The Relationship of Depression and Systemic Inflammation in Psoriasis: Findings from the UK Biobank.

Psoriasis is regarded as a systemic inflammatory disease; however, the significance of inflammatory burden for psychiatric comorbidity is not clear (Griffiths et al., 2017Griffiths C.E.M. Fava M. Miller A.H. Russell J. Ball S.G. Xu W. et al.Impact of ixekizumab treatment on depressive symptoms and systemic inflammation in patients with moderate-to-severe psoriasis: an integrated analysis of three phase 3 clinical studies.Psychother Psychosom. 2017; 86: 260-267Crossref PubMed Scopus (65) Google Scholar). A bidirectional association between psoriasis and depression is observed, negatively affecting patients’ disability, psoriasis course, and psoriatic arthritis risk (Lewinson et al., 2017Lewinson R.T. Vallerand I.A. Lowerison M.W. Parsons L.M. Frolkis A.D. Kaplan G.G. et al.Depression is associated with an increased risk of psoriatic arthritis among patients with psoriasis: a population-based study.J Invest Dermatol. 2017; 137: 828-835Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar). Mounting evidence of low-grade systemic inflammation in physically healthy individuals with depression led to the hypothesis that immunological processes may drive the psoriasis-depression comorbidity (Griffiths et al., 2017Griffiths C.E.M. Fava M. Miller A.H. Russell J. Ball S.G. Xu W. et al.Impact of ixekizumab treatment on depressive symptoms and systemic inflammation in patients with moderate-to-severe psoriasis: an integrated analysis of three phase 3 clinical studies.Psychother Psychosom. 2017; 86: 260-267Crossref PubMed Scopus (65) Google Scholar). However, there is little evidence that the mood of patients with psoriasis is associated with their inflammatory load. The correlations between C-reactive protein (CRP) decrease and depression improvement after ixekizumab were shown to be significant but weak (Pearson’s r = 0.11) (Griffiths et al., 2017Griffiths C.E.M. Fava M. Miller A.H. Russell J. Ball S.G. Xu W. et al.Impact of ixekizumab treatment on depressive symptoms and systemic inflammation in patients with moderate-to-severe psoriasis: an integrated analysis of three phase 3 clinical studies.Psychother Psychosom. 2017; 86: 260-267Crossref PubMed Scopus (65) Google Scholar); in another study (n = 53), CRP was associated with chronic stress only in women (Breuer et al., 2016Breuer K. Göldner F.M. Jäger B. Werfel T. Schmid-Ott G. Relationship between chronic stress and CRP levels in women with psoriasis.J Dtsch Dermatol Ges. 2016; 14: 528-530Google Scholar). Furthermore, depression is less consistently associated with objective psoriasis and psoriatic arthritis severity measures than with subjective symptom experience (Michelsen et al., 2017Michelsen B. Kristianslund E.K. Sexton J. Hammer H.B. Fagerli K.M. Lie E. et al.Do depression and anxiety reduce the likelihood of remission in rheumatoid arthritis and psoriatic arthritis? Data from the prospective multicentre NOR-DMARD study.Ann Rheum Dis. 2017; 76: 1906-1910Crossref PubMed Scopus (117) Google Scholar). Recent reports support a key role for neutrophils in psoriasis, not only in plaques but also in systemic inflammation and coexistent metabolic syndrome (Schön et al., 2017Schön M.P. Broekaert S.M. Erpenbeck L. Sexy again: the renaissance of neutrophils in psoriasis.Exp Dermatol. 2017; 26: 305-311Crossref PubMed Scopus (58) Google Scholar). High neutrophil counts, neutrophil-to-lymphocyte ratio (NLR), and neutrophil activation markers are found in patients (Paliogiannis et al., 2019Paliogiannis P. Satta R. Deligia G. Farina G. Bassu S. Mangoni A.A. et al.Associations between the neutrophil-to-lymphocyte and the platelet-to-lymphocyte ratios and the presence and severity of psoriasis: a systematic review and meta-analysis.Clin Exp Med. 2019; 19: 37-45Crossref PubMed Scopus (60) Google Scholar; Rocha-Pereira et al., 2004Rocha-Pereira P. Santos-Silva A. Rebelo I. Figueiredo A. Quintanilha A. Teixeira F. The inflammatory response in mild and in severe psoriasis.Br J Dermatol. 