CoChat
Sign Up
Cited by 129Functional Brain Connectivity StudiesTreatment of Major DepressionMental Health Research Topics

Mapping neurophysiological subtypes of major depressive disorder using normative models of the functional connectome

Open access
Sun, Xiaoyi, Sun, Jinrong, Lu, Xiaowen, Dong, Qiangli, Zhang, Liang, Wang, Wenxu, Liu, Jin, Ma, Qing +35 more·2023·Biological Psychiatry
DOI 10.1016/j.biopsych.2023.05.021

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

Background Major depressive disorder (MDD) is a highly heterogeneous disorder that typically emerges in adolescence and can occur throughout adulthood. Studies aimed at quantitatively uncovering the heterogeneity of individual functional connectome abnormalities in MDD and identifying reproducibly distinct neurophysiological MDD subtypes across the lifespan, which could provide promising insights for precise diagnosis and treatment prediction, are still lacking.Methods Leveraging resting-state functional magnetic resonance imaging data from 1148 patients with MDD and 1079 healthy control participants (ages 11-93), we conducted the largest multisite analysis to date for neurophysiological MDD subtyping. First, we characterized typical lifespan trajectories of functional connectivity strength based on the normative model and quantitatively mapped the heterogeneous individual deviations among patients with MDD. Then, we identified neurobiological MDD subtypes using an unsupervised clustering algorithm and evaluated intersite reproducibility. Finally, we validated the subtype differences in baseline clinical variables and longitudinal treatment predictive capacity.Results Our findings indicated great intersubject heterogeneity in the spatial distribution and severity of functional connectome deviations among patients with MDD, which inspired the identification of 2 reproducible neurophysiological subtypes. Subtype 1 showed severe deviations, with positive deviations in the default mode, limbic, and subcortical areas and negative deviations in the sensorimotor and attention areas. Subtype 2 showed a moderate but converse deviation pattern. More importantly, subtype differences were observed in depressive item scores and the predictive ability of baseline deviations for antidepressant treatment outcomes.Conclusions These findings shed light on our understanding of different neurobiological mechanisms underlying the clinical heterogeneity of MDD and are essential for developing personalized treatments for this disorder.

AI snapshot

Written from the paper’s full text.

TL;DR

Using multisite resting-state fMRI from 1,148 MDD patients and 1,079 healthy controls, this study built normative models of functional connectivity strength and mapped individual deviations to identify neurophysiological MDD subtypes. Two subtypes emerged: a severe-deviation subtype with positive deviations in default mode, limbic, and subcortical areas and a moderate, largely converse subtype, with clinical differences and treatment-response prediction in the severe subtype.

Key findings

  • MDD patients showed larger individual FCS deviation indexes than HCs: number of extremely deviated regions (t=4.22), sum of positive extreme deviations (t=4.11), and sum of negative extreme deviations (t=-2.77; p<0.05, FDR corrected). 72.82% (N=836) of patients had an extreme FCS deviation in at least one region; extreme positive deviations in 25.78% (N=296) and extreme negative deviations in 66.38% (N=762). For any single brain region, extreme deviations were low: ≤2.35% positive (N=27) and ≤3.14% negative (N=36).
  • Two MDD subtypes were identified. Subtype 1 (37%, N=425) showed severe deviations with positive deviations in DMN, LIM, and SUB areas and negative deviations in SMN, DAN, and VAN areas. Subtype 2 (63%, N=723) showed moderate but converse deviations: negative in DMN, LIM, and SUB and positive in SMN, DAN, and VAN. Subtype 1 had significantly higher deviation indexes than HCs and subtype 2; subtype 2 had lower number of extremely deviated regions and sum of negative extreme deviations than HCs.
  • The severe-deviation subtype was older (t=2.64, p=0.008), had higher medicated proportion (χ²=6.11, p=0.013), and higher Suicide item score (t=2.02, p=0.044). The moderate-deviation subtype had higher Work and Activities (t=3.11, p=0.002) and Depressed Mood (t=2.42, p=0.016) item scores. HDRS-17 score correlated negatively with onset age in the severe subtype (r=-0.24, p=0.004) but not the moderate subtype (r=-0.00, p=0.966; subtype difference F=4.41, p=0.037).
  • In a 43-patient treatment subsample (16 severe, 27 moderate), baseline deviation maps predicted 6-month HDRS changes in the severe subtype (r=0.47, p=0.019, one-tailed permutation test) but not the moderate subtype (r=-0.14, p=0.785). Most positively contributive features were in DMN (24.1%), FPN (16.1%), and VAN (15.6%); most negatively contributive were in VIS (40.5%).

