1 compared to D2 (Supplementary Fig.\u0026nbsp;3). All measures are presented in Supplementary materials.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003ePhysical Measures\u003c/h2\u003e \u003cp\u003eIn UKB general population, D2 showed significantly higher levels compared to D1 in: leg fat percentage (right) (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.052, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;6.70E-08, \u003cem\u003eη\u0026sup2;\u003c/em\u003e = 0.052), leg fat percentage (left) (β\u0026thinsp;=\u0026thinsp;0.046, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;3.15E-07, η\u0026sup2; = 0.046), body fat percentage (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.050, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.000655, \u003cem\u003eη\u0026sup2;\u003c/em\u003e = 0.050), arm fat percentage (left) (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.059, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;7.40E-05, \u003cem\u003eη\u0026sup2;\u003c/em\u003e = 0.059), and arm fat percentage (right) (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.065, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.19E-05, \u003cem\u003eη\u0026sup2;\u003c/em\u003e = 0.065). In contrast, D1 showed significantly higher levels compared to D2 in: hand grip strength (right) (\u003cem\u003eβ\u003c/em\u003e = \u0026minus;0.125, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.01E-24, \u003cem\u003eη\u0026sup2;\u003c/em\u003e = \u0026minus;0.125), hand grip strength (left) (\u003cem\u003eβ\u003c/em\u003e = \u0026minus;0.115, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;3.19E-21, \u003cem\u003eη\u0026sup2;\u003c/em\u003e = \u0026minus;0.115), trunk fat-free mass (\u003cem\u003eβ\u003c/em\u003e = \u0026minus;0.207, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;1E-10, \u003cem\u003eη\u0026sup2;\u003c/em\u003e = \u0026minus;0.207), and trunk predicted mass (\u003cem\u003eβ\u003c/em\u003e = \u0026minus;0.204, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;1E-10, \u003cem\u003eη\u0026sup2;\u003c/em\u003e = \u0026minus;0.204) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn UKB MDE cohort, D2 similarly showed significantly higher levels compared to D1 in: arm fat percentage (right) (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.161, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.028, \u003cem\u003eη\u0026sup2;\u003c/em\u003e = 0.161), and arm fat percentage (left) (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.152, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.037, \u003cem\u003eη\u0026sup2;\u003c/em\u003e = 0.152), while D1 exhibited significantly higher levels compared to D2 in: trunk fat-free mass (\u003cem\u003eβ\u003c/em\u003e = \u0026minus;0.121, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.028, \u003cem\u003eη\u0026sup2;\u003c/em\u003e = \u0026minus;0.121) and trunk predicted mass (\u003cem\u003eβ\u003c/em\u003e = \u0026minus;0.120, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.028, \u003cem\u003eη\u0026sup2;\u003c/em\u003e = \u0026minus;0.120) (Supplementary Fig.\u0026nbsp;4).\u003c/p\u003e \u003cp\u003eGenetics\u003c/p\u003e \u003cp\u003eGWAS analysis identified 7 genomic loci significantly associated with D1 and 7 genomic loci significantly associated with D2 (significance level P-value\u0026thinsp;\u0026lt;\u0026thinsp;5x10\u003csup\u003e\u0026minus;\u0026thinsp;8\u003c/sup\u003e). 3 genomic loci associated with D1 and 3 associated with D2 were new as their top lead SNP had not associated with any clinical traits in the National Human Genome Research Institute and European Bioinformatics Institute (NHGRI-EBI) GWAS Catalog.\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e GWAS analysis identified 10 independent significant SNPs associated with D1 and 9 associated with D2 (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFive independent SNPs associated with D1 had previous reports of positive associations in NHGRI-EBI GWAS Catalog, indicating that carrying the risk allele has been linked with increased depressed affect,\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e schizophrenia,\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e brain volume in dorsolateral prefrontal, posterolateral temporal and superior temporal regions,\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e right hippocampal subfield CA3 (head) volume,\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e body fat percentage,\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e and acne.\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e Three independent significant SNPs associated with D1 had previous reports of negative associations with brain age (i.e. difference between predicted and chronological age),\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e brain volume in orbitofrontal, superior parietal, and superior temporal regions, precuneus, and cortical surface regions,\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e as well as body mass index (Huang et al., 2023). Four independent significant SNPs associated with D1 showed no previous associations in GWAS Catalog.\u003c/p\u003e \u003cp\u003eTwo independent SNPs associated with D2 had shown positive associations in NHGRI-EBI GWAS Catalog with vertex-wise sulcal depth,\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e and whole-body fat free mass (UKB data field 23101),\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e and non-glioblastoma glioma.\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e SNP rs9823492 has been associated with paired helical filament tau (PHF-tau), which is a biomarker for Alzheimer\u0026rsquo;s disease,\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e and SNP rs4843547 has been associated with white matter microstructure measures.\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e Five independent significant SNPs associated with D2 showed no previous associations GWAS Catalog.\u003c/p\u003e \u003cp\u003eMAGMA-based tissue expression analysis for D1 revealed two significant associations: in cerebellum (p\u0026thinsp;=\u0026thinsp;0.00030772) and cerebellar hemisphere (p\u0026thinsp;=\u0026thinsp;0.00034503).\u003c/p\u003e \u003cp\u003eResearch Domain Criteria (RDoC) associations\u003c/p\u003e \u003cp\u003eIn MDD participants from COORDINATE-MDD consortium,\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e individual item level baseline data were available in 426 participants from MADRS rating scale (CANBIND and Remedi, N\u0026thinsp;=\u0026thinsp;130 (D1 n\u0026thinsp;=\u0026thinsp;46, D2 n\u0026thinsp;=\u0026thinsp;84)) or the HAMD rating scale (Oxford and EMBARC, N\u0026thinsp;=\u0026thinsp;296 (D1 n\u0026thinsp;=\u0026thinsp;148, D2 n\u0026thinsp;=\u0026thinsp;148)). There were no significant differences between D1 and D2 in age or sex (D1: n\u0026thinsp;=\u0026thinsp;194 (118 female), mean age 36.02\u0026thinsp;\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e\u0026plusmn;\u003c/span\u003e\u0026thinsp;12.87 years; D2: n\u0026thinsp;=\u0026thinsp;232 (160 female), mean age 35.79\u0026thinsp;\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e\u0026plusmn;\u003c/span\u003e\u0026thinsp;12.57 years). In RDoC core depression domain, D1 showed significantly greater scores compared to D2 (D1 mean\u0026thinsp;=\u0026thinsp;0.594, D2 mean\u0026thinsp;=\u0026thinsp;0.556, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.037), while in sensorimotor systems domain, D2 showed significantly greater scores compared to D1 (D1 mean\u0026thinsp;=\u0026thinsp;0.233, D2 mean\u0026thinsp;=\u0026thinsp;0.315, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.005). Negative valance systems showed a trend towards significance with D2 exhibiting greater scores than D1 (D1 mean\u0026thinsp;=\u0026thinsp;0.395, D2 mean\u0026thinsp;=\u0026thinsp;0.423, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.052) (Supplementary Materials). There were no significant differences in positive valance systems, cognitive systems, arousal/regulatory systems, anxiety phenotype or neurovegetative symptoms of melancholia phenotype.\u003c/p\u003e \u003cp\u003eSensitivity analyses\u003c/p\u003e \u003cp\u003e Repeating the external validation in two UKB subgroups with overlapping ages from the COORDINATE-MDD sample: (1) having a lower age of 56 years and an upper age of 65 years (n\u0026thinsp;=\u0026thinsp;6523), and (2) having an age range of ten years of the youngest UKB participant (ages 45 to 55 years) (n\u0026thinsp;=\u0026thinsp;3112), D1 and D2 individuals were identified in both subgroups and showed a similar pattern of results in cognitive functioning, self-harm, adverse life events history, as well as metabolomics (Supplementary Tables\u0026nbsp;30 and 31).\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eNeuroimaging-based dimensions derived from first episode and recurrent MDD showed robust external validation in a large UKB general population cohort. D1 was characterised by relatively preserved grey and white matter volumes, while D2 showed reduced grey and white matter volumes that were associated with a pattern of cognitive impairments, increased adverse life events in adulthood and childhood, increased self-harm and suicide attempts, a pro-atherogenic lipid profile, higher body fat and systemic inflammation compared to D1. These patterns were observed both in a large cohort from the general population and in individuals who had core depressive symptoms at the time of their brain scan in UK Biobank.\u003c/p\u003e \u003cp\u003eA pattern of cognitive impairments was evident in D2 relative to D1 in a wide range of domains from executive function, working memory, and processing speed to nonverbal reasoning. Brain volume has been associated with general cognitive functioning in which cognitive decline is linked with reduced grey matter volume.