2004; 150: 917-928Crossref PubMed Scopus (212) Google Scholar); increased aged neutrophil levels interact with T cells, modulating IL-17 release, and decrease with anticytokine therapy (Rodriguez-Rosales et al., 2021Rodriguez-Rosales Y.A. Langereis J.D. Gorris M.A.J. van den Reek J.M.P.A. Fasse E. Netea M.G. et al.Immunomodulatory aged neutrophils are augmented in blood and skin of psoriasis patients.J Allergy Clin Immunol. 2021; 148: 1030-1040Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar). We cross-sectionally investigated associations of current and lifetime depression with neutrophil count, NLR, and high-sensitivity CRP in psoriasis, using the UK Biobank, a UK population-based cohort enrolling middle-to-older-aged adults (Smith et al., 2013Smith D.J. Nicholl B.I. Cullen B. Martin D. Ul-Haq Z. Evans J. et al.Prevalence and characteristics of probable major depression and bipolar disorder within UK Biobank: cross-sectional study of 172,751 participants.PLoS One. 2013; 8e75362Crossref Scopus (190) Google Scholar). Baseline data were used. Psoriasis was identified in linked primary care and hospital records and in self-reports where additional criteria were met (Supplementary Materials and Methods). Current depression levels were assessed with the Patient Health Questionnaire-2 (PHQ-2) (Löwe et al., 2005Löwe B. Kroenke K. Gräfe K. Detecting and monitoring depression with a two-item questionnaire (PHQ-2).J Psychosom Res. 2005; 58: 163-171Crossref PubMed Scopus (914) Google Scholar). We identified lifetime depression with the UK Biobank Smith et al. criteria for probable major depressive disorder (Smith et al., 2013Smith D.J. Nicholl B.I. Cullen B. Martin D. Ul-Haq Z. Evans J. et al.Prevalence and characteristics of probable major depression and bipolar disorder within UK Biobank: cross-sectional study of 172,751 participants.PLoS One. 2013; 8e75362Crossref Scopus (190) Google Scholar). Linear regression models were fitted; CRP and NLR were log-transformed. Models were adjusted for confounders, including a treatment-based classification (conventional systemics/biologics, topicals, untreated) as a psoriasis severity proxy (Supplementary Materials and Methods). Depression × sex interaction terms were entered, based on earlier evidence of modifying gender effects (Ye et al., 2021Ye Z. Kappelmann N. Moser S. Davey Smith G.D. Burgess S. Jones P.B. et al.Role of inflammation in depression and anxiety: tests for disorder specificity, linearity and potential causality of association in the UK Biobank.EClinicalMedicine. 2021; 38100992Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar), and were dropped if nonsignificant. A total of 5,485 patients with psoriasis were included for analysis (5,184 had PHQ-2 data; 4,796 had lifetime depression data) (Table 1). Using Mann-Whitney U tests, patients on systemic treatment exhibited higher CRP, NLR, and neutrophil counts than patients on topicals (Pcrp < 0.001, Pneutrophil count = 0.02, PNLR < 0.001) or untreated patients (P < 0.001; all markers), suggesting higher systemic disease activity, as expected (Supplementary Materials and Methods).Table 1Baseline Characteristics of UK Biobank Participants with Psoriasis Included in the AnalysesCharacteristicsPsoriasis with Lifetime Depression (n = 462)Psoriasis without Lifetime Depression (n = 4,334)P-valuePsoriasis, Total (n = 5,485)1Including cohort with available Patient Health Questionnaire-2 but no case-control lifetime depression data (missing or borderline cases).Sex, female267 (57.79%)1,920 (44.30%)<0.0012,604 (47.47%)Age, y64.01 (7.93)65.69 (7.97)<0.00165.54 (7.96)Ethnicity, white444 (96.10%)4,184 (96.54%)0.065,302 (96.66%)Body mass index, kg/m228.69 (5.51)28.19 (4.83)0.0628.41 (5.05)C-reactive protein, mg/l, median (IQR)1.89 (3.15)1.70 (2.74)0.0551.77 (2.89)Neutrophil counts, 109 cells/l, median (IQR)4.52 (2.09)4.27 (1.77)<0.0014.30 (1.80)Neutrophil-to-lymphocyte ratio, median (IQR)2.29 (1.26)2.31 (1.24)0.742.30 (1.24)Treatment0.24 Biologics/conventional systemics34 (7.36%)276 (6.37%)359 (6.54 %) Topicals154 (33.33%)1,608 (37.10%)1,984 (36.17%) No treatment274 (59.31%)2,450 (56.53%)3,142 (57.28%)Psoriasis duration, y2Data available in 3,079 (self-reported onset).25.46 (16.94)26.09 (16.19)0.5826.02 (16.37)Psoriatic arthritis37 (8.01%)385 (8.88%)0.59514 (9.37%)Patient Health Questionnaire-2 scores3.20 (1.67)2.49 (0.98)<0.0012.66 (1.20)Hypertension and other cardiovascular disorders165 (35.71%)1,535 (35.42%)0.941,999 (36.44%)Other major comorbidities3Other systemic inflammatory or neuroinflammatory disorders, including malignancy.207 (44.81%)1,461 (33.71%)<0.0012,012 (36.68%)Alcohol consumption0.009 Frequent drinker196 (42.42%)1,992 (45.96%)2,441 (44.50%) Occasional drinker159 (34.42%)1,589 (36.66%)1,972 (35.95%)Smoker87 (18.83%)610 (14.07%)0.007813 (14.82%)Townsend Deprivation Index4Higher scores indicate higher material deprivation for the area of participant residence at enrollment.−0.53 (3.11)−1.23 (3.09)<0.001−1.09 (3.14)Abbreviation: IQR, interquartile range.Numerical variables reported as mean (SD) unless otherwise specified. P-values relate to χ2Data available in 3,079 (self-reported onset)., t-tests or Mann-Whitney U tests, as appropriate.1 Including cohort with available Patient Health Questionnaire-2 but no case-control lifetime depression data (missing or borderline cases).2 Data available in 3,079 (self-reported onset).3 Other systemic inflammatory or neuroinflammatory disorders, including malignancy.4 Higher scores indicate higher material deprivation for the area of participant residence at enrollment. Open table in a new tab Abbreviation: IQR, interquartile range. Numerical variables reported as mean (SD) unless otherwise specified. P-values relate to χ2Data available in 3,079 (self-reported onset)., t-tests or Mann-Whitney U tests, as appropriate. We found no association of CRP or NLR with current or lifetime depression in psoriasis, although we noted a trend toward a lifetime depression effect on CRP. CRP-depression associations were significant when body mass index was omitted from models (Supplementary Materials and Methods). An effect modification by sex was observed for between neutrophils and PHQ-2 (Table 2). The association of PHQ-2 scores with neutrophil count was greater and significant only for women, even after accounting for lifetime depression. Lifetime depression was associated with higher neutrophil counts in both sexes.Table 2Associations of Depression with CRP, Neutrophil Counts, and NLR in Patients with PsoriasisModelsPredictor VariablesOutcome VariablesCRP1Log-transformed.NeutrophilsNLR1Log-transformed.β (95% CI)P-Valueβ (95% CI)P-Valueβ (95% CI)P-ValueModel 1: Current depressionCurrent depression2Median-centered Patient Health Questionnaire-2 score. × sex interaction term−0.07 (−0.14 to −0.005)0.04Sex0.21 (0.01 to 0.40)0.04Current depression2Median-centered Patient Health Questionnaire-2 score.0.06 (0.01 to 0.11)0.009Current depression coefficients by sex:Women (reference)0.06 (0.01 to 0.11)3Stratified models by sex yielded similar results, showing significant associations among women only.Men−0.01 (−0.06 to 0.04)3Stratified models by sex yielded similar results, showing significant associations among women only.Model 2: Current depression4Models 2 and 3 do not include depression × sex interaction terms; interactions were not significant at 0.05 level and were dropped.Current depression2Median-centered Patient Health Questionnaire-2 score.0.01 (−0.01 to 0.03)0.42−0.001 (−1.09 to 0.01)0.88Model 3: Lifetime depression4Models 2 and 3 do not include depression × sex interaction terms; interactions were not significant at 0.05 level and were dropped.Lifetime depression0.08 (−0.01 to 0.18)0.070.26 (0.12 to 0.40)<0.0010.01 (−3.11 to 0.05)0.64Model 4: Current depression, lifetime-depression-adjusted (in women only)5Post hoc model for the significant result; it did not include patients not meeting case-control criteria for lifetime