Methods

  • Multisite resting-state fMRI data from the DIDA-MDD working group: initially 2,414 participants (1,276 MDD, 1,138 HCs); final sample after quality control was 1,148 MDD patients (aged 11–93) and 1,079 HCs (aged 13–81). A subsample of 43 patients received 6-month paroxetine treatment.
  • Functional connectivity strength (FCS) was computed for 220 cerebral regions using a predefined parcellation; Pearson correlations were thresholded at r=0.2, Fisher r-to-z transformed, summed per region, z-normalized, and ComBat-harmonized for site effects. Normative models of FCS as a function of age and gender were estimated in HCs using Gaussian process regression with 10-fold cross-validation; individual deviations were expressed as Z values, with extreme deviations thresholded at Z=±2.6 (p<0.005).
  • K-means clustering was applied to individual deviation maps, with cluster number 2–10 assessed by 22 effective indexes via NbClust. Subtype differences in deviation patterns and demographic/clinical variables were tested; support vector regression with embedded 5-fold cross-validation and permutation tests examined prediction of treatment response. Validation included leave-one-site-out and alternative thresholds.

Limitations

  • Stated by the authors: the analysis is cross-sectional, so age-related trajectories are population-level rather than individual trajectories; longitudinal samples are needed. Clinical characterization used HDRS-17 item scores, and cognitive impairments were not collected. Treatment prediction included only paroxetine responders, so nonresponders are needed to model treatment-resistant depression. Subtypes were based on FCS alteration patterns, and future work should incorporate genetic and environmental information.
  • Evident from the text: the treatment-response prediction subsample was very small (43 patients; 16 severe, 27 moderate), limiting generalizability. The normative models and subtype solutions depend on analytic choices such as correlation threshold and extreme-deviation threshold, although the authors report validation across several alternatives.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1016/j.biopsych.2023.05.021

Cites

89 of 89 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 46 cited sources checked · 0 supported · 0 flagged

  • Findingown data
    § Normative models of functional connectivity strength

    Regions with decreased age-related FCS values were mainly in the precuneus, posterior cingulate cortex, medial prefrontal cortex, angular gyrus, insula, and medial temporal areas.

    “while those with decreased FCS were mainly in the precuneus, posterior cingulate cortex, medial prefrontal cortex, angular gyrus, insula, and medial temporal areas”
    Quote in the paperNot a substantial claim
  • Findingown data
    § Normative models of functional connectivity strength

    Normative models of FCS in HCs showed high generalizability in 10-fold cross-validation, with overall standardized mean squared error close to 1 (0.996 ± 0.013) and mean squared log-loss close to 0 (-0.001 ± 0.007).

    “The 10-fold cross-validation in the HCs revealed a high generalizability of the fitting performance of normative models for FCS, as indicated by overall standardized mean squared error close to 1 (0.996 ± 0.013) and mean squared log-loss close to 0 (-0.001 ± 0.007) (Supplementary Fig. 2).”
    Quote in the paperSubstantial claim
  • Findingown data
    § Highly heterogenous individual deviations from normative models in patients with MDD

    72.82% (N=836) of MDD patients showed extreme FCS deviations from the normative model in at least one brain region.

    “A total of 72.82% (N=836) of the patients with MDD showed extreme FCS deviations from the normative model in at least one brain region”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.
  • Findingown data
    § Normative models of functional connectivity strength

    Regions with increased age-related FCS values were located mostly in the lateral frontoparietal cortices, dorsal anterior cingulate cortex, medial occipital cortices, sensorimotor areas, and subcortical areas.

    “Specifically, regions with increased age-related FCS values were located mostly in the lateral frontoparietal cortices, dorsal anterior cingulate cortex, medial occipital cortices, sensorimotor areas, and subcortical areas”
    Quote in the paperNot a substantial claim
  • Findingown data
    § Normative models of functional connectivity strength

    Brain regions can be clustered into two categories according to their age-related FCS change trajectories in both female and male groups.

    “we found that the brain regions can be clustered (Supplement) into two categories according to their age-related FCS change trajectories in both female (Fig. 2a) and male groups (Supplementary Fig. 3).”
    Quote in the paperNot a substantial claim
  • May combine two claims. 34
  • Classified as probably not a substantial claim. 14

  • May combine two claims.