\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e,\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e In the present analysis, we observed similar pattern of cognitive deficits in individuals who had endorsed current depressive symptoms at the time of their scan who were classified as D2 compared to D1. The findings suggest that the widespread subtle deficits in grey and white matter in D2 underlie the wide-ranging cognitive deficits observed in D2 compared to D1, which is observable in the general population, suggesting that dimension assignment is a trait feature as well as a state feature in individuals from the general population who had endorsed experiencing a current depressive episode.\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eIn the general population, D2 had a significantly greater history of self-harm and suicide attempts than did D1. In addition, trauma-related experiences, such as childhood abuse and exposure to interpersonal violence, were more prevalent among D2 than among D1 individuals. Cumulative grey matter reductions have been consistently observed in trauma-exposed individuals, suggesting a potential contribution to the regional deficits in the development of MDD in D2 individuals.\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eWith respect to metabolomics, D2 had a pro-atherogenic lipid profile, higher body fat, and impaired glucose metabolism, reflecting poorer metabolic health, while D1 had a healthier lipid metabolism. D2 was associated with higher levels of very low-density lipoproteins (VLDL), chylomicrons, triglycerides, and cholesterol as compared to D1. D2 also exhibited significantly higher body fat percentages, including leg and arm fat, as well as overall body fat. This was accompanied by reduced muscle mass, as evidenced by lower trunk fat-free mass and trunk predicted mass, and diminished hand grip strength on both sides as compared to D1. This profile is commonly associated with increased cardiovascular risk and metabolic dysregulation, potentially linked to metabolic syndrome, insulin resistance, and obesity.\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eHigh triglycerides have been associated with an increased risk of depression, anxiety, and stress-related disorders.\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e Furthermore, D2’s higher pyruvate levels, a key glycolysis metabolite, suggest disrupted glucose metabolism, consistent with observations that elevated glucose levels are associated with increased risk of psychiatric disorders\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e and Lv et al.\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e identified a relationship between increased body fat and higher risks of depression. Higher body fat percentages correlate with greater hormonal stress vulnerability (e.g., elevated cortisol levels), which in turn impairs cognitive performance under stress,\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e and is in itself associated with reduced regional brain volumes.\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e,\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e In contrast, D1 displayed a healthier lipid metabolism profile, with higher levels of high-density lipoproteins (HDL) and intermediate-density lipoproteins (IDL). Elevated HDL levels, particularly in large and very large HDL particles, are indicative of better cardiovascular protection and reduced systemic inflammation, which have been found to be protective against psychiatric disorders.\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eConsistent with the metabolic profile observed in D2, Lamers et al.\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e had identified an \"immuno-metabolic depression\" (IMD) profile which is characterized by increased inflammatory markers and metabolic dysregulation, such as dyslipidemia and higher body fat, and reduced physical fitness. The IMD profile was identified from the epidemiological population-based Netherlands Study of Depression and Anxiety (NESDA) and has been replicated in treatment-naïve adults with MDD.\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e Furthermore, longitudinal findings over six years observed a long-term persistence of high symptom burden in the IMD profile.\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e Our findings suggest that the limited effectiveness of SSRI antidepressant as well as placebo medications might be observed early in the course of illness in D2 dimension individual.\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e This highlights the need for additional treatment options to mitigate long term morbidity and improve clinical outcomes.\u003c/p\u003e \u003cp\u003eGWAS analyses revealed distinct genomic loci and independent significant SNPs that were uniquely associated with either D1 or D2 scores. For D1, a positive association was found with SNP \u003cem\u003ers6782581\u003c/em\u003e, which has a known negative association with BMI.\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e Grey matter volume is linked to BMI, with increased BMI (indicating being overweight or obesity) associated with reductions in grey matter, while normal BMI is linked with preserved grey matter.\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e,\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e However, \u003cem\u003ers6782581\u003c/em\u003e has also been associated with a slight increase in body fat percentage (0.01%),\u003csup\u003e27\u003c/sup\u003e which appears contradictory as this is also associated with reduced grey matter volume. Indicators of general obesity though, such as BMI or body-fat percentage, are less informative at predicting brain volumes than indicators of central obesity, such as increased waist-to-hip ratio (WHR), with increases associated with reduced grey matter volume.\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e,\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e Supplementary analysis indicated significantly larger WHR in D2 compared to D1 in the UKBB general population. For D2, GWAS identified a positive association with SNP \u003cem\u003ers1926034\u003c/em\u003e with previous studies linking this SNP to increased whole-body-fat free mass (all mass in the body that is not fat).\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e D2 is characterized by reduced grey matter volume compared to D1,\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e and Pflanz et al.\u003csup\u003e\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e found an that higher whole-body fat-free mass was associated with lower grey matter volume and total brain volume. The findings suggest that variations in genetic associations and body composition metrics contribute to the distinct profiles of D1 and D2.\u003c/p\u003e \u003cp\u003eD1 score demonstrated a significant negative association with SNP \u003cem\u003ers199505\u003c/em\u003e, which has been linked to an increased likelihood of experiencing a depressed affect.\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e Within the RDoC framework, D1 was associated with significantly higher core depression scores and lower sensorimotor system scores compared to D2. This pattern highlights the persistence of core symptoms in the D1 profile, potentially linking its neurobiological features to distinct genetic and phenotypic domains.\u003c/p\u003e \u003cp\u003eD2 score showed a negative association with \u003cem\u003ers12076373\u003c/em\u003e, a variant linked to vertex-wide sulcal depth.\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e Increased sulcal depth reflects enhanced cortical folding, which has been positively associated with cognitive ability and cortical volume.\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e Conversely, reduced cortical folding has been observed in MDD.\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e Cortical folding is closely tied with genetic processes\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e and is predominantly completed before birth,\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e with sulcal patterns present at birth serving as predictors of neurobehavioral outcomes.\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e The negative association with D2 suggests a predisposition to reduced cortical folding, aligning with the reduced cortical volumes observed in D2 individuals.\u003c/p\u003e \u003cp\u003eWhile genes direct brain development, environmental experiences shape the emerging structures and their relationships with each other. The greatest increase in brain weight occurs from birth to 3 years and there is a further five fold increase from 3 to 18 years.\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e However, childhood maltreatment has a pronounced impact on brain development, from grey matter deficits and cortical thinning in anterior cingulate/paracingulate and middle frontal regions\u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e to widespread microstructural white matter reductions, particularly in fornix, corpus callosum and optic radiations as well as altered brain activity in the default mode and central executive network and increased responsivity in the amygdala and anterior cingulate to socioaffective cues.\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e,\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e Childhood maltreatment refers to physical, sexual or emotional abuse or neglect and is associated with deleterious effects on cognitive and emotional functioning and increased risk of mental health disorders.