Major Depressive Disorder <-> Vitiligo

Comorbidity BIDIRECTIONAL CANDIDATE
A large UK primary-care cohort supports a temporal association in both directions, but it does not establish a shared mechanism. The mechanistic evidence is substantially weaker than the epidemiology: the cross-disease transcriptomic study compared separate MDD and vitiligo expression datasets, used a small vitiligo discovery set, and nominated MAPK14 under permissive hypothesis-generating thresholds. The authors did check the candidates in independent disease-specific GEO datasets, but that discriminatory validation did not test a shared cell state or mediator. Patient-derived nonsegmental vitiligo keratinocytes show pan-p38 activation, whereas the isoform-specific MDD-arm evidence is a mouse serotonergic-neuron model. A confirmatory human MDD study found no advantage for the tested losmapimod regimen after a smaller prematurely terminated study had favored treatment. These compartment-, species-, and assay-specific findings constrain but do not refute an acquired bridge. A bidirectional Mendelian-randomization study found no significant directional causal effect between generalized vitiligo and broad mental-disorder phenotypes, including depression; that design does not adjudicate shared pleiotropic genetic liability. Psychosocial burden, healthcare surveillance, treatment, nonspecific inflammatory burden, shared pleiotropy, parallel tissue-specific p38 use, and an acquired cell-state-specific pathway therefore remain competing explanations.
A

Disease A

Major Depressive Disorder
Slug:Major_Depressive_Disorder
B

Disease B

Vitiligo
Slug:Vitiligo
G

Causal Mechanism Graphs

Major Depressive Disorder

graph LR
    Neurogenic_Niche_Interferon_Signaling_Activation["Neurogenic Niche Interferon Signaling Activation"]
    Electroconvulsive_Therapy["Electroconvulsive Therapy"]
    PENK+_GABAergic_Interneuron_Activation["PENK+ GABAergic Interneuron Activation"]
    Lithium_Augmentation["Lithium Augmentation"]
    IL1B["IL1B"]
    Excitation_Inhibition_Imbalance["Excitation-Inhibition Imbalance"]
    Ketamine_Esketamine["Ketamine/Esketamine"]
    Stress_Responsive_Transcription_Factor_Reprogramming["Stress-Responsive Transcription Factor Reprogramming"]
    Monoamine_Deficiency["Monoamine Deficiency"]
    Psilocybin_Assisted_Therapy["Psilocybin-Assisted Therapy"]
    C_511T["C-511T"]
    Serotonin_Norepinephrine_Reuptake_Inhibitors_SNRIs["Serotonin-Norepinephrine Reuptake Inhibitors (SNRIs)"]
    Stalled_Adult_Hippocampal_Neurogenesis["Stalled Adult Hippocampal Neurogenesis"]
    Neuroinflammation["Neuroinflammation"]
    Oligodendrocyte_Dysfunction_and_Demyelination["Oligodendrocyte Dysfunction and Demyelination"]
    Selective_Serotonin_Reuptake_Inhibitors_SSRIs["Selective Serotonin Reuptake Inhibitors (SSRIs)"]
    Transcranial_Magnetic_Stimulation["Transcranial Magnetic Stimulation"]
    Neuroplasticity_Deficits["Neuroplasticity Deficits"]

    Oligodendrocyte_Dysfunction_and_Demyelination --> Neuroplasticity_Deficits
    Oligodendrocyte_Dysfunction_and_Demyelination --> Neuroinflammation
    Stalled_Adult_Hippocampal_Neurogenesis --> Neuroplasticity_Deficits
    Neurogenic_Niche_Interferon_Signaling_Activation --> Stalled_Adult_Hippocampal_Neurogenesis
    Neurogenic_Niche_Interferon_Signaling_Activation --> Neuroinflammation
    Stress_Responsive_Transcription_Factor_Reprogramming --> Stalled_Adult_Hippocampal_Neurogenesis
    Stress_Responsive_Transcription_Factor_Reprogramming --> Excitation_Inhibition_Imbalance
    PENK+_GABAergic_Interneuron_Activation --> Excitation_Inhibition_Imbalance
    Selective_Serotonin_Reuptake_Inhibitors_SSRIs --> Monoamine_Deficiency
    Serotonin_Norepinephrine_Reuptake_Inhibitors_SNRIs --> Monoamine_Deficiency
    Electroconvulsive_Therapy --> Neuroplasticity_Deficits
    Ketamine_Esketamine --> Excitation_Inhibition_Imbalance
    Transcranial_Magnetic_Stimulation --> Neuroplasticity_Deficits
    Lithium_Augmentation --> Neuroplasticity_Deficits
    Psilocybin_Assisted_Therapy --> Neuroplasticity_Deficits
    C_511T --> IL1B