    Findingown data
    § Highly heterogenous individual deviations from normative models in patients with MDD

    Compared to HCs, MDD patients exhibited significantly larger individual FCS deviation indexes, including number of extremely deviated regions (t=4.22) and sum of positive (t=4.11) and negative (t=-2.77) extreme deviations.

    “Compared to the HCs, patients with MDD exhibited significantly larger individual FCS deviation indexes, including the number of extremely deviated regions ( t =4.22) and the sum of positive ( t =4.11) and negative ( t =-2.77) extreme deviations”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • May combine two claims.

    Findingown data
    § Highly heterogenous individual deviations from normative models in patients with MDD

    Regionally, MDD patients had significantly larger FCS deviations than HCs, with positive deviations mainly in bilateral lateral frontal cortex, precuneus, angular gyrus, and subcortical areas and negative deviations in left parahippocampal gyrus, right Rolandic operculum, and middle cingulum gyrus.

    “Regionally, the patient group had significantly larger FCS deviations than the HC group, with positive deviations mainly in the bilateral lateral frontal cortex, precuneus, angular gyrus, and subcortical areas and negative deviations in the left parahippocampal gyrus, right Rolandic operculum, and middle cingulum gyrus”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • May combine two claims.

    Findingown data
    § Highly heterogenous individual deviations from normative models in patients with MDD

    Extreme negative deviations in MDD patients were widespread over the whole brain, especially in medial sensorimotor cortex and temporal lobe.

    “the extreme negative deviations were widespread over the whole brain, especially in the medial sensorimotor cortex and the temporal lobe”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • May combine two claims.

    Findingown data
    § Highly heterogenous individual deviations from normative models in patients with MDD

    25.78% (N=296) of MDD patients showed extreme positive deviations and 66.38% (N=762) showed extreme negative deviations.

    “including extreme positive deviations in 25.78% (N=296) of patients and extreme negative deviations in 66.38% (N=762) of patients”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • May combine two claims.

    Findingown data
    § FCS deviation-based MDD subtypes

    HDRS-17 score was negatively correlated with onset age in severe-deviation subtype (r=-0.24, p=0.004) but not in moderate-deviation subtype (r=-0.00, p=0.966).

    “The HDRS-17 score was negatively correlated with onset age in patients with the severe-deviation subtype ( r =-0.24, p =0.004) but not in patients with the moderate-deviation subtype ( r =-0.00, p =0.966)”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • May combine two claims.

    Findingown data
    § FCS deviation-based MDD subtypes

    Patients with moderate-deviation subtype exhibited more severe symptoms in Work and Activities (t=3.11, p=0.002) and Depressed Mood items (t=2.42, p=0.016).

    “while patients with the moderate-deviation subtype exhibited more severe symptoms in the Work and Activities ( t =3.11, p =0.002) and Depressed Mood items ( t =2.42, p =0.016)”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • May combine two claims.

    Findingown data
    § FCS deviation-based MDD subtypes

    Patients with severe-deviation subtype were significantly older (t=2.64, p=0.008) and had higher medicated proportion (χ2=6.11, p=0.013) than moderate-deviation subtype.

    “patients with the severe-deviation subtype were significantly older ( t =2.64, p =0.008) and had a higher medicated proportion ( χ 2 =6.11, p =0.013) than patients with the moderate-deviation subtype”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • May combine two claims.

    Findingown data
    § FCS deviation-based MDD subtypes

    Most positively contributive features for prediction in severe-deviation subtype were located in DMN (24.1%), FPN (16.1%), and VAN (15.6%), and most negatively contributive in VIS (40.5%).

    “The most positively contributive features were located in the DMN (24.1%), FPN (16.1%), and VAN (15.6%), and the most negatively contributive features were in the VIS (40.5%)”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.
    • Rests on an assumption: The 43-patient subsample receiving 6-month paroxetine treatment is representative of MDD patients' antidepressant treatment response.

  • May combine two claims.

    Findingown data
    § FCS deviation-based MDD subtypes

    Among patients with follow-up treatment outcomes, 16 were clustered into severe-deviation subtype and 27 into moderate-deviation subtype.