\u003csup\u003e\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e Structural deficits and functional alterations linked with childhood maltreatment are a potential vulnerability which might then interacts with stressful life events in adulthood leading to the development of MDD.\u003csup\u003e\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e Our findings suggest a potential pathophysiological mechanism for D2 MDD whereby polymorphisms associated with reduced cortical folding reflect a predisposition to reduced cortical volumes that is exacerbated by childhood maltreatment, leading to cognitive impairments and an immuno-metabolic form of MDD in adulthood which is associated with increased rates of self-harm as observed in D2 individuals in UKB general population,.\u003c/p\u003e \u003cp\u003eLimitations of the present study include the binary coding of variables and reliance on self-reported experiences that may have introduced potential biases, including recall inaccuracies or social desirability effects. Additionally, in the demographic characteristics of the UK Biobank cohort, participants tend to have relatively high socioeconomic status and limited ethnic diversity, which does not fully reflect the general population.\u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e The UK Biobank cohort was an older age group (mean 64 years) than our COORDINATE-MDD cohort (mean 38 years) which may have introduced a domain shift (Fu et al, 2024). However, sensitivity analysis stratified by age within the UK Biobank cohort showed comparable findings in subgroups with overlapping age ranges (Supplementary Materials). Although D1 and D2 dimensions were validated in the UK Biobank cohort, about 30% of UK Biobank participants had not been classified to either dimension, suggesting the possibility of additional dimensions. Furthermore, while functional connectivity measures have been associated with subtype classifications in MDD, the present analysis was limited to structural neuroimaging, which reduces our ability to fully capture the dynamic and network-level alterations in MDD. Integrating structural and functional measures has the potential to increase the precision in delineation of the biomarkers that comprise MDD.\u003c/p\u003e \u003cp\u003eTwo neuroanatomical dimensions were identified in medication-free individuals with first episode and recurrent MDD during a current depressive episode, in which D2 was characterized by widespread subtle deficits and showed limited treatment response to SSRI or placebo medication compared to D1. The present findings demonstrated strong generalizability in an independent general population cohort. D2 showed further associations with poorer cognitive functioning, greater exposure to adverse life events, including childhood trauma, an increased history of self-harm and suicide attempts, a pro-atherogenic lipid profile, and genetic links to white matter microstructure and neurodegenerative traits. The findings suggest that D1 and D2 represent biologically meaningful dimensions underlying MDD, associated with distinct neurobiological mechanisms and treatment responses. This highlights the shared and distinct neurobiological mechanisms of these dimensions that are currently categorised under the same clinical diagnosis. External validation demonstrates their potential as biomarkers to elucidate the heterogeneity within the current clinical MDD diagnosis that can aid in predicting treatment response and guiding treatment approaches.\u003c/p\u003e "},{"header":"Methods","content":"\u003cp\u003e We adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines, as detailed in the Supplementary Materials.\u003c/p\u003e\u003cp\u003eParticipants\u003c/p\u003e\u003cp\u003eIn the COORDINATE-MDD consortium, raw MRI data were shared from international samples (n = 1,384) in medication-free first episode and recurrent MDD (n = 685), all in a current depressive episode of at least moderate severity, and healthy controls (n = 699). Prospective longitudinal data on treatment response were available for a MDD subset (n = 359). Treatments were either SSRI antidepressant medication (escitalopram, citalopram, sertraline) or placebo (COORDINATE-MDD consortium).\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003cp\u003eThe UK Biobank cohort represents a general population with recruitment taking place from 2006 to 2010 in 22 assessment centres in England, Wales and Scotland, with an age range of 40–69 years and total sample size of 500,000 participants.\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e All participants provided informed written consent, and ethical approval for the study was granted by the National Research Ethics Service Committee North West–Haydock (reference 11/NW/0382). Participants provided sociodemographic, cognitive, and medical information through questionnaires and physical assessments. A subset of participants completed magnetic resonance imaging (MRI), beginning in 2014 (UK Biobank Brain Imaging Documentation; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.ukbiobank.ac.uk\u003c/span\u003e\u003cspan address=\"http://www.ukbiobank.ac.uk\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Imaging was conducted on a 3T Siemens Skyra scanner with a T1-weighted MPRAGE protocol, resolution of 1 x 1 x 1 mm and a time to echo (TE) of 2000 ms.\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e UK Biobank MRI data in the present study were acquired from 2014 to 2019 (n = 37,235 (19,736 women (53.0%)); mean age 64.14 years (SD = 7.50)).\u003c/p\u003e\u003cp\u003eWithin the UKB general population cohort, a subset of individuals with current depressive symptoms indicating a major depressive episode (MDE) was identified using the following two UKB data fields: item 2050 (\"frequency of depressed mood over the past two weeks\") and item 2060 (\"frequency of unenthusiasm or disinterest over the past two weeks\"), with endorsement of \"More than half the days\" or \"Nearly every day\" for either criterion. Exclusion criteria included comorbid psychiatric disorders (e.g., schizophrenia, bipolar disorder, psychotic symptoms), neurological (e.g., Parkinson’s disease, epilepsy), and medical (e.g., diabetes, hypertension) disorders from ICD-10 diagnostic codes (UKB data field: 41202) (Supplementary Materials).\u003c/p\u003e\u003cp\u003eImage preprocessing and harmonization\u003c/p\u003e\u003cp\u003eAll raw T1-weighted MRI data were first manually assessed for head motion, image artifacts or restricted field-of-view for quality assurance. Images were corrected for magnetic field inhomogeneity, and a multi-atlas segmentation (MUSE) was applied to obtain brain volumes in 145 anatomical regions of interest (ROI).\u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e,\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u003c/sup\u003e ROI volumes were linearly harmonized to correct for sex and deep learning-based intracranial volume (DLICV), an ICV estimation method that uses deep learning algorithms to harmonize variability in volumetric measurements across MRI scan, and non-linearly harmonized to correct for age-wise differences.\u003csup\u003e\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u003c/sup\u003e Harmonization was performed independently in the test dataset (UK Biobank) to prevent information leakage. For voxel-based analyses, T1-weighted images were transformed to the Regional Analysis of Volumes Examined in Normalized Space (RAVENS) maps for grey and white matter, which preserves total tissue volumes following normalization to ensure that regional differences are retained following registration and warping, enhancing their precision and reliability in assessing regional volumes.\u003csup\u003e\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e,\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003cp\u003eGeneralization with external validation in UK Biobank\u003c/p\u003e\u003cp\u003eThe semi-supervised Heterogeneity through Discriminative Analysis (HYDRA) model had been implemented to identify the optimal \u003cem\u003ek-\u003c/em\u003edimensional space for delineating the neuroanatomical heterogeneity in individuals with first episode and recurrent MDD, all of whom were unmedicated and experiencing a current depressive episode.\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e Determination of the optimal number of clusters (\u003cem\u003ek\u003c/em\u003e) was based on the highest Adjusted Rand Index (ARI), a metric used to evaluate clustering solutions by quantifying the agreement between partitions while remaining relatively insensitive to the variations in the number of clusters. Based on ARI, \u003cem\u003ek\u003c/em\u003e = 2 was computed as the optimal number of clusters, identifying two neuroanatomical dimensions which captured the heterogeneity in MDD neuroanatomy. The final HYDRA model was then applied to compute the MDD dimension scores in an independent external cohort derived from UKB.\u003c/p\u003e\u003cp\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"580\" height=\"105\"\u003e\u003c/p\u003e\u003cp\u003eThis angle ranges from 0° to 180°. As the dimensionality of the two axes increases, there is a greater likelihood for the angle to fall closer to 90°. 145 brain ROIs were used as training features in the HYDRA models, making each dimension axis 145-dimensional.