    style Neurogenic_Niche_Interferon_Signaling_Activation fill:#dbeafe
    style Electroconvulsive_Therapy fill:#fce7f3
    style PENK+_GABAergic_Interneuron_Activation fill:#dbeafe
    style Lithium_Augmentation fill:#fce7f3
    style IL1B fill:#f3e8ff
    style Excitation_Inhibition_Imbalance fill:#dbeafe
    style Ketamine_Esketamine fill:#fce7f3
    style Stress_Responsive_Transcription_Factor_Reprogramming fill:#dbeafe
    style Monoamine_Deficiency fill:#dbeafe
    style Psilocybin_Assisted_Therapy fill:#fce7f3
    style C_511T fill:#f3e8ff
    style Serotonin_Norepinephrine_Reuptake_Inhibitors_SNRIs fill:#fce7f3
    style Stalled_Adult_Hippocampal_Neurogenesis fill:#dbeafe
    style Neuroinflammation fill:#dbeafe
    style Oligodendrocyte_Dysfunction_and_Demyelination fill:#dbeafe
    style Selective_Serotonin_Reuptake_Inhibitors_SSRIs fill:#fce7f3
    style Transcranial_Magnetic_Stimulation fill:#fce7f3
    style Neuroplasticity_Deficits fill:#dbeafe

Vitiligo

graph LR
    Chemical_Exposure["Chemical Exposure"]
    Increased_Sensitivity_To_Sunlight["Increased Sensitivity To Sunlight"]
    Skin_depigmentation["Skin depigmentation"]
    Melanocyte_Destruction["Melanocyte Destruction"]
    UV_Exposure["UV Exposure"]
    Oxidative_Stress["Oxidative Stress"]
    Stress["Stress"]
    Depigmented_Patches["Depigmented Patches"]

    Depigmented_Patches --> Increased_Sensitivity_To_Sunlight
    UV_Exposure --> Skin_depigmentation
    Stress --> Oxidative_Stress
    Chemical_Exposure --> Melanocyte_Destruction

    style Chemical_Exposure fill:#dcfce7
    style Increased_Sensitivity_To_Sunlight fill:#fef3c7
    style Skin_depigmentation fill:#fef3c7
    style Melanocyte_Destruction fill:#dbeafe
    style UV_Exposure fill:#dcfce7
    style Oxidative_Stress fill:#dbeafe
    style Stress fill:#dcfce7
    style Depigmented_Patches fill:#fef3c7
S