    “Among the patients who had follow-up treatment outcomes, 16 patients were clustered into the severe-deviation subtype and the other 27 patients were clustered into the moderate-deviation subtype.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.
    • Rests on an assumption: The 43-patient subsample receiving 6-month paroxetine treatment is representative of MDD patients' antidepressant treatment response.

  • May combine two claims.

    Findingown data
    § FCS deviation-based MDD subtypes

    In subtype 1, number of extremely deviated regions, sum of positive extreme deviations, and sum of negative extreme deviations were significantly higher than in HCs and subtype 2.

    “Statistical comparisons showed that the number of extremely deviated regions, the sum of positive extreme deviations, and the sum of negative extreme deviations observed in subtype 1 patients were significantly higher than those observed in HCs and subtype 2 patients”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • May combine two claims.

    Findingown data
    § FCS deviation-based MDD subtypes

    Patients with subtype 2 (63%, N=723) showed moderate deviations with negative deviations in DMN, LIM, and SUB areas and positive deviations in SMN, DAN, and VAN areas.

    “However, the deviations observed in patients with subtype 2 (63%, N=723) were moderate, and the deviation patterns were significantly different, with negative deviations in the DMN, LIM, and SUB areas and positive deviations in the SMN, DAN, and V AN areas”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • May combine two claims.

    Findingown data
    § FCS deviation-based MDD subtypes

    In subtype 2, number of extremely deviated regions and sum of negative extreme deviations were significantly lower than in HCs.

    “while the number of extremely deviated regions and the sum of negative extreme deviations observed in subtype 2 patients were significantly lower than those observed in HCs”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • May combine two claims.

    Findingown data
    § Highly heterogenous individual deviations from normative models in MDD patients

    Positive FCS deviations mainly in DMN and SUB areas and negative deviations mainly in sensorimotor and lateral temporal cortices.

    “We found positive FCS deviations mainly in the DMN and SUB areas and negative deviations mainly in the sensorimotor and lateral temporal cortices.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • May combine two claims.

    Findingown data
    § FCS deviation-based MDD subtypes

    FCS deviations in DMN, LIM, SUB, SMN, DAN, and VAN exhibited significant differences and opposite alterations between the two subtypes, and DMN showed the most.

    “The FCS deviations in the DMN, LIM, SUB, SMN, DAN, and VAN exhibited significant differences and opposite alterations between the two subtypes, and the DMN showed the most.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • May combine two claims.

    Findingown data
    § FCS deviation-based MDD subtypes

    FCS deviation patterns have predictive power for treatment effects in the severe-deviation subtype but not in the other subtype.

    “More importantly, we found the predictive power of FCS deviation patterns for treatment effects in the severe-deviation subtype but not found in the other subtype.”
    • Also stated in: § Figure legends · plain
      “The prediction ability of deviation values for treatment response in patients of the severe-deviation subtype.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data
    § Discussion

    The study uncovered neurophysiological heterogeneity and subtypes of MDD through mapping deviations from normative models of functional connectome, using the largest R-fMRI dataset in MDD.

    “In this study, we uncovered the neurophysiological heterogeneity and subtypes of patients with MDD through mapping deviations from the normative models of functional connectome, by leveraging the currently largest R-fMRI dataset in MDD.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.
    • Rests on an assumption: The DIDA-MDD sample of 1,148 MDD patients and 1,079 HCs is representative of the broader MDD and healthy populations, so findings generalize.

  • May combine two claims.

    Premisenone given
    § Normative models of functional connectivity strength

    Normative model does not require assumptions about change trajectories and provides a useful tool to characterize nonlinear changes in features.

    “Compared to the traditional general linear model, the novel framework normative model does not require assumptions about the change trajectories and provides a useful tool to characterize any nonlinear changes in features.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data
    § Discussion

    Results highlight differences in spatial distribution of functional connectome abnormalities and significant disparities in demographic and clinical characteristics between the two identified neurobiological subtypes of MDD.

    “our results highlight not only differences in the spatial distribution of functional connectome abnormalities but also significant disparities in demographic and clinical characteristics between the two identified neurobiological subtypes of MDD.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.
    • Rests on an assumption: The DIDA-MDD sample of 1,148 MDD patients and 1,079 HCs is representative of the broader MDD and healthy populations, so findings generalize.

  • May combine two claims.

    Study design and conductown data
    § Validation results

    Under different thresholds in FCS calculation (r=0.15, 0.25), normative models and patient deviations were similar to main results, overlap rates >96%, and subtype differences largely remained.