depression.Current depression2Median-centered Patient Health Questionnaire-2 score.0.06 (0.004 to 0.12)0.03Lifetime depression0.24 (0.05 to 0.43)0.01Abbreviations: CI, confidence interval; CRP, C-reactive protein; NLR, neutrophil-to-lymphocyte ratio.All models were adjusted for sex, age, ethnicity, body mass index, alcohol consumption (frequent, occasional, or rare), smoking, Townsend Deprivation Index, psoriatic arthritis, cardiovascular comorbidity, other systemic inflammatory or neuroinflammatory disease or malignancy, and psoriasis severity (treatment-based classification). Patients with zero leukocytes/lymphocytes were excluded. The variance inflation factor was <1.5 for all variables in models, excluding interactions (∼2.2); no influential outliers were found. Unadjusted estimates and sensitivity analyses are reported in the Supplementary Materials and Methods; omitting the severity variable did not change results.1 Log-transformed.2 Median-centered Patient Health Questionnaire-2 score.3 Stratified models by sex yielded similar results, showing significant associations among women only.4 Models 2 and 3 do not include depression × sex interaction terms; interactions were not significant at 0.05 level and were dropped.5 Post hoc model for the significant result; it did not include patients not meeting case-control criteria for lifetime depression. Open table in a new tab Abbreviations: CI, confidence interval; CRP, C-reactive protein; NLR, neutrophil-to-lymphocyte ratio. All models were adjusted for sex, age, ethnicity, body mass index, alcohol consumption (frequent, occasional, or rare), smoking, Townsend Deprivation Index, psoriatic arthritis, cardiovascular comorbidity, other systemic inflammatory or neuroinflammatory disease or malignancy, and psoriasis severity (treatment-based classification). Patients with zero leukocytes/lymphocytes were excluded. The variance inflation factor was <1.5 for all variables in models, excluding interactions (∼2.2); no influential outliers were found. Unadjusted estimates and sensitivity analyses are reported in the Supplementary Materials and Methods; omitting the severity variable did not change results. Our findings suggest an association of current depressive burden with systemic inflammation in the form of increased neutrophils in psoriasis, only among women. These sex effects align with previous reports on CRP, both in patients with psoriasis (Breuer et al., 2016Breuer K. Göldner F.M. Jäger B. Werfel T. Schmid-Ott G. Relationship between chronic stress and CRP levels in women with psoriasis.J Dtsch Dermatol Ges. 2016; 14: 528-530Google Scholar) and the general population (Ye et al., 2021Ye Z. Kappelmann N. Moser S. Davey Smith G.D. Burgess S. Jones P.B. et al.Role of inflammation in depression and anxiety: tests for disorder specificity, linearity and potential causality of association in the UK Biobank.EClinicalMedicine. 2021; 38100992Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar). Higher systemic inflammation is consistently found in women, may be linked to hormonal physiology and visceral adiposity, and predicts poorer antidepressant outcomes in women (Jha et al., 2019Jha M.K. Minhajuddin A. Chin-Fatt C. Greer T.L. Carmody T.J. Trivedi M.H. Sex differences in the association of baseline C-reactive protein (CRP) and acute-phase treatment outcomes in major depressive disorder: findings from the EMBARC study.J Psychiatr Res. 2019; 113: 165-171Crossref PubMed Scopus (31) Google Scholar). Interestingly, however, we did not find similar results for CRP. Lifetime depression predicted higher neutrophil counts in psoriasis regardless of sex, and also when accounting for current depression levels in women. For the majority of patients, major depressive disorder is a chronic relapsing-remitting condition with evidence of ongoing biochemical abnormalities, including elevated serum proinflammatory markers, even when the mood is normal (Dittrich et al., 2021Dittrich K. Boedeker K. Kluczniok D. Hindi Attar C.H. Winter S.M. Roepke S. et al.Elevated inflammatory markers in women with remitted major depressive disorder and the role of early life maltreatment.Brain Behav Immun. 