Association Signals

Signal 1
LITERATURE EHR_COHORT_ASSOCIATION A_BEFORE_B
Population:The Health Improvement Network UK primary-care database; incident MDD cohort (n=405,397) and referent cohort (n=5,739,048), followed for incident vitiligo with covariate-adjusted Cox proportional-hazards models.
Signal IDs: MONDO:0002009 dismech -> MONDO:0008661 dismech
Temporal: A before B: , B before A: , Same time:
HR: 1.64
CI: 1.43 - 1.87
p:
FDR:
The publication reported P<0.0001; no exact p-value is available.
PMID:30528503 (SUPPORT)
Source: HUMAN_CLINICAL
"In adjusted models, MDD patients (n = 405,397) were at a 64% increased risk for vitiligo (hazard ratio 1.64"
Quantifies the adjusted MDD-to-vitiligo temporal association.
PMID:30528503 (SUPPORT)
Source: HUMAN_CLINICAL
"1.43-1.87, P < .0001) compared with the referent cohort (n = 5,739,048)."
Supplies the reported confidence interval and thresholded P value for the same adjusted hazard ratio; the publication does not provide an exact P value.
Signal 2
LITERATURE EHR_COHORT_ASSOCIATION B_BEFORE_A
Population:The Health Improvement Network UK primary-care database; incident vitiligo patients diagnosed before age 30 and the referent cohort, followed for incident MDD with covariate-adjusted Cox proportional-hazards models.
Demographics: Age younger than 30 years at vitiligo diagnosis
Signal IDs: MONDO:0002009 dismech -> MONDO:0008661 dismech
Temporal: A before B: , B before A: , Same time:
HR: 1.31
CI: 1.14 - 1.5
p:
FDR:
The publication reported P<0.0001; no exact p-value is available.
PMID:30528503 (SUPPORT)
Source: HUMAN_CLINICAL
"Compared with the referent cohort (n = 6,137,696), patients with vitiligo (n = 7104) that were <30 years of age at diagnosis had a higher risk of developing MDD than patients ≥30 years of age (hazard ratio 1.31, 95% CI 1.14-1.50, P < .0001 vs 1.22, 95% CI 1.08-1.37, P = .001, respectively)."
Quantifies the younger-age vitiligo-to-MDD association.
Signal 3
LITERATURE EHR_COHORT_ASSOCIATION B_BEFORE_A
Population:The Health Improvement Network UK primary-care database; incident vitiligo patients diagnosed at age 30 or older and the referent cohort, followed for incident MDD with covariate-adjusted Cox proportional-hazards models.
Demographics: Age 30 years or older at vitiligo diagnosis
Signal IDs: MONDO:0002009 dismech -> MONDO:0008661 dismech
Temporal: A before B: , B before A: , Same time:
HR: 1.22
CI: 1.08 - 1.37
p: 0.001
FDR:
Adjusted hazard ratio for incident MDD.
PMID:30528503 (SUPPORT)
Source: HUMAN_CLINICAL
"Compared with the referent cohort (n = 6,137,696), patients with vitiligo (n = 7104) that were <30 years of age at diagnosis had a higher risk of developing MDD than patients ≥30 years of age (hazard ratio 1.31, 95% CI 1.14-1.50, P < .0001 vs 1.22, 95% CI 1.08-1.37, P = .001, respectively)."
Quantifies the older-age vitiligo-to-MDD association.
H