    “Under different thresholds in FCS calculation (r=0.15, 0.25), the normative models and patient deviations were similar to our main results, the overlap rates of the resulting subtype indexes with the clustered indexes in the main results were &gt;96%, and the subtype differences largely remained”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data
    § Validation results

    When different threshold was used to define extreme deviations (FDR p<0.05), spatial overlap maps were slightly sparser, but specific brain regions affected by MDD still varied markedly among individual patients.

    “When different threshold was used to define extreme deviations (FDR p &lt;0.05), the spatial overlap maps were slightly sparser, but the specific brain regions affected by MDD still varied markedly among individual patients”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § Normative models of functional connectivity strength

    Several previous studies found linear age-related FCS decreases in medial prefrontal cortex, precuneus, insula, and calcarine and linear FCS increases in sensorimotor areas based on general linear model.

    “Similar to our findings, several previous studies found linear-age-related FCS decreases in the medial prefrontal cortex, precuneus, insula, and calcarine and linear FCS increases in sensorimotor areas based on the general linear model [36-38].”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [36] Filley, 1994, Attention and vigilance functions in normal aging, Appl Neuropsychol source

      Not checked in the free run.

    • cites [37] Sambataro, 2010, Age-related alterations in default mode network: Impact on working memory performance, Neurobiol Aging source

      Not checked in the free run.

    • cites [38] Feczko, 2019, The heterogeneity problem: Approaches to identify psychiatric subtypes, Trends Cogn Sci source

      Not checked in the free run.

  • May combine two claims.

    Study design and conductown data
    § Validation results

    Overlap rates of resulting clustered indexes in leave-one-site-out validation with main results were all >92%, and subtype differences were largely unchanged.

    “The overlap rates of the resulting clustered indexes in the leave-one-site-out validation with the clustered indexes in the main results were all &gt;92%, and the subtype differences were largely unchanged”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § Normative models of functional connectivity strength

    Sensorimotor areas are the least affected by aging.

    “Conversely, the sensorimotor areas are the least affected by aging [36].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [36] Filley, 1994, Attention and vigilance functions in normal aging, Appl Neuropsychol source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data
    § Normative models of functional connectivity strength

    Although brain regions had overall increased or decreased change trajectories, changes did not always follow linear or quadratic change, demonstrating high value of normative model in characterizing natural FCS change trajectories more accurately.

    “Notably, in our study, although brain regions had overall increased or decreased change trajectories, the changes did not always follow a linear or quadratic change, which demonstrates the high value of the normative model in characterizing the natural FCS change trajectories more accurately.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Premisenone given
    § Highly heterogenous individual deviations from normative models in MDD patients

    The model can recognize all sources of variance and reduce overly optimistic inferences and thus obtain more accurate and patient-specific individual deviations.

    “Importantly, the model can recognize all sources of variance and reduce overly optimistic inferences and thus obtain more accurate and patient-specific individual deviations for patients.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § Highly heterogenous individual deviations from normative models in MDD patients

    FCS alterations in these regions have been proven to be related to regulation of widespread cognitive, emotional, and executive control functions in MDD.

    “The FCS alterations in these regions have been proven to be related to the regulation of widespread cognitive, emotional, and executive control functions in patients with MDD [6, 42-50].”
    Quote in the paperNot a substantial claim0 of 10 sources checked
    • cites [6] Harald, 2012, Meta-review of depressive subtyping models, J Affect Disord source

      Not checked in the free run.

    • cites [42] Wang, 2014, Overlapping and segregated resting-state functional connectivity in patients with major depressive disorder with and without childhood neglect, Hum Brain Mapp source

      Not checked in the free run.

    • cites [43] Wang, 2015, The effects of antidepressant treatment on resting-state functional brain networks in patients with major depressive disorder, Hum Brain Mapp source

      Not checked in the free run.

    • cites [44] Shi, 2020, Abnormal functional connectivity strength in first-episode, drug-naive adult patients with major depressive disorder, Prog Neuropsychopharmacol Biol Psychiatry source

      Not checked in the free run.

    • cites [45] Zhang, 2021, Identification of psychiatric disorder subtypes from functional connectivity patterns in resting-state electroencephalography, Nat Biomed Eng source

      Not checked in the free run.