2021; 97: 219-225Crossref PubMed Scopus (3) Google Scholar). Our observed lack of association between depression and CRP adds to the limited, inconsistent evidence in psoriasis (Breuer et al., 2016Breuer K. Göldner F.M. Jäger B. Werfel T. Schmid-Ott G. Relationship between chronic stress and CRP levels in women with psoriasis.J Dtsch Dermatol Ges. 2016; 14: 528-530Google Scholar; Griffiths et al., 2017Griffiths C.E.M. Fava M. Miller A.H. Russell J. Ball S.G. Xu W. et al.Impact of ixekizumab treatment on depressive symptoms and systemic inflammation in patients with moderate-to-severe psoriasis: an integrated analysis of three phase 3 clinical studies.Psychother Psychosom. 2017; 86: 260-267Crossref PubMed Scopus (65) Google Scholar) and supports findings in psoriatic arthritis (Michelsen et al., 2017Michelsen B. Kristianslund E.K. Sexton J. Hammer H.B. Fagerli K.M. Lie E. et al.Do depression and anxiety reduce the likelihood of remission in rheumatoid arthritis and psoriatic arthritis? Data from the prospective multicentre NOR-DMARD study.Ann Rheum Dis. 2017; 76: 1906-1910Crossref PubMed Scopus (117) Google Scholar). The reason for the discrepancy between neutrophils and CRP/NLR is unclear. In psoriasis, NLR does not appear to mirror disease severity (Paliogiannis et al., 2019Paliogiannis P. Satta R. Deligia G. Farina G. Bassu S. Mangoni A.A. et al.Associations between the neutrophil-to-lymphocyte and the platelet-to-lymphocyte ratios and the presence and severity of psoriasis: a systematic review and meta-analysis.Clin Exp Med. 2019; 19: 37-45Crossref PubMed Scopus (60) Google Scholar). Although CRP correlates with depression in the general population (Ye et al., 2021Ye Z. Kappelmann N. Moser S. Davey Smith G.D. Burgess S. Jones P.B. et al.Role of inflammation in depression and anxiety: tests for disorder specificity, linearity and potential causality of association in the UK Biobank.EClinicalMedicine. 2021; 38100992Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar), it may be a less sensitive marker among patients with psoriasis or other immune-mediated diseases. In addition, several depression-associated serum chemokine increases observed in physically healthy patients are not detected among individuals with inflammatory illnesses (Leighton et al., 2018Leighton S.P. Nerurkar L. Krishnadas R. Johnman C. Graham G.J. Cavanagh J. Chemokines in depression in health and in inflammatory illness: a systematic review and meta-analysis.Mol Psychiatry. 2018; 23: 48-58Crossref PubMed Scopus (147) Google Scholar). One possible explanation for these and our findings is the small relative contributions of depression to some inflammatory markers compared with the underlying disease (Leighton et al., 2018Leighton S.P. Nerurkar L. Krishnadas R. Johnman C. Graham G.J. Cavanagh J. Chemokines in depression in health and in inflammatory illness: a systematic review and meta-analysis.Mol Psychiatry. 2018; 23: 48-58Crossref PubMed Scopus (147) Google Scholar). Furthermore, obesity may exert mediator or collider effects in addition to confounding, although its role remains obscure. Depression is considered markedly heterogeneous regarding its biological and immunometabolic phenotypes, found to affect antidepressant treatment response (Milaneschi et al., 2020Milaneschi Y. Lamers F. Berk M. Penninx B.W.J.H. Depression heterogeneity and its biological underpinnings: toward immunometabolic depression.Biol Psychiatry. 