Hypotheses

Acquired, cell-context-specific p38-alpha/MAPK14 activation may be a shared mediator. Current arm-specific findings instead resolve to non-equivalent contexts: p38-alpha-specific mouse serotonergic-neuron genetics for stress-related behavior and pan-p38 signaling in patient-derived nonsegmental-vitiligo keratinocytes. The p38 MAPK inhibitor losmapimod also lacked benefit in the larger of two human MDD studies. This remains a bridge hypothesis, not an established shared pathway; no study demonstrates the same MAPK14-dependent state in comorbid humans.
PMID:42418414 (SUPPORT)
Source: COMPUTATIONAL
"These findings primarily serve to generate hypotheses regarding shared mechanisms and prioritize targets for subsequent experimental validation."
The cross-disease transcriptomic reanalysis nominates MAPK14 but explicitly frames the result as hypothesis-generating.
PMID:42418414 (SUPPORT)
Source: COMPUTATIONAL
"For MDD, the GSE98793 dataset was utilized, which includes 192 whole blood samples: 64 from healthy controls, 128 from patients with MDD, using the GPL570 platform for sequencing. For vitiligo, the GSE65127 and GSE53146 datasets were integrated using the R package ‘sva’ to correct batch effects. The combined datasets included 15 healthy samples and 15 vitiligo samples, using the GPL570 and GPL14951 platforms for sequencing."
The analysis combined separate disease datasets and had only 15 vitiligo and 15 healthy samples, rather than paired samples from a comorbid cohort.
PMID:42418414 (SUPPORT)
Source: COMPUTATIONAL
"Differentially expressed genes (DEGs) between disease and control groups were identified using the R package ‘limma’, applying a threshold of |log2FC| > 0 and P-value < 0.05."
The permissive nonzero fold-change and nominal P-value screen limits the specificity of the candidate-gene intersection.
PMID:42418414 (SUPPORT)
Source: COMPUTATIONAL
"It should be noted that these ssGSEA scores represent inferred immune signatures from bulk RNA-seq data, rather than direct measurements of actual immune cell abundances."
The immune-cell results are computational enrichment scores rather than measured cell abundances.
PMID:42418414 (SUPPORT)
Source: COMPUTATIONAL
"Two independent GEO datasets (GSE52790 and GSE80009) were utilized as external validation sets to examine the consistency and stability of these findings."
Independent disease-specific datasets strengthen the candidate-marker screen relative to discovery alone, but ROC-style discriminatory validation does not establish a shared cell state, p38 activity, mediation, or causality.
PMID:21835346 (SUPPORT)
Source: MODEL_ORGANISM
"Social defeat stress produced social avoidance (a model of depression-like behaviors) and reinstatement of cocaine preference (a measure of addiction risk) in wild-type mice, but not in mice having p38α MAPK selectively deleted in serotonin-producing neurons of the dorsal raphe nucleus."
Cell-specific mouse genetics supports p38-alpha involvement in a stress-related behavioral model, but it is not human MDD evidence and does not connect to vitiligo.
PMID:24699061 (SUPPORT)
Source: HUMAN_CLINICAL
"A subsequent study, Study 009 (n=128), designed using a Bayesian approach based on a prior derived from Study 574, showed no advantage for losmapimod (Bech, 6 weeks: endpoint drug vs. placebo difference = 1.11; 95% credible interval, -0.22, 2.50). Biomarker data showed no significant changes. In conclusion 7.5 mg BID losmapimod was not effective in MDD."
The larger human MDD study and its biomarkers were null, although a smaller prematurely terminated study had favored treatment. A single regimen, duration, and selected MDD population without demonstrated biomarker modulation is a narrow pharmacologic constraint, not a refutation of cell-specific MAPK14 activity or a shared acquired mediator.
PMID:20085492 (SUPPORT)
Source: IN_VITRO
"In keratinocytes from perilesional vitiligo skin, we observed high levels of activated p38, NF-kB p65 subunit, p53, and Smac/DIABLO proteins."
Patient-derived keratinocytes from 12 people with nonsegmental vitiligo support a skin-compartment pan-p38 signal. The cultured-cell study does not resolve MAPK14 from other p38 isoforms or connect the signal to MDD.
PMID:18575770 (SUPPORT)
Source: IN_VITRO
"Also, H2O2 treatment activates JNK and p38 MAPK, and executive caspase 3 in B10BR cells."
Oxidative stress activates p38-family signaling in cultured mouse melanocytes, but the study is neither MAPK14-specific nor evidence of a shared human mechanism.
A directional causal effect of liability to one disease on the other could produce the bidirectional temporal association. Current bidirectional Mendelian-randomization evidence found no detectable directional causal effect between generalized vitiligo and the study's broad depression phenotype, although it cannot exclude weak or subtype-specific effects.
Source: COMPUTATIONAL
"In our findings, none of the rigorous bidirectional MR analyses uncovered a significant causal association."
The null bidirectional MR result constrains genetically instrumented disease-to-disease effects between generalized vitiligo and the broad depression phenotype used in that analysis, rather than specifically adjudicating clinically defined MDD subgroups.
Shared pleiotropic genetic liability could independently increase susceptibility to both MDD and vitiligo without either disease causing the other. Bidirectional Mendelian randomization does not test this common-cause model; cross-trait global and local genetic correlation followed by locus-level colocalization would be needed to distinguish shared variants from directional causality and linkage.
Psychosocial effects of visible skin disease, healthcare-contact surveillance, treatment, and nonspecific chronic inflammatory burden may explain some or all of the association without a vitiligo-specific immune bridge to MDD.
PMID:30528503 (SUPPORT)
Source: HUMAN_CLINICAL
"This risk was decreased in patients using antidepressants."
The cohort reported an unquantified treatment-associated attenuation. Antidepressant pharmacology, indication, adherence, healthcare behavior, and residual confounding remain inseparable, so this observation cannot distinguish a biological mediator from treatment, psychosocial, or surveillance explanations.
PMID:41781039 (SUPPORT)
Source: HUMAN_CLINICAL
"In matched comparisons with atopic dermatitis, only a few and inconsistent differences were observed."
Similar mental-health profiles in another visible inflammatory skin disease favor nonspecific or psychosocial alternatives that must be tested against the shared-immune hypothesis.
Y