    • cites [46] Price, 2017, Data-driven subgroups in depression derived from directed functional connectivity paths at rest, Neuropsychopharmacology source

      Not checked in the free run.

    • cites [47] Chang, 2021, Identifying and validating subtypes within major psychiatric disorders based on frontal-posterior functional imbalance via deep learning, Mol Psychiatry source

      Not checked in the free run.

    • cites [48] Greene, 2022, Brain-phenotype models fail for individuals who defy sample stereotypes, Nature source

      Not checked in the free run.

    • cites [49] Dhamala, 2023, One size does not fit all: Methodological considerations for brain-based predictive modeling in psychiatry, Biol Psychiatry source

      Not checked in the free run.

    • cites [50] Auerbach, 2021, Neural correlates associated with suicide and nonsuicidal self-injury in youth, Biol Psychiatry source

      Not checked in the free run.

  • May combine two claims.

    Interpretationcitation
    § Normative models of functional connectivity strength

    The areas of FCS decrease are prominent hubs of global and local functional connectivity, and age-related decrease could underlie performance decline in working memory and visual sustained attention.

    “The areas of FCS decrease are the prominent hubs of global and local functional connectivity, and the age-related decrease could underlie the performance decline in working memory and visual sustained attention, which are the most affected cognitive functions that occur with aging [39-41].”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites [39] Feczko, 2020, Methods and challenges for assessing heterogeneity, Biol Psychiatry source

      Not checked in the free run.

    • cites [40] Yang, 2020, Brain structural and functional alterations specific to low sleep efficiency in major depressive disorder, Front Neurosci source

      Not checked in the free run.

    • cites [41] Zhang, 2011, Disrupted brain connectivity networks in drug-naive, first-episode major depressive disorder, Biol Psychiatry source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data
    § Highly heterogenous individual deviations from normative models in MDD patients

    FCS alteration is an important neuropathological feature of MDD, while alteration patterns among patients are largely different and there might be multiple forms of MDD.

    “Our results suggest that FCS alteration is an important neuropathological feature of MDD, while the alteration patterns among patients are largely different and there might be multiple forms of MDD.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.
    • Rests on an assumption: The DIDA-MDD sample of 1,148 MDD patients and 1,079 HCs is representative of the broader MDD and healthy populations, so findings generalize.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § FCS deviation-based MDD subtypes

    When individuals regurgitate negative emotions about themselves, suicidal thoughts and behaviors occur in response to the individual's desire to escape from both self-awareness and the associated unpleasant feelings.

    “When individuals are involved in regurgitating negative emotions about themselves, suicidal thoughts and behaviors occur in response to the individual's desire to escape from both self-awareness and the associated unpleasant feelings [59].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [59] Felger, 2016, Inflammation is associated with decreased functional connectivity within corticostriatal reward circuitry in depression, Mol Psychiatry source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation
    § FCS deviation-based MDD subtypes

    Increased functional connectomes and activities of the DMN and LIM areas are related to suicide, including the orbitofrontal cortex, medial prefrontal cortex, cingulate cortex, and striatum.

    “Studies have shown that the increased functional connectomes and activities of the DMN and LIM areas are related to suicide, including the orbitofrontal cortex, medial prefrontal cortex, cingulate cortex, and striatum [56-59].”
    Quote in the paperSubstantial claim0 of 4 sources checked
    • cites [56] First, 1997

      Not checked in the free run.

    • cites [57] Höflich, 2019, Circuit mechanisms of reward, anhedonia, and depression, Int J Neuropsychopharmacol source

      Not checked in the free run.

    • cites [58] Su, 2022, Progress and challenges in research of the mechanisms of anhedonia in major depressive disorder, Gen Psychiatr

      Not checked in the free run.

    • cites [59] Felger, 2016, Inflammation is associated with decreased functional connectivity within corticostriatal reward circuitry in depression, Mol Psychiatry source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § FCS deviation-based MDD subtypes

    Alterations in self-referential thinking may be associated with suicidal behavior.

    “Increasing evidence suggests that alterations in self-referential thinking may be associated with suicidal behavior [61].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [61] Geller, 2021, Specificity of anhedonic alterations in resting-state network connectivity and structure: A transdiagnostic approach, Psychiatry Res Neuroimaging source

      Not checked in the free run.

  • May combine two claims.

    Interpretationreasoning
    § FCS deviation-based MDD subtypes

    Functional connectome and activity of DMN areas are important biomarkers for neurophysiological subtyping of MDD.