2020; 88: 369-380Abstract Full Text Full Text PDF PubMed Scopus (118) Google Scholar). Neutrophils and putatively proinflammatory cytokines key to psoriasis, such as IL-17, may better reflect its inflammatory interplay with mood and constitute promising targets for tailored management in comorbid patients. To our knowledge, this is the largest report investigating the association between depression and systemic inflammation in psoriasis. The use of a multicenter, national cohort, and PHQ-2 as a validated tool adds to the study’s validity. Limitations include the misclassification potential for psoriasis and the lack of clinician-rated psoriasis severity assessments. Furthermore, the overall mild psoriasis severity and middle-to-old age range may limit the generalizability of findings in broader psoriasis populations, whereas the cross-sectional design does not allow for determining directionality. Finally, although the depression coefficients are significant, they are relatively small, indicating that caution is required in the interpretation of our results and that future replication is warranted. Taken together, however, the observed associations and their independence from lifestyle factors, comorbidities, and treatment-based psoriasis severity are suggestive of underlying immunological relationships with mood, particularly among women. It is important to further elucidate these associations in psoriasis through large prospective studies including multiple inflammatory markers. Data may be obtained from a third party and are not publicly available. Data used in the study can be accessed following application through the UK Biobank Access Management System (www.ukbiobank.ac.uk/register-apply) and approval by the UK Biobank. Georgia Lada: http://orcid.org/0000-0003-1926-2599 Hector Chinoy: http://orcid.org/0000-0001-6492-1288 Peter S. Talbot: http://orcid.org/0000-0003-3492-0801 Richard B. Warren: http://orcid.org/0000-0002-2918-6481 C. Elise Kleyn: http://orcid.org/0000-0003-4038-5833 CEK has received honoraria and has served as consultant and/or received research funding from Janssen, Eli Lilly, LEO, Novartis, AbbVie, UCB, Almirall, Pfizer, and L’Oréal. HC has received personal compensation for activities with Novartis, UCB, Lilly, Biogen, Orphazyme as a speaker, advisory board member, or consultancy; grants through The University of Manchester from Novartis, UCB, and MedImmune, and has received travel support from AbbVie and Janssen. RBW has received research grants from AbbVie, Almirall, Amgen, Celgene, Janssen, Lilly, Leo, Medac, Novartis, Pfizer, and UCB and consulting fees from AbbVie, Almirall, Amgen, Arena, Astellas, Avillion, Biogen, Boehringer Ingelheim, Bristol Myers Squibb, Celgene, DiCE, GalaxoSmithKline, Janssen, Eli Lilly, Leo, Medac, Novartis, Pfizer, Sanofi, Sun Pharma, UCB, and UNION. GL has received speaker honoraria from Janssen, Lilly, Leo, and Novartis. PST has no competing interests to report. This research has been conducted using the UK Biobank resource, a major biomedical database, under Application 63178 (https://www.ukbiobank.ac.uk). GL, CEK, RBW, and HC are supported by the NIHR Manchester Biomedical Research Centre Funding Scheme. The views expressed in this publication are those of the authors and not necessarily those of the NHS, the NIHR, or the Department of Health. Conceptualization: GL, PST, CEK, HC, RBW; Data Curation: GL; Formal Analysis: GL; Funding Acquisition: CEK, RBW, HC; Investigation: GL; Methodology: GL, PST, CEK; Project Administration: GL, CEK; Resources: GL, HC, PST, RBW, CEK; Software: GL, Supervision: CEK, PST, HC, RBW; Visualization: GL; Writing - Original Draft Preparation: GL; Writing - Review and Editing: GL, HC, PST, RBW, CEK A detailed description of methods and sensitivity analyses is provided below. The UK Biobank is a national, multicenter resource that collects a wide range of health data (Sudlow et al., 2015Sudlow C. Gallacher J. Allen N. Beral V. Burton P. Danesh J. 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. 2015; 12e1001779Crossref PubMed Scopus (3820) Google Scholar). About half a million adults, aged 40 to 69 years, enrolled in 22 sites across the United Kingdom between 2006 and 2010 (Sudlow et al., 2015Sudlow C. Gallacher J. Allen N. Beral V. Burton P. Danesh J. 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. 