Raw YAML

Show YAML 
name: com_Major_Depressive_Disorder__Vitiligo
creation_date: '2026-07-26T08:22:20Z'
curation_status: CANDIDATE
notes: >-
  A large UK primary-care cohort supports a temporal association in both
  directions, but it does not establish a shared mechanism. The mechanistic
  evidence is substantially weaker than the epidemiology: the cross-disease
  transcriptomic study compared separate MDD and vitiligo expression datasets,
  used a small vitiligo discovery set, and nominated MAPK14 under permissive
  hypothesis-generating thresholds. The authors did check the candidates in
  independent disease-specific GEO datasets, but that discriminatory validation
  did not test a shared cell state or mediator. Patient-derived nonsegmental
  vitiligo keratinocytes show pan-p38 activation, whereas the isoform-specific
  MDD-arm evidence is a mouse serotonergic-neuron model. A confirmatory human
  MDD study found no advantage for the tested losmapimod regimen after a smaller
  prematurely terminated study had favored treatment.
  These compartment-, species-, and assay-specific findings constrain but do
  not refute an acquired bridge. A bidirectional Mendelian-randomization study
  found no significant directional causal effect between generalized vitiligo
  and broad mental-disorder phenotypes, including depression; that design does
  not adjudicate shared pleiotropic genetic liability. Psychosocial burden,
  healthcare surveillance, treatment, nonspecific inflammatory burden, shared
  pleiotropy, parallel tissue-specific p38 use, and an acquired
  cell-state-specific pathway therefore remain competing explanations.

disease_a:
  slug: Major_Depressive_Disorder
  preferred_term: major depressive disorder
  term:
    id: MONDO:0002009
    label: major depressive disorder

disease_b:
  slug: Vitiligo
  preferred_term: vitiligo
  term:
    id: MONDO:0008661
    label: vitiligo

directionality: BIDIRECTIONAL
effect_direction: RISK

literature_evidence:
- reference: PMID:30528503
  reference_title: "Vitiligo and major depressive disorder: A bidirectional population-based cohort study."
  supports: SUPPORT
  evidence_source: HUMAN_CLINICAL
  snippet: "In adjusted models, MDD patients (n = 405,397) were at a 64% increased risk for vitiligo (hazard ratio 1.64"
  explanation: >-
    In a large UK primary-care cohort, incident MDD preceded an increased
    hazard of subsequently recorded vitiligo after covariate adjustment.
- reference: PMID:30528503
  reference_title: "Vitiligo and major depressive disorder: A bidirectional population-based cohort study."
  supports: SUPPORT
  evidence_source: HUMAN_CLINICAL
  snippet: "Compared with the referent cohort (n = 6,137,696), patients with vitiligo (n = 7104) that were <30 years of age at diagnosis had a higher risk of developing MDD than patients ≥30 years of age (hazard ratio 1.31, 95% CI 1.14-1.50, P < .0001 vs 1.22, 95% CI 1.08-1.37, P = .001, respectively)."
  explanation: >-
    The reverse analysis also found an increased hazard of subsequently
    recorded MDD after incident vitiligo, with age-stratified estimates.
- reference: PMID:41781039
  reference_title: "Prevalence and comparative risk of mental health disorders in persons with vitiligo: a retrospective matched cohort study using claims data with expert-informed case validation."
  supports: SUPPORT
  evidence_source: HUMAN_CLINICAL
  snippet: "The mental health comorbidity profile was largely comparable to that of atopic dermatitis, whereas more pronounced differences were observed in comparisons with psoriasis."
  explanation: >-
    A dermatologic comparator analysis confirms mental-health burden but
    weakens the specificity of a vitiligo-specific biological explanation.