    “Combined with these findings, our results indicate that the functional connectome and activity of DMN areas are important biomarkers for the neurophysiological subtyping of MDD.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation
    § FCS deviation-based MDD subtypes

    The orbitofrontal cortex is involved in learning, prediction, and decision-making for emotional and reward-related behaviors and is important in the regulation of behavioral impulsivity and response inhibition.

    “The orbitofrontal cortex is involved in learning, prediction, and decision-making for emotional and reward-related behaviors and is important in the regulation of behavioral impulsivity and response inhibition [60].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [60] Gabbay, 2013, Striatum-based circuitry of adolescent depression and anhedonia, J Am Acad Child Adolesc Psychiatry source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data
    § Highly heterogenous individual deviations from normative models in MDD patients

    Normative model of functional connectome is useful for identifying individual abnormalities and parsing heterogeneity of MDD.

    “Also, these findings reflect the useful application of normative model of functional connectome for identifying individual abnormities and parsing heterogeneity of MDD.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

  • May combine two claims.

    Premisecitation
    § FCS deviation-based MDD subtypes

    Brain resilience increases during development and early adulthood and then decreases during aging.

    “Moreover, evidence has shown that brain resilience increases during development and early adulthood and then decreases during aging [70].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [70] Dichter, 2015, A systematic review of relations between resting-state functional-MRI and treatment response in major depressive disorder, J Affect Disord source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation
    § FCS deviation-based MDD subtypes

    Studies of late-life depression showed increased FCS in the inferior parietal lobule and parahippocampal gyrus and decreased FCS in the somatosensory and motor cortices compared to HCs.

    “Similar to the alteration patterns in our results, studies of late-life depression showed increased FCS in the inferior parietal lobule and parahippocampal gyrus and decreased FCS in the somatosensory and motor cortices compared to HCs [69].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [69] Chin Fatt, 2020, Effect of intrinsic patterns of functional brain connectivity in moderating antidepressant treatment response in major depression, Am J Psychiatry source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation
    § FCS deviation-based MDD subtypes

    Several studies have explored the association between onset age and HDRS-17 score in patients with MDD, but the results were inconsistent.

    “Several studies have explored the association between the onset age and HDRS-17 score in patient with MDD, but the results were inconsistent [71-74].”
    Quote in the paperNot a substantial claim0 of 4 sources checked
    • cites [71] Flint, 2014, The genetics of major depression, Neuron source

      Not checked in the free run.

    • cites [72] Nguyen, 2022, Genetic heterogeneity and subtypes of major depression, Mol Psychiatry source

      Not checked in the free run.

    • cites [73] Tokuda, 2018, Identification of depression subtypes and relevant brain regions using a data-driven approach, Sci Rep source

      Not checked in the free run.

    • cites [74]

      Not checked in the free run.

  • May combine two claims.

    Interpretationcitation
    § FCS deviation-based MDD subtypes

    The ability of the brain to withstand disease may decrease and thus experience more severe alterations in older patients.

    “Thus, the ability of the brain to withstand disease may decrease and thus experience more severe alterations in older patients.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationcitation
    § FCS deviation-based MDD subtypes

    Alterations in these areas might be related to increased negative self-focused thought, impaired visuospatial and episodic memory, poor sleep quality, and deficits in physical health and functions in late-life patients.

    “The alterations in these areas might be related to the increased negative self-focused thought, impaired visuospatial and episodic memory, poor sleep quality, and deficits in physical health and functions in late-life patients.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductnone given
    § Limitations and future directions

    The age-related change trajectories shown here reveal age-specific population-level means and individual variabilities.

    “The age-related change trajectories shown here do not represent the trajectories of each participant, and they reveal age-specific population-level means and individual variabilities.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given
    § Limitations and future directions

    Cognitive impairments were not collected in the current retrospective study.

    “The patients with MDD also had varied cognitive impairments, which were not collected in the current retrospective study.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductreasoning
    § Limitations and future directions

    The age-related change trajectories shown here do not represent the trajectories of each participant.

    “The age-related change trajectories shown here do not represent the trajectories of each participant, and they reveal age-specific population-level means and individual variabilities.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given
    § Limitations and future directions

    The analysis was performed based on a cross-sectional sample.

    “First, our analysis was performed based on a cross-sectional sample.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given
    § Figure legends

    Ten-fold cross-validation was performed to assess the generalizability of the models.