\u003c/p\u003e\u003cp\u003eFor external validation, the HYDRA model was applied to the UK Biobank general population. Dimension membership (D1, D2) and expression scores (E1, E2) of the \u003cem\u003ek = 2\u003c/em\u003e dimensions for each participant were derived: 1) Dimension 1 (D1) was designated from E1 \u0026gt; = 0.3 and E2 \u0026lt;= -0.3; 2) Dimension 2 (D2) was E2 \u0026gt; = 0.3 and E1 \u0026lt;= -0.3; 3) combined D1 and D2 was E1 \u0026gt; = 0.3 and E2 \u0026gt; = 0.3; and 4) neither D1 or D2 was E1 \u0026lt;= -0.3 and E2 \u0026lt;= -0.3 (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eTo evaluate differences in brain volumetric measures in dimension membership, a post-hoc analysis of MUSE features was conducted using one-way ANOVA for groups: 1) D1; 2) D2; 3) combined D1 and D2; and 4) neither D1 or D2, with False Discovery Rate (FDR) correction to account for multiple comparisons with significance set at an FDR-adjusted threshold of \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05. For each MUSE feature, the group means, and proportion of variance explained by the group differences was calculated.\u003c/p\u003e\u003cp\u003eAssociations with cognitive functioning, depressive and anxiety symptoms, neuroticism-related traits, adverse life events, self-harm and suicide, lifestyle items and metabolomics were assessed for the dimensions (D1, D2). For each item, variables were coded as binary (1 = \"Yes,\" 0 = \"No\"), with missing data or responses excluded from analysis. Full descriptions of the variables and their classifications are presented in Supplementary Materials. Chi-squared tests were applied to identify significant differences in response patterns between groups. Standardised residuals and effect sizes (Cramér’s V) were computed to specify response patterns driving significant differences. A Cramér’s V of less than 0.2 indicates a weak effect, between 0.2 and 0.6 indicates a moderate effect, and greater than 0.6 indicates a strong effect.\u003c/p\u003e\u003cp\u003eCognitive functioning\u003c/p\u003e\u003cp\u003eCognitive functioning was assessed in the following seven domains:\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e,\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e\u003c/sup\u003e executive function (trail-making test (TMT) A and B); fluid intelligence (verbal and numerical reasoning); working memory (backward digit span); verbal memory (paired associate learning); complex processing speed (symbol-digit substitution test) and nonverbal reasoning (matrix pattern completion). Poorer cognitive performance is indicated by higher scores in TMT, indicating increased time for completion, and lower scores in all other tests (Full descriptions are presented in Supplementary Materials).\u003c/p\u003e\u003cp\u003eStructured analysis was applied to examine differences in cognitive functioning between D1 and D2. For each cognitive test, a general linear model adjusted for age, sex, and dimension membership (D1, D2) was assessed. False Discovery Rate (FDR) correction was applied, setting significance at an FDR-adjusted threshold of \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05. Outliers were addressed by calculating z-scores based on middle 80% of data values, thereby establishing robust mean and standard deviation estimates. Extreme outliers having z-score exceeding 5 were excluded, and z-scores were recalculated for accuracy. All dependent variables were standardized. The resulting beta coefficients represent the standardized strength of associations. Positive beta coefficients indicate higher scores for D2 relative to D1, while negative coefficients indicate lower scores for D2 relative to D1. Partial eta squared (η²) measure effect size for the variance explained by specific factors in ANOVA.\u003c/p\u003e\u003cp\u003eDepressive and anxiety symptoms, neuroticism-related traits\u003c/p\u003e\u003cp\u003eDepressive symptom items were selected based on Howard et al.:\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e endorsement of at least one core symptom experienced for two weeks or more, namely persistent sadness or loss of interest. Additional six depressive symptoms reflecting functional impairments during the worst period of depression were included: tiredness, changes in sleep patterns, difficulty concentrating, feelings of worthlessness, thoughts of death, and weight changes.\u003c/p\u003e\u003cp\u003eFollowing Thorp et al.,\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e depressive symptoms were also selected from endorsement of the following items based on Patient Health Questionnaire-9 (PHQ-9):\u003csup\u003e\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e\u003c/sup\u003e recent lack of interest or pleasure in doing things, recent poor appetite or overeating, recent trouble concentrating on things, recent feelings of depression, recent feelings of tiredness or low energy, recent feelings of inadequacy, recent changes in speed or amount of moving or speaking, trouble falling or staying asleep or sleeping too much, and recent thoughts of suicide or self-harm.\u003c/p\u003e\u003cp\u003eAnxiety-related symptoms were selected based on Thorp et al. \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e in the seven items: feeling nervous or anxious, feeling of foreboding, easy annoyance or irritability, restlessness, trouble relaxing, worrying too much about different things, inability to stop worrying over the last two weeks. Items are based on Generalised Anxiety Disorder-7 scale (GAD-7).\u003csup\u003e\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\u003cp\u003eNeuroticism-related traits were based on endorsement of twelve items following Okbay et al.,\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e which had been assessed from the 12-item Eysenck Personality Inventory framework:\u003csup\u003e\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e\u003c/sup\u003e mood swings, miserableness, irritability, sensitivity or hurt feelings, fed-up feelings, nervous feelings, worrier or anxious feelings, tense or highly strung, worrying too long after embarrassment, suffering from nerves, loneliness or isolation, and guilt feelings.\u003c/p\u003e\u003cp\u003eAdverse life events, self-harm and suicide, lifestyle items\u003c/p\u003e\u003cp\u003eAdverse life events were based on data fields assessing five items reflecting experiences of violence related to physical or sexual as an adult and three items related to childhood experiences of physical abuse, sexual molestation and feeling hated by a family member as a child. Self-harm and suicide attempts were assessed in individual items. Lifestyle items which had shown a significant association with suicide attempts in Zhang et al.\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e were selected: ever smoked, current smoking, alcohol use, age of first sexual intercourse, number of sexual partners, and sleep disturbances.\u003c/p\u003e\u003cp\u003eMetabolomics\u003c/p\u003e\u003cp\u003eA total of 68 measures were examined, reflecting respiratory health, cardiovascular health, metabolic processes, lipid profiles, blood cell counts, lipid metabolism, and inflammation, which have been associated with depression,\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e biological age\u003csup\u003e\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e\u003c/sup\u003e and metabolic syndrome.\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e Each measure was standardized, allowing beta coefficients to reflect standardized effect sizes. Positive beta coefficients indicate higher scores for D2 relative to D1, while negative coefficients indicate lower scores for D2 relative to D1. Full description of all measures is presented in Supplementary Materials.\u003c/p\u003e\u003cp\u003eGenome-wide association studies (GWAS) analysis\u003c/p\u003e\u003cp\u003eImputed genetic data were downloaded from UK Biobank (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ukbiobank.ac.uk/enable-your-research/about-our-data/genetic-data\u003c/span\u003e\u003cspan address=\"https://www.ukbiobank.ac.uk/enable-your-research/about-our-data/genetic-data\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) in July 2021.\u003csup\u003e\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e\u003c/sup\u003e Genotyped and imputed single nucleotide polymorphisms (SNP) (or single nucleotide variations (SNV)) were pre-processed following a quality check protocol. We extracted and excluded participants with mismatched genetically identified sex and self-reported sex; chromosome aneuploidy; and related individuals (2nd-degree related individuals) via King software relationship inference.\u003csup\u003e\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e\u003c/sup\u003e Duplicate variants, variants with minor allele frequency less than 1%, variants with missing rate higher than 3%, variants which failed the Hardy-Weinberg test at threshold \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:p\u0026lt;1\\times\\:{10}^{-10}\\)\u003c/span\u003e\u003c/span\u003e. Participants with more than 3% missing genotypes were excluded. To adjust for population stratification, the first 40 principal components (PC) were derived using PLINK 2 (v2.0.0).\u003csup\u003e\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e\u003c/sup\u003e 30,376 samples and 6,288,959 variants for GWAS analysis in UKB participants with European ancestry were filtered.\u003c/p\u003e\u003cp\u003eWe performed linear regression for D1 and D2 dimensional scores respectively on autosome variants. Age at imaging scan, sex, ICV and first 40 PC components were included as covariates. After calculating the association analysis via PLINK2, Functional Mapping and Annotation (FUMA) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://fuma.ctglab.nl/tutorial#snp2gene\u003c/span\u003e\u003cspan address=\"https://fuma.ctglab.nl/tutorial#snp2gene\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) identified the significant independent SNPs having genome-wide significant threshold with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:p\\le\\:5\\times\\:{10}^{-8}\\)\u003c/span\u003e\u003c/span\u003e and are independent of each other at \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{r}^{2}\u0026lt;0.6\\)\u003c/span\u003e\u003c/span\u003e. Each candidate SNP was queried in the GWAS Catalog to check for any published associations with previous GWAS studies.