association_signals:
- source: LITERATURE
  method: EHR_COHORT_ASSOCIATION
  signal_disorder_a_id: MONDO:0002009
  signal_disorder_b_id: MONDO:0008661
  population: >-
    The Health Improvement Network UK primary-care database; incident MDD
    cohort (n=405,397) and referent cohort (n=5,739,048), followed for incident
    vitiligo with covariate-adjusted Cox proportional-hazards models.
  directionality: A_BEFORE_B
  effect_direction: RISK
  statistics:
    metrics:
    - metric_type: HR
      metric_value: 1.64
      metric_ci_lower: 1.43
      metric_ci_upper: 1.87
      notes: >-
        The publication reported P<0.0001; no exact p-value is available.
    evidence:
    - reference: PMID:30528503
      reference_title: "Vitiligo and major depressive disorder: A bidirectional population-based cohort study."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: "In adjusted models, MDD patients (n = 405,397) were at a 64% increased risk for vitiligo (hazard ratio 1.64"
      explanation: Quantifies the adjusted MDD-to-vitiligo temporal association.
    - reference: PMID:30528503
      reference_title: "Vitiligo and major depressive disorder: A bidirectional population-based cohort study."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: "1.43-1.87, P < .0001) compared with the referent cohort (n = 5,739,048)."
      explanation: >-
        Supplies the reported confidence interval and thresholded P value for
        the same adjusted hazard ratio; the publication does not provide an
        exact P value.

- source: LITERATURE
  method: EHR_COHORT_ASSOCIATION
  signal_disorder_a_id: MONDO:0002009
  signal_disorder_b_id: MONDO:0008661
  population: >-
    The Health Improvement Network UK primary-care database; incident vitiligo
    patients diagnosed before age 30 and the referent cohort, followed for
    incident MDD with covariate-adjusted Cox proportional-hazards models.
  demographics:
    age_range: younger than 30 years at vitiligo diagnosis
  directionality: B_BEFORE_A
  effect_direction: RISK
  statistics:
    metrics:
    - metric_type: HR
      metric_value: 1.31
      metric_ci_lower: 1.14
      metric_ci_upper: 1.50
      notes: >-
        The publication reported P<0.0001; no exact p-value is available.
    evidence:
    - reference: PMID:30528503
      reference_title: "Vitiligo and major depressive disorder: A bidirectional population-based cohort study."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: "Compared with the referent cohort (n = 6,137,696), patients with vitiligo (n = 7104) that were <30 years of age at diagnosis had a higher risk of developing MDD than patients ≥30 years of age (hazard ratio 1.31, 95% CI 1.14-1.50, P < .0001 vs 1.22, 95% CI 1.08-1.37, P = .001, respectively)."
      explanation: Quantifies the younger-age vitiligo-to-MDD association.

- source: LITERATURE
  method: EHR_COHORT_ASSOCIATION
  signal_disorder_a_id: MONDO:0002009
  signal_disorder_b_id: MONDO:0008661
  population: >-
    The Health Improvement Network UK primary-care database; incident vitiligo
    patients diagnosed at age 30 or older and the referent cohort, followed for
    incident MDD with covariate-adjusted Cox proportional-hazards models.
  demographics:
    age_range: 30 years or older at vitiligo diagnosis
  directionality: B_BEFORE_A
  effect_direction: RISK
  statistics:
    metrics:
    - metric_type: HR
      metric_value: 1.22
      metric_ci_lower: 1.08
      metric_ci_upper: 1.37
      p_value: 0.001
      notes: Adjusted hazard ratio for incident MDD.
    evidence:
    - reference: PMID:30528503
      reference_title: "Vitiligo and major depressive disorder: A bidirectional population-based cohort study."
      supports: SUPPORT
      evidence_source: HUMAN_CLINICAL
      snippet: "Compared with the referent cohort (n = 6,137,696), patients with vitiligo (n = 7104) that were <30 years of age at diagnosis had a higher risk of developing MDD than patients ≥30 years of age (hazard ratio 1.31, 95% CI 1.14-1.50, P < .0001 vs 1.22, 95% CI 1.08-1.37, P = .001, respectively)."
      explanation: Quantifies the older-age vitiligo-to-MDD association.