    “Ten-fold cross-validation were performed to assess the generalizability of the models.”
    Quote in the paperNot a substantial claim
44 of 44

  • The paper relies on this without support.

    The normative model of FCS built from healthy controls provides a valid reference for estimating disease-related deviations in MDD patients.

    “For each patient with MDD, the FCS of the brain regions were positioned on the normative percentile charts from HCs to estimate individual deviation (Fig. 1b).”

  • The paper relies on this without support.

    The DIDA-MDD sample of 1,148 MDD patients and 1,079 HCs is representative of the broader MDD and healthy populations, so findings generalize.

    “we conducted a comprehensive investigation into the neurobiological heterogeneity and subtypes of MDD using a large multisite r-fMRI dataset of 1,148 patients with MDD and 1,079 matched healthy controls (HCs).”

  • The paper relies on this without support.

    The 43-patient subsample receiving 6-month paroxetine treatment is representative of MDD patients' antidepressant treatment response.

    “Additionally, a subsample of 43 patients (Supplementary Table 1) received a 6-month treatment with paroxetine (an antidepressant of selective serotonin reuptake inhibitor, SSRI), and the treatment outcomes were recorded (Supplement).”

  • The paper relies on this without support.

    The areas of the DMN are related to self-referential processing.

    “The areas of the DMN are related to self-referential processing.”

  • The paper relies on this without support.

    Previous studies have found different alteration patterns between patients in different age stages.

    “Previous studies have found different alteration patterns between patients in different age stages.”

  • The paper relies on this without support.

    Patients with MDD have varied cognitive impairments.

    “The patients with MDD also had varied cognitive impairments, which were not collected in the current retrospective study.”
6 of 6

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Sun, X., Sun, J., Lu, X., Dong, Q., Zhang, L., Wang, W., Liu, J., Ma, Q., Wang, X., Wei, D., Chen, Y., Liu, B., Huang, C.-C., Zheng, Y., Wu, Y., Chen, T., Cheng, Y., Xu, X., Gong, Q., Si, T., Qiu, S., Lin, C.-P., Cheng, J., Tang, Y., Wang, F., Qiu, J., Xie, P., Li, L., He, Y., He, Y., Li, L., Cheng, J., Gong, Q., Lin, C.-P., Qiu, J., Qiu, S., Si, T., Tang, Y., Wang, F., Xie, P., Xu, X., Xia, M., & Xia, M.
Affiliations
  • Beijing Normal University (CN)
  • Central South University (CN)
  • Shanghai Medical College of Fudan University (CN)
  • Southwest University (CN)
  • First Affiliated Hospital of Zhengzhou University (CN)
  • East China Normal University (CN)
  • First Affiliated Hospital of Guangzhou University of Chinese Medicine (CN)
  • Peking University (CN)
  • Sichuan University (CN)
  • First Affiliated Hospital of Kunming Medical University (CN)
ORCID
Venue
Biological Psychiatry
Publisher
Elsevier BV
Published
2023-12
Volume
94
Issue
12
Pages
936-947
ISSN
0006-3223 (ISSN-L)
DOI
10.1016/j.biopsych.2023.05.021
OpenAlex
W4379741254
Cited by
129
References
88
Citations by year
2026: 54 · 2025: 53 · 2024: 20 · 2023: 2
FWCI
19.27
Type
journal article
Language
en
Open access
Licence
Locations
  • Biological Psychiatry (publishedVersion)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc-nd)
  • PubMed (publishedVersion)
Funders
  • Science and Technology Planning Project of Fuzhou (2018-1002-SF-0442)
  • National Natural Science Foundation of China (81471251, 81771344, 81920108019, 82021004, 82071998, 91649117)
  • Beijing Nova Program (Z191100001119023)
Topics
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Functional Brain Connectivity Studies
  • Health Sciences › Medicine › Pharmacology › Treatment of Major Depression
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Mental Health Research Topics
Keywords
major depressive disorder, functional connectome, resting-state functional connectivity, normative model, lifespan trajectories, default mode network, limbic system, antidepressant response, treatment prediction, personalized treatment, clinical heterogeneity
MeSH terms
Adolescent, Adult, Brain, Brain Mapping, Major Depressive Disorder, Humans, Magnetic Resonance Imaging, Reproducibility of Results, Connectome

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-08.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-08.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review