\u003c/p\u003e\u003cp\u003eResearch Domain Criteria\u003c/p\u003e\u003cp\u003eThe Research Domain Criteria (RDoC) framework aims to understand mental illness according to domains which exemplify types of neurobiological functioning based on biological underpinnings,\u003csup\u003e\u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e\u003c/sup\u003e consisting of six major functional domains: negative valence systems (NVS), positive valence systems (PVS), cognitive systems (CS), arousal / regulatory systems (ARS), sensorimotor systems (SS), and social processes (SP). Citrome et al.\u003csup\u003e\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e\u003c/sup\u003e identified HAMD and MADRS items which align with five RDoC domains (NVS, PVS, CS, ARS, SS), and Ahmed et al.\u003csup\u003e\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e\u003c/sup\u003e defined three MDD phenotypes (core depression (CD), anxiety (ANX), and neurovegetative symptoms of melancholia) based on HAMD items to represent RDoC constructs, calculating a threshold for scoring across items to determine if an individual would be classified as positive or negative for each phenotype. We applied these items,\u003csup\u003e\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e,\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e\u003c/sup\u003e to examine RDoC domain phenotype in the present D1 and D2 classification.\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e Total scores for all items within each phenotype were transformed into a percentage score of the total possible score to enable standardised comparison between scores assessed using MADRS and HAMD. Additionally, we identified corresponding items on the MADRS scale to measure the three phenotypes that were measured using HAMD items by Ahmed et al.\u003csup\u003e\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e\u003c/sup\u003e and created comparative thresholds to determine positive or negative phenotype classification (Supplementary Table\u0026nbsp;10). Phenotype scores between D1 and D2 groups were analysed by general linear modelling ANOVA. Proportion of participants who were phenotype positive or negative at baseline was analysed by Chi-square test. P-values were corrected using FDR.\u003c/p\u003e\u003cp\u003eSensitivity analyses\u003c/p\u003e\u003cp\u003eWe repeated the external validation and phenotype associations in UKB subgroups which overlapped with the COORDINATE-MDD sample age range:\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e (1) having a lower age of 56 years and an upper age of 65 years (n = 6523), and (2) having an age range of ten years of the youngest UKB participant (ages 45 to 55 years) (n = 3112) (Supplementary Tables\u0026nbsp;30 and 31).\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData availability \u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUK Biobank data is accessible only to approved researchers through the UK Biobank Research Analysis Platform (UKB-RAP) upon application and approval by the UK Biobank Access Management Committee. CAN-BIND data are available from https://www.braincode.ca/; EMBARC data are available from https://nda.nih.gov/edit_collection.html?id=2199; original data are available from individual co-authors, and the derived data are available on reasonable request to corresponding authors C.H.Y.F. and C.D.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCode availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe MUSE algorithm for image segmentation is available at https://www.nitrc.org/projects/cbica_muse. The HYDRA algorithm is available at https://github.com/evarol/HYDRA. The MIDAS algorithm is available at https://github.com/evarol/MIDAS. The following R packages were used: WebPower 0.8.6 (https://cran.r-project.org/web/packages/WebPower/WebPower.pdf), effectsize 0.8.2, and ggplot2 3.4.0. (https://cran.r-project.org/web/packages/ggplot2/ggplot2.pdf).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eS.R.A. has consulted for Indoc Research Canada. \u003c/p\u003e\n\u003cp\u003eB.W.D. has received research support from Boehringer-Ingelheim, Compass, Pathways, NIMH, Otsuka, and Uson and honoraria for consulting from Aya Biosciences, Myriad Neuroscience, Otsuka, Sophren Therapeutics, Cerebral Therapeutics, Sage. \u003c/p\u003e\n\u003cp\u003eC.H.Y.F. has received grant funding from Brain and Behavior (NARSAD), Eli Lilly and Co. Milken Institute, Flow Neuroscience, MRC, NIMH, Rosetrees Trust, and Wellcome Trust and is Section Editor of Brain Research Bulletin. \u003c/p\u003e\n\u003cp\u003eC.J.H. serves as a consultant for P1vital, Lundbeck, Servier, and Compass Pathways. She holds grant income from Zogenix and J\u0026amp;J.\u003c/p\u003e\n\u003cp\u003eS.H.K. has received funding for consulting or speaking engagements from Abbvie, Boehringer-Ingelheim, Janssen, Lundbeck, Lundbeck Institute, Merck, Otsuka Pfizer, Sunovion, and Servier. He has received research support from Abbott, Brain Canada, CIHR (Canadian Institutes of Health Research), Janssen, Lundbeck, Ontario Brain Institute, Otsuka, Pfizer, and SPOR (Canada’s Strategy for Patient-Oriented Research). He has stock/stock options in Field Trip Health. \u003c/p\u003e\n\u003cp\u003eG.M.K. has served as a speaker for Angelini, Abbvie, Cybin, and H. Lundbeck and as an advisor for Sanos, Onsero, Pangea Botanica, Gilgamesh, and Seaport Therapeutics.\u003c/p\u003e\n\u003cp\u003eH.S.M. has received grant funding from NIH, grant support from Wellcome Leap and Hope for Depression Research Foundation, and consulting and IP licenses fees from Abbott Labs. \u003c/p\u003e\n\u003cp\u003eA.M.M. has received research support from Eli Lilly, Janssen, and The Sackler Trust. A.M.M. has also received speaker fees from Illumina and Janssen. \u003c/p\u003e\n\u003cp\u003eW.E.C. serves on the National Advisory Board of the George West Mental health Foundation, as a board member of Hugarheill ehf (an Icelandic company dedicated to the prevention of depression), and on the scientific advisory boards of AIM for Mental Health and the Anxiety and Depression Association of America; he is supported by the Mary and John Brock Foundation, the Pitts Foundation, and the Fuqua family foundations, and he receives book royalties from John Wiley. \u003c/p\u003e\n\u003cp\u003eH.S. has received grant funding from NIH. \u003c/p\u003e\n\u003cp\u003eS.C.S. received research support from Brain Canada, CIHR (Canadian Institutes of Health Research), Ontario Brain Institute, and CFI (Canadian Foundation for Innovation). S.C.S. is a founder and shareholder of ADMdx, Inc. \u003c/p\u003e\n\u003cp\u003eD.T. has received grant funding from NIH. \u003c/p\u003e\n\u003cp\u003eI.H.G. has received grant funding from NIH. \u003c/p\u003e\n\u003cp\u003eM.H.T. received research support from NIH, PCORI, and AFSP, is a consultant for Alkermes Inc., Alto Neuroscience Inc, Axsome Therapeutics, Boegringer Ingelheim, GH Research, GreenLight VitalSign6 Inc, Heading Health, Inc., Janssen Pharmaceutical, Legion Health, Merck Sharp \u0026amp; Dohme Corp., Mind Medicine Inc., Navitor, Neurocrine Biosciences Inc., Noema Pharma AG, Orexo US Inc., Otsuka Canada Pharmaceutical Inc, Otsuka Pharmaceutical Development \u0026amp; Commercialization, Inc. (MDD section adviser), SAGE Therapeutics, Signant Health, and Takeda Pharmaceuticals Inc and receives editorial compensation from Oxford University Press. \u003c/p\u003e\n\u003cp\u003eA.H.Y. reports the following conflicts of interest: paid lectures and advisory boards for the following companies: Astrazenaca, Eli Lilly, Lundbeck, Sunovion, Servier, Livanova, Janssen, Allegan, Bionomics, Sumitomo Dainippon Pharma, COMPASS, Sage, Novartis; consultant to Johnson \u0026amp; Johnson and to Livanova; received honoraria for attending advisory boards and presenting talks at meetings organized by LivaNova; principal investigator in the Restore-Life VNS registry study funded by LivaNova; UK chief investigator for Novartis MDD study MIJ821A12201; principal investigator on ESKETINTRD3004: ‘An Open-label, Long-term, Safety and Efficacy Study of Intranasal Esketamine in Treatment-resistant Depression’; principal investigator on ‘The Effects of Psilocybin on Cognitive Function in Healthy Participants’; principal investigator on ‘The Safety and Efficacy of Psilocybin in Participants with Treatment-Resistant Depression (P-TRD)’; no shareholdings in pharmaceutical companies; deputy editor of BJPsych Open. Grant funding (past and present): NIMH (USA); CIHR (Canada); NARSAD (USA); Stanley Medical Research Institute (USA); MRC (UK); Wellcome Trust (UK); Royal College of Physicians (Edin); BMA (UK); UBC-VGH Foundation (Canada); WEDC (Canada); CCS Depression Research Fund (Canada); MSFHR (Canada); NIHR (UK); Janssen (UK). \u003c/p\u003e\n\u003cp\u003eR.Z. is a private psychiatrist service provider at The London Depression Institute and co-investigator on a Livanova-funded observational study of vagus nerve stimulation for depression. R.Z. has received honoraria for talks at medical symposia sponsored by Lundbeck as well as Janssen. He has collaborated with EMIS PLC and advises Depsee Ltd. He is affiliated with the D’Or Institute of Research and Education, Rio de Janeiro, and advises the Scients Institute, USA.\u003c/p\u003e\n\u003cp\u003eAuthors A. Singh, A. Stolicyn, B.R.G., B.N.F., C.C.F., C.-G.Y., C.D., D.A., D.S., D.W., G.E., H.C.W., I.M.A., I.S., J.F.W.D., J.Q., J.S., J.A.G., K.H., K.S.C., K.Q., M.P.P., M.A., M.D.S., M.G., Q.G., R.E., R.D.W., S.H., S.R., T.A.V., T.C., V.G.F., Y.C., Y.F., and W.X declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e \u003c/p\u003e\n\u003cp\u003eC.H.Y.F. and C.D. led the project. \u003c/p\u003e\n\u003cp\u003eC.H.Y.F., G.E., W.X., and C.D. were responsible for the study concept and the design of the study\u003cstrong\u003e.