hypotheses:
- description: >-
    Acquired, cell-context-specific p38-alpha/MAPK14 activation may be a shared
    mediator. Current arm-specific findings instead resolve to non-equivalent
    contexts: p38-alpha-specific mouse serotonergic-neuron genetics for
    stress-related behavior and pan-p38 signaling in patient-derived
    nonsegmental-vitiligo keratinocytes. The p38 MAPK inhibitor losmapimod also
    lacked benefit in the larger of two human MDD studies. This remains a bridge
    hypothesis, not an established shared pathway; no study demonstrates the
    same MAPK14-dependent state in comorbid humans.
  evidence:
  - reference: PMID:42418414
    reference_title: "Bioinformatics analysis reveals the characteristics of immune microenvironment in major depressive disorder and vitiligo."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: "These findings primarily serve to generate hypotheses regarding shared mechanisms and prioritize targets for subsequent experimental validation."
    explanation: >-
      The cross-disease transcriptomic reanalysis nominates MAPK14 but explicitly
      frames the result as hypothesis-generating.
  - reference: PMID:42418414
    reference_title: "Bioinformatics analysis reveals the characteristics of immune microenvironment in major depressive disorder and vitiligo."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: "For MDD, the GSE98793 dataset was utilized, which includes 192 whole blood samples: 64 from healthy controls, 128 from patients with MDD, using the GPL570 platform for sequencing. For vitiligo, the GSE65127 and GSE53146 datasets were integrated using the R package ‘sva’ to correct batch effects. The combined datasets included 15 healthy samples and 15 vitiligo samples, using the GPL570 and GPL14951 platforms for sequencing."
    explanation: >-
      The analysis combined separate disease datasets and had only 15 vitiligo
      and 15 healthy samples, rather than paired samples from a comorbid cohort.
  - reference: PMID:42418414
    reference_title: "Bioinformatics analysis reveals the characteristics of immune microenvironment in major depressive disorder and vitiligo."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: "Differentially expressed genes (DEGs) between disease and control groups were identified using the R package ‘limma’, applying a threshold of |log2FC| > 0 and P-value < 0.05."
    explanation: >-
      The permissive nonzero fold-change and nominal P-value screen limits the
      specificity of the candidate-gene intersection.
  - reference: PMID:42418414
    reference_title: "Bioinformatics analysis reveals the characteristics of immune microenvironment in major depressive disorder and vitiligo."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: "It should be noted that these ssGSEA scores represent inferred immune signatures from bulk RNA-seq data, rather than direct measurements of actual immune cell abundances."
    explanation: >-
      The immune-cell results are computational enrichment scores rather than
      measured cell abundances.
  - reference: PMID:42418414
    reference_title: "Bioinformatics analysis reveals the characteristics of immune microenvironment in major depressive disorder and vitiligo."
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: "Two independent GEO datasets (GSE52790 and GSE80009) were utilized as external validation sets to examine the consistency and stability of these findings."
    explanation: >-
      Independent disease-specific datasets strengthen the candidate-marker
      screen relative to discovery alone, but ROC-style discriminatory
      validation does not establish a shared cell state, p38 activity,
      mediation, or causality.
  - reference: PMID:21835346
    reference_title: "Selective p38α MAPK deletion in serotonergic neurons produces stress resilience in models of depression and addiction."
    supports: SUPPORT
    evidence_source: MODEL_ORGANISM
    snippet: "Social defeat stress produced social avoidance (a model of depression-like behaviors) and reinstatement of cocaine preference (a measure of addiction risk) in wild-type mice, but not in mice having p38α MAPK selectively deleted in serotonin-producing neurons of the dorsal raphe nucleus."
    explanation: >-
      Cell-specific mouse genetics supports p38-alpha involvement in a
      stress-related behavioral model, but it is not human MDD evidence and
      does not connect to vitiligo.
  - reference: PMID:24699061
    reference_title: "Evaluation of antidepressant properties of the p38 MAP kinase inhibitor losmapimod (GW856553) in Major Depressive Disorder: Results from two randomised, placebo-controlled, double-blind, multicentre studies using a Bayesian approach."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "A subsequent study, Study 009 (n=128), designed using a Bayesian approach based on a prior derived from Study 574, showed no advantage for losmapimod (Bech, 6 weeks: endpoint drug vs. placebo difference = 1.11; 95% credible interval, -0.22, 2.50). Biomarker data showed no significant changes. In conclusion 7.5 mg BID losmapimod was not effective in MDD."
    explanation: >-
      The larger human MDD study and its biomarkers were null, although a
      smaller prematurely terminated study had favored treatment. A single
      regimen, duration, and selected MDD population without demonstrated
      biomarker modulation is a narrow pharmacologic constraint, not a
      refutation of cell-specific MAPK14 activity or a shared acquired mediator.
  - reference: PMID:20085492
    reference_title: "The involvement of Smac/DIABLO, p53, NF-kB, and MAPK pathways in apoptosis of keratinocytes from perilesional vitiligo skin: Protective effects of curcumin and capsaicin."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: "In keratinocytes from perilesional vitiligo skin, we observed high levels of activated p38, NF-kB p65 subunit, p53, and Smac/DIABLO proteins."
    explanation: >-
      Patient-derived keratinocytes from 12 people with nonsegmental vitiligo
      support a skin-compartment pan-p38 signal. The cultured-cell study does
      not resolve MAPK14 from other p38 isoforms or connect the signal to MDD.
  - reference: PMID:18575770
    reference_title: "Minocycline protects melanocytes against H2O2-induced cell death via JNK and p38 MAPK pathways."
    supports: SUPPORT
    evidence_source: IN_VITRO
    snippet: "Also, H2O2 treatment activates JNK and p38 MAPK, and executive caspase 3 in B10BR cells."
    explanation: >-
      Oxidative stress activates p38-family signaling in cultured mouse
      melanocytes, but the study is neither MAPK14-specific nor evidence of a
      shared human mechanism.