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eB.W.D., C.C.F., S.H.K., M.H.T., T.A.V., M.P.P., D.A., I.M.A., B.R.G., H.S.M., W.E.C., Q.G., M.D.S., I.H.G., A.M.M., A. Stolicyn, and H.C.W. contributed to the data acquisition. \u003c/p\u003e\n\u003cp\u003eW.X., R.D.W., Y.C., M.A., and D.S. conducted the data analysis and created the figures. \u003c/p\u003e\n\u003cp\u003eJ.W., H.S., A. Singh, C.H.Y.F., and C.D. supervised the statistical analysis. \u003c/p\u003e\n\u003cp\u003eC.H.Y.F. and C.D. interpreted the data. \u003c/p\u003e\n\u003cp\u003eJ.W., Y.F., D.A., S.H., A.M.M., R.Z., H.S.M., and C.C.F. provided crucial advice for the study. \u003c/p\u003e\n\u003cp\u003eW.X, R.D.W., Y.C. and C.H.Y.F. wrote the manuscript. \u003c/p\u003e\n\u003cp\u003eS.R.A., T.C., K.S.C., B.N.F., V.G.F., M.G., B.R.G., K.H., K.Q., S.R., M.D.S., J.S., J.A.G., A. Stolicyn, I.S., S.C.S., D.T., T.A.V., D.W., I.M.A., W.E.C., J.F.W.D., B.W.D., R.E., Q.G., I.H.G., C.J.H., S.H.K., G.M.K., H.S.M., M.P.P., J.Q., M.H.T., H.C.W., C.-G.Y., and A.H.Y. made substantial contributions to the manuscript and provided critical comments. \u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eJames, S. L. \u003cem\u003eet al.\u003c/em\u003e Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990\u0026ndash;2017: a systematic analysis for the Global Burden of Disease Study 2017. \u003cem\u003eThe Lancet\u003c/em\u003e\u003cstrong\u003e392\u003c/strong\u003e, 1789\u0026ndash;1858 (2018). https://doi.org/10.1016/S0140-6736(18)32279-7\u003c/li\u003e\n\u003cli\u003eAmerican Psychiatric Association. \u003cem\u003eDiagnostic and Statistical Manual of Mental Disorders, (DSM-5\u0026reg;)\u003c/em\u003e. (American Psychiatric Association, 2013).\u003c/li\u003e\n\u003cli\u003eFu, C. H. Y., Fan, Y. \u0026amp; Davatzikos, C. Addressing heterogeneity (and homogeneity) in treatment mechanisms in depression and the potential to develop diagnostic and predictive biomarkers. \u003cem\u003eNeuroImage Clin.\u003c/em\u003e\u003cstrong\u003e24\u003c/strong\u003e, 101997 (2019). https://doi.org/10.1016/j.nicl.2019.101997\u003c/li\u003e\n\u003cli\u003eFu, C. H. Y. \u0026amp; Costafreda, S. G. Neuroimaging-based biomarkers in psychiatry: Clinical opportunities of a paradigm shift. \u003cem\u003eCan. J. Psychiatry\u003c/em\u003e\u003cstrong\u003e58\u003c/strong\u003e, 499\u0026ndash;508 (2013). https://doi.org/10.1177/070674371305800904\u003c/li\u003e\n\u003cli\u003eFu, C. H. Y. \u003cem\u003eet al.\u003c/em\u003e AI-based dimensional neuroimaging system for characterizing heterogeneity in brain structure and function in major depressive disorder: COORDINATE-MDD consortium design and rationale. \u003cem\u003eBMC Psychiatry\u003c/em\u003e\u003cstrong\u003e23\u003c/strong\u003e, 59 (2023). https://doi.org/10.1186/s12888-022-04509-7\u003c/li\u003e\n\u003cli\u003eCostafreda, S. G., Chu, C., Ashburner, J. \u0026amp; Fu, C. H. Y. Prognostic and diagnostic potential of the structural neuroanatomy of depression. \u003cem\u003ePLoS ONE\u003c/em\u003e\u003cstrong\u003e4\u003c/strong\u003e, e6353 (2009). https://doi.org/10.1371/journal.pone.0006353\u003c/li\u003e\n\u003cli\u003eSankar, A. \u003cem\u003eet al.\u003c/em\u003e Diagnostic potential of structural neuroimaging for depression from a multi-ethnic community sample. \u003cem\u003eBJPsych Open\u003c/em\u003e\u003cstrong\u003e2\u003c/strong\u003e, 247\u0026ndash;254 (2016). https://doi.org/10.1192/bjpo.bp.115.002493\u003c/li\u003e\n\u003cli\u003eWen, J. \u003cem\u003eet al.\u003c/em\u003e Characterizing heterogeneity in neuroimaging, cognition, clinical symptoms, and genetics among patients with late-life depression. \u003cem\u003eJAMA Psychiatry\u003c/em\u003e\u003cstrong\u003e79\u003c/strong\u003e, 464\u0026ndash;474 (2022). https://doi.org/10.1001/jamapsychiatry.2022.0020\u003c/li\u003e\n\u003cli\u003eEnneking, V., Leehr, E. J., Dannlowski, U. \u0026amp; Redlich, R. Brain structural effects of treatments for depression and biomarkers of response: a systematic review of neuroimaging studies. \u003cem\u003ePsychol. Med.\u003c/em\u003e\u003cstrong\u003e50\u003c/strong\u003e, 187\u0026ndash;209 (2020). https://doi.org/10.1017/S0033291719003660\u003c/li\u003e\n\u003cli\u003eDrysdale, A. T. \u003cem\u003eet al.\u003c/em\u003e Resting-state connectivity biomarkers define neurophysiological subtypes of depression. \u003cem\u003eNat. Med.\u003c/em\u003e\u003cstrong\u003e23\u003c/strong\u003e, 28\u0026ndash;38 (2017). https://doi.org/10.1038/nm.4246\u003c/li\u003e\n\u003cli\u003eFu, C. H. Y. \u003cem\u003eet al.\u003c/em\u003e Neuroanatomical dimensions in medication-free individuals with major depressive disorder and treatment response to SSRI antidepressant medications or placebo. \u003cem\u003eNat. Ment. Health\u003c/em\u003e\u003cstrong\u003e2\u003c/strong\u003e, 164\u0026ndash;176 (2024). https://doi.org/10.1038/s44220-023-00187-w\u003c/li\u003e\n\u003cli\u003eMiller, K. L. \u003cem\u003eet al.\u003c/em\u003e Multimodal population brain imaging in the UK Biobank prospective epidemiological study. \u003cem\u003eNat. Neurosci.\u003c/em\u003e\u003cstrong\u003e19\u003c/strong\u003e, 1523\u0026ndash;1536 (2016). https://doi.org/10.1038/nn.4393\u003c/li\u003e\n\u003cli\u003eAlfaro-Almagro, F. \u003cem\u003eet al.\u003c/em\u003e Image processing and Quality Control for the first 10,000 brain imaging datasets from UK Biobank. \u003cem\u003eNeuroImage\u003c/em\u003e\u003cstrong\u003e166\u003c/strong\u003e, 400\u0026ndash;424 (2018). https://doi.org/10.1016/j.neuroimage.2017.10.034\u003c/li\u003e\n\u003cli\u003eQureshi, D. \u003cem\u003eet al.\u003c/em\u003e Association of Metabolic Syndrome with Neuroimaging and Cognitive Outcomes in the UK Biobank. \u003cem\u003eDiabetes Care\u003c/em\u003e\u003cstrong\u003e47\u003c/strong\u003e, 1415\u0026ndash;1423 (2024). https://doi.org/10.2337/dc24-0537\u003c/li\u003e\n\u003cli\u003eColombo, F. \u003cem\u003eet al.\u003c/em\u003e Multimodal brain-derived subtypes of Major depressive disorder differentiate patients for anergic symptoms, immune-inflammatory markers, history of childhood trauma and treatment-resistance. \u003cem\u003eEur. Neuropsychopharmacol.\u003c/em\u003e\u003cstrong\u003e85\u003c/strong\u003e, 45\u0026ndash;57 (2024). https://doi.org/10.1016/j.euroneuro.2024.05.015\u003c/li\u003e\n\u003cli\u003eDavis, K. A. S. \u003cem\u003eet al.\u003c/em\u003e Mental health in UK Biobank - development, implementation and results from an online questionnaire completed by 157 366 participants: a reanalysis. \u003cem\u003eBJPsych Open\u003c/em\u003e\u003cstrong\u003e6\u003c/strong\u003e, e18 (2020). https://doi.org/10.1192/bjo.2019.100\u003c/li\u003e\n\u003cli\u003eHoward, D. M. \u003cem\u003eet al.\u003c/em\u003e Genetic stratification of depression in UK Biobank. \u003cem\u003eTransl. Psychiatry\u003c/em\u003e\u003cstrong\u003e10\u003c/strong\u003e, 163 (2020). https://doi.org/10.1038/s41398-020-0848-0\u003c/li\u003e\n\u003cli\u003eThorp, J. G. \u003cem\u003eet al.\u003c/em\u003e Symptom-level modelling unravels the shared genetic architecture of anxiety and depression. \u003cem\u003eNat. Hum. Behav.\u003c/em\u003e\u003cstrong\u003e5\u003c/strong\u003e, 1432\u0026ndash;1442 (2021). https://doi.org/10.1038/s41562-021-01094-9\u003c/li\u003e\n\u003cli\u003eOkbay, A. \u003cem\u003eet al.\u003c/em\u003e Genome-wide association study identifies 74 loci associated with educational attainment. \u003cem\u003eNature\u003c/em\u003e\u003cstrong\u003e533\u003c/strong\u003e, 539\u0026ndash;542 (2016). https://doi.org/10.1038/nature17671\u003c/li\u003e\n\u003cli\u003eZhang, B. \u003cem\u003eet al.\u003c/em\u003e Identifying behaviour-related and physiological risk factors for suicide attempts in the UK Biobank. \u003cem\u003eNat. Hum. Behav.\u003c/em\u003e\u003cstrong\u003e8\u003c/strong\u003e, 1784\u0026ndash;1797 (2024). https://doi.org/10.1038/s41562-024-01903-x\u003c/li\u003e\n\u003cli\u003eAmin, N. \u003cem\u003eet al.\u003c/em\u003e Interplay of Metabolome and Gut Microbiome in Individuals with Major Depressive Disorder vs Control Individuals. \u003cem\u003eJAMA Psychiatry\u003c/em\u003e\u003cstrong\u003e80\u003c/strong\u003e, 597\u0026ndash;609 (2023). https://doi.org/10.1001/jamapsychiatry.2023.0685\u003c/li\u003e\n\u003cli\u003eSollis, E. \u003cem\u003eet al.\u003c/em\u003e The NHGRI-EBI GWAS Catalog: knowledgebase and deposition resource. \u003cem\u003eNucleic Acids Res.\u003c/em\u003e\u003cstrong\u003e51\u003c/strong\u003e, D977\u0026ndash;D985 (2023). https://doi.org/10.1093/nar/gkac1010\u003c/li\u003e\n\u003cli\u003eNagel, M. \u003cem\u003eet al.\u003c/em\u003e Meta-analysis of genome-wide association studies for neuroticism in 449,484 individuals identifies novel genetic loci and pathways. \u003cem\u003eNat. Genet.\u003c/em\u003e\u003cstrong\u003e50\u003c/strong\u003e, 920\u0026ndash;927 (2018). https://doi.org/10.1038/s41588-018-0151-7\u003c/li\u003e\n\u003cli\u003eTrubetskoy, V. \u003cem\u003eet al.\u003c/em\u003e Mapping genomic loci implicates genes and synaptic biology in schizophrenia. \u003cem\u003eNature\u003c/em\u003e\u003cstrong\u003e604\u003c/strong\u003e, 502\u0026ndash;508 (2022). https://doi.org/10.1038/s41586-022-04434-5\u003c/li\u003e\n\u003cli\u003eMakowski, C. \u003cem\u003eet al.\u003c/em\u003e Larger cerebral cortex is genetically correlated with greater frontal area and dorsal thickness. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e\u003cstrong\u003e120\u003c/strong\u003e, e2214834120 (2023). https://doi.org/10.1073/pnas.2214834120\u003c/li\u003e\n\u003cli\u003eLiu, N. \u003cem\u003eet al.\u003c/em\u003e Cross-ancestry genome-wide association meta-analyses of hippocampal and subfield volumes. \u003cem\u003eNat. Genet.\u003c/em\u003e\u003cstrong\u003e55\u003c/strong\u003e, 1126\u0026ndash;1137 (2023). https://doi.org/10.1038/s41588-023-01425-8\u003c/li\u003e\n\u003cli\u003eMartin, S. \u003cem\u003eet al.\u003c/em\u003e Genetic Evidence for Different Adiposity Phenotypes and Their Opposing Influences on Ectopic Fat and Risk of Cardiometabolic Disease. \u003cem\u003eDiabetes\u003c/em\u003e\u003cstrong\u003e70\u003c/strong\u003e, 1843\u0026ndash;1856 (2021). https://doi.org/10.2337/db21-0129\u003c/li\u003e\n\u003cli\u003eTeder-Laving, M. \u003cem\u003eet al.\u003c/em\u003e Genome-wide meta-analysis identifies novel loci conferring risk of acne vulgaris. \u003cem\u003eEur. J. Hum. Genet.\u003c/em\u003e\u003cstrong\u003e32\u003c/strong\u003e, 1136\u0026ndash;1143 (2024). https://doi.org/10.1038/s41431-023-01326-8\u003c/li\u003e\n\u003cli\u003eKim, J., Lee, J., Nam, K. \u0026amp; Lee, S. Investigation of genetic variants and causal biomarkers associated with brain aging. \u003cem\u003eSci. Rep.\u003c/em\u003e\u003cstrong\u003e13\u003c/strong\u003e, 1526 (2023). https://doi.org/10.1038/s41598-023-27903-x\u003c/li\u003e\n\u003cli\u003evan der Meer, D. \u003cem\u003eet al.\u003c/em\u003e The genetic architecture of human cortical folding. \u003cem\u003eSci. Adv.\u003c/em\u003e\u003cstrong\u003e7\u003c/strong\u003e, eabj9446 (2021). https://doi.org/10.1126/sciadv.abj9446\u003c/li\u003e\n\u003cli\u003eHarris, B. H. L. \u003cem\u003eet al.\u003c/em\u003e New role of fat-free mass in cancer risk linked with genetic predisposition. \u003cem\u003eSci. Rep.\u003c/em\u003e\u003cstrong\u003e14\u003c/strong\u003e, 7270 (2024). https://doi.org/10.1038/s41598-024-54291-7\u003c/li\u003e\n\u003cli\u003eMelin, B. S. \u003cem\u003eet al.\u003c/em\u003e Genome-wide association study of glioma subtypes identifies specific differences in genetic susceptibility to glioblastoma and non-glioblastoma tumors. \u003cem\u003eNat. Genet.\u003c/em\u003e\u003cstrong\u003e49\u003c/strong\u003e, 789\u0026ndash;794 (2017). https://doi.org/10.1038/ng.3823\u003c/li\u003e\n\u003cli\u003eWang, H. \u003cem\u003eet al.\u003c/em\u003e Genome-wide interaction analysis of pathological hallmarks in Alzheimer\u0026rsquo;s disease. \u003cem\u003eNeurobiol. Aging\u003c/em\u003e\u003cstrong\u003e93\u003c/strong\u003e, 61\u0026ndash;68 (2020). https://doi.org/10.1016/j.neurobiolaging.2020.04.025\u003c/li\u003e\n\u003cli\u003eZhao, B. \u003cem\u003eet al.\u003c/em\u003e Large-scale GWAS reveals genetic architecture of brain white matter microstructure and genetic overlap with cognitive and mental health traits (n = 17,706). \u003cem\u003eMol. Psychiatry\u003c/em\u003e\u003cstrong\u003e26\u003c/strong\u003e, 3943\u0026ndash;3955 (2021). https://doi.org/10.1038/s41380-019-0569-z\u003c/li\u003e\n\u003cli\u003eCox, S. R., Ritchie, S. J., Fawns-Ritchie, C., Tucker-Drob, E. M. \u0026amp; Deary, I. J. Structural brain imaging correlates of general intelligence in UK Biobank. \u003cem\u003eIntelligence\u003c/em\u003e\u003cstrong\u003e76\u003c/strong\u003e, 101376 (2019). https://doi.org/10.1016/j.intell.2019.101376\u003c/li\u003e\n\u003cli\u003eLammer, L. \u003cem\u003eet al.\u003c/em\u003e Impact of social isolation on grey matter structure and cognitive functions: A population-based longitudinal neuroimaging study. \u003cem\u003eeLife\u003c/em\u003e\u003cstrong\u003e12\u003c/strong\u003e, e83660 (2023). https://doi.org/10.7554/eLife.83660\u003c/li\u003e\n\u003cli\u003eLiu, R. T., Kraines, M. A., Massing-Schaffer, M. \u0026amp; Alloy, L. B. Rejection Sensitivity and Depression: Mediation by Stress Generation. \u003cem\u003ePsychiatry\u003c/em\u003e\u003cstrong\u003e77\u003c/strong\u003e, 86\u0026ndash;97 (2014). https://doi.org/10.1521/psyc.2014.77.1.86\u003c/li\u003e\n\u003cli\u003eVekic, J., Stefanovic, A. \u0026amp; Zeljkovic, A. Obesity and Dyslipidemia: A Review of Current Evidence. \u003cem\u003eCurr. Obes. Rep.\u003c/em\u003e\u003cstrong\u003e12\u003c/strong\u003e, 207\u0026ndash;222 (2023). https://doi.org/10.1007/s13679-023-00518-z\u003c/li\u003e\n\u003cli\u003eChourpiliadis, C. \u003cem\u003eet al.\u003c/em\u003e Metabolic Profile and Long-Term Risk of Depression, Anxiety, and Stress-Related Disorders. \u003cem\u003eJAMA Netw. Open\u003c/em\u003e\u003cstrong\u003e7\u003c/strong\u003e, e244525 (2024). https://doi.org/10.1001/jamanetworkopen.2024.4525\u003c/li\u003e\n\u003cli\u003eLv, X. \u003cem\u003eet al.\u003c/em\u003e Body composition, lifestyle, and depression: a prospective study in the UK biobank. \u003cem\u003eBMC Public Health\u003c/em\u003e\u003cstrong\u003e24\u003c/strong\u003e, 393 (2024). https://doi.org/10.1186/s12889-024-17891-6\u003c/li\u003e\n\u003cli\u003eMujica-Parodi, L. R., Renelique, R. \u0026amp; Taylor, M. K. Higher body fat percentage is associated with increased cortisol reactivity and impaired cognitive resilience in response to acute emotional stress. \u003cem\u003eInt. J. Obes.\u003c/em\u003e\u003cstrong\u003e33\u003c/strong\u003e, 157\u0026ndash;165 (2009). https://doi.org/10.1038/ijo.2008.218\u003c/li\u003e\n\u003cli\u003eCole, J. H. \u003cem\u003eet al.\u003c/em\u003e Body mass index, but not FTO genotype or major depressive disorder, influences brain structure. \u003cem\u003eNeuroscience\u003c/em\u003e\u003cstrong\u003e252\u003c/strong\u003e, 109\u0026ndash;117 (2013). https://doi.org/10.1016/j.neuroscience.2013.07.015\u003c/li\u003e\n\u003cli\u003eHidese, S. \u003cem\u003eet al.\u003c/em\u003e Association of body mass index and its classifications with gray matter volume in individuals with a wide range of body mass index group: A whole-brain magnetic resonance imaging study. \u003cem\u003eFront. Hum. Neurosci.\u003c/em\u003e\u003cstrong\u003e16\u003c/strong\u003e, (2022). https://doi.org/10.3389/fnhum.2022.926804\u003c/li\u003e\n\u003cli\u003eLamers, F. \u003cem\u003eet al.\u003c/em\u003e Depression profilers and immuno-metabolic dysregulation: Longitudinal results from the NESDA study. \u003cem\u003eBrain. Behav. Immun.\u003c/em\u003e\u003cstrong\u003e88\u003c/strong\u003e, 174\u0026ndash;183 (2020). https://doi.org/10.1016/j.bbi.2020.04.002\u003c/li\u003e\n\u003cli\u003eBrydges, C. R. \u003cem\u003eet al.\u003c/em\u003e Metabolomic and inflammatory signatures of symptom dimensions in major depression. \u003cem\u003eBrain. Behav. Immun.\u003c/em\u003e\u003cstrong\u003e102\u003c/strong\u003e, 42\u0026ndash;52 (2022). https://doi.org/10.1016/j.bbi.2022.02.003\u003c/li\u003e\n\u003cli\u003eHuang, J. \u003cem\u003eet al.\u003c/em\u003e Genomics and phenomics of body mass index reveals a complex disease network. \u003cem\u003eNat. Commun.\u003c/em\u003e\u003cstrong\u003e13\u003c/strong\u003e, 7973 (2022). https://doi.org/10.1038/s41467-022-35553-2\u003c/li\u003e\n\u003cli\u003eHamer, M. \u0026amp; Batty, G. D. Association of body mass index and waist-to-hip ratio with brain structure. \u003cem\u003eNeurology\u003c/em\u003e\u003cstrong\u003e92\u003c/strong\u003e, e594\u0026ndash;e600 (2019). https://doi.org/10.1212/WNL.0000000000006879\u003c/li\u003e\n\u003cli\u003ePflanz, C.-P. \u003cem\u003eet al.\u003c/em\u003e Central obesity is selectively associated with cerebral gray matter atrophy in 15,634 subjects in the UK Biobank. \u003cem\u003eInt. J. Obes.\u003c/em\u003e\u003cstrong\u003e46\u003c/strong\u003e, 1059\u0026ndash;1067 (2022). https://doi.org/10.1038/s41366-021-00992-2\u003c/li\u003e\n\u003cli\u003eGautam, P., Anstey, K. J., Wen, W., Sachdev, P. S. \u0026amp; Cherbuin, N. Cortical gyrification and its relationships with cortical volume, cortical thickness, and cognitive performance in healthy mid-life adults. \u003cem\u003eBehav. Brain Res.\u003c/em\u003e\u003cstrong\u003e287\u003c/strong\u003e, 331\u0026ndash;339 (2015). https://doi.org/10.1016/j.bbr.2015.03.018\u003c/li\u003e\n\u003cli\u003ePenttil\u0026auml;, J. \u003cem\u003eet al.\u003c/em\u003e Cortical folding in patients with bipolar disorder or unipolar depression. \u003cem\u003eJ. Psychiatry Neurosci. JPN\u003c/em\u003e\u003cstrong\u003e34\u003c/strong\u003e, 127\u0026ndash;135 (2009).\u003c/li\u003e\n\u003cli\u003eGarel, C. \u003cem\u003eet al.\u003c/em\u003e Fetal cerebral cortex: normal gestational landmarks identified using prenatal MR imaging. \u003cem\u003eAJNR Am. J. Neuroradiol.\u003c/em\u003e\u003cstrong\u003e22\u003c/strong\u003e, 184\u0026ndash;189 (2001).\u003c/li\u003e\n\u003cli\u003eDubois, J. \u003cem\u003eet al.\u003c/em\u003e Primary cortical folding in the human newborn: an early marker of later functional development. \u003cem\u003eBrain\u003c/em\u003e\u003cstrong\u003e131\u003c/strong\u003e, 2028\u0026ndash;2041 (2008). https://doi.org/10.1093/brain/awn137\u003c/li\u003e\n\u003cli\u003eHedman, A. M., van Haren, N. E. M., Schnack, H. G., Kahn, R. S. \u0026amp; Hulshoff Pol, H. E. Human brain changes across the life span: a review of 56 longitudinal magnetic resonance imaging studies. \u003cem\u003eHum. Brain Mapp.\u003c/em\u003e\u003cstrong\u003e33\u003c/strong\u003e, 1987\u0026ndash;2002 (2012).\u003c/li\u003e\n\u003cli\u003eYang, W. \u003cem\u003eet al.\u003c/em\u003e The effects of childhood maltreatment on cortical thickness and gray matter volume: a coordinate-based meta-analysis. \u003cem\u003ePsychol. Med.\u003c/em\u003e\u003cstrong\u003e53\u003c/strong\u003e, 1681\u0026ndash;1699 (2023). https://doi.org/10.1017/S0033291723000661\u003c/li\u003e\n\u003cli\u003eIreton, R., Hughes, A. \u0026amp; Klabunde, M. A Functional Magnetic Resonance Imaging Meta-Analysis of Childhood Trauma. \u003cem\u003eBiol. Psychiatry Cogn. Neurosci. Neuroimaging\u003c/em\u003e\u003cstrong\u003e9\u003c/strong\u003e, 561\u0026ndash;570 (2024). https://doi.org/10.1016/j.bpsc.2024.01.009\u003c/li\u003e\n\u003cli\u003eLim, L., Howells, H., Radua, J. \u0026amp; Rubia, K. Aberrant structural connectivity in childhood maltreatment: A meta-analysis. \u003cem\u003eNeurosci. Biobehav. Rev.