- description: >-
    A directional causal effect of liability to one disease on the other could
    produce the bidirectional temporal association. Current bidirectional
    Mendelian-randomization evidence found no detectable directional causal
    effect between generalized vitiligo and the study's broad depression
    phenotype, although it cannot exclude weak or subtype-specific effects.
  evidence:
  - reference: DOI:10.1111/exd.14979
    reference_title: "Exploring genetic associations between vitiligo and mental disorders using Mendelian randomization"
    supports: SUPPORT
    evidence_source: COMPUTATIONAL
    snippet: "In our findings, none of the rigorous bidirectional MR analyses uncovered a significant causal association."
    explanation: >-
      The null bidirectional MR result constrains genetically instrumented
      disease-to-disease effects between generalized vitiligo and the broad
      depression phenotype used in that analysis, rather than specifically
      adjudicating clinically defined MDD subgroups.

- description: >-
    Shared pleiotropic genetic liability could independently increase
    susceptibility to both MDD and vitiligo without either disease causing the
    other. Bidirectional Mendelian randomization does not test this common-cause
    model; cross-trait global and local genetic correlation followed by
    locus-level colocalization would be needed to distinguish shared variants
    from directional causality and linkage.

- description: >-
    Psychosocial effects of visible skin disease, healthcare-contact
    surveillance, treatment, and nonspecific chronic inflammatory burden may
    explain some or all of the association without a vitiligo-specific immune
    bridge to MDD.
  evidence:
  - reference: PMID:30528503
    reference_title: "Vitiligo and major depressive disorder: A bidirectional population-based cohort study."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "This risk was decreased in patients using antidepressants."
    explanation: >-
      The cohort reported an unquantified treatment-associated attenuation.
      Antidepressant pharmacology, indication, adherence, healthcare behavior,
      and residual confounding remain inseparable, so this observation cannot
      distinguish a biological mediator from treatment, psychosocial, or
      surveillance explanations.
  - reference: PMID:41781039
    reference_title: "Prevalence and comparative risk of mental health disorders in persons with vitiligo: a retrospective matched cohort study using claims data with expert-informed case validation."
    supports: SUPPORT
    evidence_source: HUMAN_CLINICAL
    snippet: "In matched comparisons with atopic dermatitis, only a few and inconsistent differences were observed."
    explanation: >-
      Similar mental-health profiles in another visible inflammatory skin
      disease favor nonspecific or psychosocial alternatives that must be tested
      against the shared-immune hypothesis.
Source:GitHub