\u003c/em\u003e\u003cstrong\u003e116\u003c/strong\u003e, 406\u0026ndash;414 (2020). https://doi.org/10.1016/j.neubiorev.2020.07.004\u003c/li\u003e\n\u003cli\u003eGreen, J. G. \u003cem\u003eet al.\u003c/em\u003e Childhood Adversities and Adult Psychiatric Disorders in the National Comorbidity Survey Replication I: Associations With First Onset of DSM-IV Disorders. \u003cem\u003eArch. Gen. Psychiatry\u003c/em\u003e\u003cstrong\u003e67\u003c/strong\u003e, 113\u0026ndash;123 (2010). https://doi.org/10.1001/archgenpsychiatry.2009.186\u003c/li\u003e\n\u003cli\u003eRingwald, K. G. \u003cem\u003eet al.\u003c/em\u003e Interaction of recent stressful life events and childhood abuse on orbitofrontal grey matter volume in adults with depression. \u003cem\u003eJ. Affect. Disord.\u003c/em\u003e\u003cstrong\u003e312\u003c/strong\u003e, 122\u0026ndash;127 (2022). https://doi.org/10.1016/j.jad.2022.06.050\u003c/li\u003e\n\u003cli\u003eKeyes, K. M. \u0026amp; Westreich, D. UK Biobank, big data, and the consequences of non-representativeness. \u003cem\u003eThe Lancet\u003c/em\u003e\u003cstrong\u003e393\u003c/strong\u003e, 1297 (2019). https://doi.org/10.1016/S0140-6736(18)33067-8\u003c/li\u003e\n\u003cli\u003eSudlow, C. \u003cem\u003eet al.\u003c/em\u003e UK Biobank: An Open Access Resource for Identifying the Causes of a Wide Range of Complex Diseases of Middle and Old Age. \u003cem\u003ePLOS Med.\u003c/em\u003e\u003cstrong\u003e12\u003c/strong\u003e, e1001779 (2015). https://doi.org/10.1371/journal.pmed.1001779\u003c/li\u003e\n\u003cli\u003eDoshi, J. \u003cem\u003eet al.\u003c/em\u003e MUSE: MUlti-atlas region Segmentation utilizing Ensembles of registration algorithms and parameters, and locally optimal atlas selection. \u003cem\u003eNeuroImage\u003c/em\u003e\u003cstrong\u003e127\u003c/strong\u003e, 186\u0026ndash;195 (2016). https://doi.org/10.1016/j.neuroimage.2015.11.073\u003c/li\u003e\n\u003cli\u003eVarol, E., Sotiras, A. \u0026amp; Davatzikos, C. HYDRA: Revealing heterogeneity of imaging and genetic patterns through a multiple max-margin discriminative analysis framework. \u003cem\u003eNeuroImage\u003c/em\u003e\u003cstrong\u003e145\u003c/strong\u003e, 346\u0026ndash;364 (2017). https://doi.org/10.1016/j.neuroimage.2016.02.041\u003c/li\u003e\n\u003cli\u003eFortin, J.-P. \u003cem\u003eet al.\u003c/em\u003e Harmonization of cortical thickness measurements across scanners and sites. \u003cem\u003eNeuroImage\u003c/em\u003e\u003cstrong\u003e167\u003c/strong\u003e, 104\u0026ndash;120 (2018). https://doi.org/10.1016/j.neuroimage.2017.11.024\u003c/li\u003e\n\u003cli\u003eDavatzikos, C., Genc, A., Xu, D. \u0026amp; Resnick, S. M. Voxel-based morphometry using the RAVENS maps: methods and validation using simulated longitudinal atrophy. \u003cem\u003eNeuroImage\u003c/em\u003e\u003cstrong\u003e14\u003c/strong\u003e, 1361\u0026ndash;1369 (2001). https://doi.org/10.1006/nimg.2001.0937\u003c/li\u003e\n\u003cli\u003eSrinivasan, D. \u003cem\u003eet al.\u003c/em\u003e A comparison of Freesurfer and multi-atlas MUSE for brain anatomy segmentation: Findings about size and age bias, and inter-scanner stability in multi-site aging studies. \u003cem\u003eNeuroImage\u003c/em\u003e\u003cstrong\u003e223\u003c/strong\u003e, 117248 (2020). https://doi.org/10.1016/j.neuroimage.2020.117248\u003c/li\u003e\n\u003cli\u003eFawns-Ritchie, C. \u0026amp; Deary, I. J. Reliability and validity of the UK Biobank cognitive tests. \u003cem\u003ePLoS ONE\u003c/em\u003e\u003cstrong\u003e15\u003c/strong\u003e, e0231627 (2020). https://doi.org/10.1371/journal.pone.0231627\u003c/li\u003e\n\u003cli\u003eKroenke, K., Spitzer, R. \u0026amp; Williams, W. The PHQ-9: Validity of a brief depression severity measure. \u003cem\u003eJ. Gen. Intern. Med.\u003c/em\u003e\u003cstrong\u003e16\u003c/strong\u003e, 606\u0026ndash;616 (2001).\u003c/li\u003e\n\u003cli\u003eSpitzer, R. L., Kroenke, K., Williams, J. B. W. \u0026amp; L\u0026ouml;we, B. A brief measure for assessing generalized anxiety disorder: the GAD-7. \u003cem\u003eArch. Intern. Med.\u003c/em\u003e\u003cstrong\u003e166\u003c/strong\u003e, 1092\u0026ndash;1097 (2006).\u003c/li\u003e\n\u003cli\u003eEysenck, H. J. \u0026amp; Eysenck, S. B. G. \u003cem\u003eManual of the Eysenck Personality Questionnaire (Junior and Adult)\u003c/em\u003e. (Hodder and Stoughton Educational, 1975).\u003c/li\u003e\n\u003cli\u003eMao, R. \u003cem\u003eet al.\u003c/em\u003e Association of biological age acceleration with cardiac morphology, function, and incident heart failure: insights from UK Biobank participants. \u003cem\u003eEur. Heart J. - Cardiovasc. Imaging\u003c/em\u003e\u003cstrong\u003e25\u003c/strong\u003e, 1315\u0026ndash;1323 (2024). https://doi.org/10.1093/ehjci/jeae126\u003c/li\u003e\n\u003cli\u003eBycroft, C. \u003cem\u003eet al.\u003c/em\u003e The UK Biobank resource with deep phenotyping and genomic data. \u003cem\u003eNature\u003c/em\u003e\u003cstrong\u003e562\u003c/strong\u003e, 203\u0026ndash;209 (2018). https://doi.org/10.1038/s41586-018-0579-z\u003c/li\u003e\n\u003cli\u003eManichaikul, A. \u003cem\u003eet al.\u003c/em\u003e Robust relationship inference in genome-wide association studies. \u003cem\u003eBioinformatics\u003c/em\u003e\u003cstrong\u003e26\u003c/strong\u003e, 2867\u0026ndash;2873 (2010). https://doi.org/10.1093/bioinformatics/btq559\u003c/li\u003e\n\u003cli\u003ePurcell, S. \u003cem\u003eet al.\u003c/em\u003e PLINK: A Tool Set for Whole-Genome Association and Population-Based Linkage Analyses. \u003cem\u003eAm. J. Hum. Genet.\u003c/em\u003e\u003cstrong\u003e81\u003c/strong\u003e, 559\u0026ndash;575 (2007).\u003c/li\u003e\n\u003cli\u003eInsel, T. \u003cem\u003eet al.\u003c/em\u003e Research Domain Criteria (RDoC): Toward a new classification framework for research on mental disorders. \u003cem\u003eAm. J. Psychiatry\u003c/em\u003e\u003cstrong\u003e167\u003c/strong\u003e, 748\u0026ndash;751 (2010). https://doi.org/10.1176/appi.ajp.2010.09091379\u003c/li\u003e\n\u003cli\u003eCitrome, L. \u003cem\u003eet al.\u003c/em\u003e Making sense of the matrix: A qualitative assessment and commentary on connecting psychiatric symptom scale items to the Research Domain Criteria (RDoC). \u003cem\u003eInnov. Clin. Neurosci.\u003c/em\u003e\u003cstrong\u003e19\u003c/strong\u003e, 26\u0026ndash;32 (2022).\u003c/li\u003e\n\u003cli\u003eAhmed, A. T. \u003cem\u003eet al.\u003c/em\u003e Mapping depression rating scale phenotypes onto research domain criteria (RDoC) to inform biological research in mood disorders. \u003cem\u003eJ. Affect. Disord.\u003c/em\u003e\u003cstrong\u003e238\u003c/strong\u003e, 1\u0026ndash;7 (2018). https://doi.org/10.1016/j.jad.2018.05.005\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-5975731/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5975731/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eMajor depressive disorder (MDD) is a leading cause of disability worldwide, yet its diagnosis relies on clinical symptoms alone. Using machine learning applied to deeply phenotyped, medication-free participants with MDD, we identified two neuroanatomical dimensions. Dimension 2 (D2), compared to Dimension 1 (D1), was characterized by reductions in grey and white matter and was associated with limited treatment response to both antidepressant and placebo medications. Validation in UK Biobank general population cohort (n\u0026thinsp;=\u0026thinsp;37,235) confirmed that D2 is characterized by reduced grey and white matter, alongside widespread cognitive impairments, adverse events in both adulthood and childhood, increased self-harm and suicide attempts, a pro-atherogenic lipid profile, and genetic associations with neurodegenerative traits. These findings suggest that D1 and D2 reflect distinct neurobiological mechanisms underlying MDD, with important implications for and treatment outcomes. External validation was demonstrated in a general population-based cohort that delineated mechanisms underlying heterogeneity in MDD, identifying potential biomarkers that could aid in personalising treatment approaches for this debilitating disorder.\u003c/p\u003e","manuscriptTitle":"Neuroanatomical dimensions in major depression: external validation and links with cognition, adverse life events, self-harm, metabolomics and genetics","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-03-26 09:02:16","doi":"10.21203/rs.3.rs-5975731/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"communications-medicine","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"commsmed","sideBox":"Learn more about [Communications Medicine](http://www.nature.com/commsmed)","snPcode":"43856","submissionUrl":"https://mts-commsmed.nature.com/cgi-bin/main.plex","title":"Communications Medicine","twitterHandle":"@commsmedicine","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Communications Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"75a19aeb-a505-4298-8106-d006be453157","owner":[],"postedDate":"March 26th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":44300249,"name":"Health sciences/Medical research/Biomarkers/Diagnostic markers"},{"id":44300250,"name":"Health sciences/Medical research/Genetics research"}],"tags":[],"updatedAt":"2025-11-29T08:10:22+00:00","versionOfRecord":{"articleIdentity":"rs-5975731","link":"https://doi.org/10.1038/s43856-025-01219-5","journal":{"identity":"communications-medicine","isVorOnly":false,"title":"Communications Medicine"},"publishedOn":"2025-11-15 05:00:00","publishedOnDateReadable":"November 15th, 2025"},"versionCreatedAt":"2025-03-26 09:02:16","video":"","vorDoi":"10.1038/s43856-025-01219-5","vorDoiUrl":"https://doi.org/10.1038/s43856-025-01219-5","workflowStages":[]},"version":"v1","identity":"rs-5975731","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5975731","identity":"rs-5975731","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}