2015; 12e1001779Crossref PubMed Scopus (3820) Google Scholar). All participants provided written informed consent; data from participants who withdraw are excluded from analysis. The UK Biobank received approval by the North West Multi-centre Research Ethics Committee, reference: 16/NW/0274 (https://www.ukbiobank.ac.uk). Participants attended a baseline visit, where medical history data were obtained through self-reported questionnaires and interviews, and blood samples (blood counts and selected biochemistry markers) were taken (Sudlow et al., 2015Sudlow C. Gallacher J. Allen N. Beral V. Burton P. Danesh J. 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. 2015; 12e1001779Crossref PubMed Scopus (3820) Google Scholar). Hematology data were analyzed on a Beckman Coulter, and high-sensitivity serum C-reactive protein (CRP) assays were analyzed on a Beckman Coulter AU5800 using immunoturbidimetry. Primary care data for about half of the cohort and hospital admission data are available through linkage. Our criteria for identifying psoriasis in the UK Biobank were psoriasis in primary care or hospital admission records at baseline, or self-reported psoriasis at baseline, if at least one of the following was present: psoriasis-indicated treatment, psoriatic arthritis, or physician-coded psoriasis at any instance. Self-reports were included to reduce linkage missingness–associated selection bias. Lifetime depression/major depressive disorder was defined with the widely used Smith et al. criteria for probable major depressive disorder in the UK Biobank (Milton et al., 2021Milton D.C. Ward J. Ward E. Lyall D.M. Strawbridge R.J. Smith D.J. et al.The association between C-reactive protein, mood disorder, and cognitive function in UK Biobank.Eur Psychiatry. 2021; 64: e14Crossref PubMed Scopus (11) Google Scholar; Smith et al., 2013Smith D.J. Nicholl B.I. Cullen B. Martin D. Ul-Haq Z. Evans J. et al.Prevalence and characteristics of probable major depression and bipolar disorder within UK Biobank: cross-sectional study of 172,751 participants.PLoS One. 2013; 8e75362Crossref Scopus (191) Google Scholar). Major depressive disorder controls did not meet these criteria and did not have self-reported or physician-coded depression or antidepressant treatment at baseline. Lifetime depression, the Patient Health Questionnaire-2, and blood data were collected at baseline. To investigate associations of depression with inflammation in psoriasis, we used linear regression, controlling for sex, age, ethnicity, body mass index, alcohol consumption (frequent, occasional, or rare), smoking, Townsend Deprivation Index, a treatment-based psoriasis severity proxy, psoriatic arthritis, cardiovascular comorbidity, and presence of other systemic inflammatory or neuroinflammatory disease or malignancy. The Townsend Deprivation Index is a composite material deprivation index that incorporates noncar ownership, nonhome ownership, overcrowding, and unemployment; higher scores indicate higher deprivation for the area of participants’ residence (Townsend et al., 1988Townsend P, Phillimore P, Beattie A. Health and deprivation: inequality and the north. Kent: Croom Helm; 1988.Google Scholar; Yousaf and Bonsall, 2017Yousaf S. Bonsall A. UK Townsend Deprivation Scores from 2011 census data. UK Data Service, Colchester, UK2017Google Scholar). There were no influential outliers in any model. In the first of a sensitivity analysis series, after removing all values outside of median ± 3 interquartile range, our results did not change. Of 425,933 UK Biobank participants with inflammatory marker and depression (Patient Health Questionnaire-2 or major depressive disorder) data, 5,549 patients were identified as having psoriasis at baseline; of these, 5,485 patients had data for covariates and were included in the analytical models: 5,184 in the models for current depression (Patient Health Questionnaire-2) and 4,796 in the models for lifetime depression. Although covariate missingness was low (1.2%), we ran a sensitivity analysis using multiple imputation to also account for the missing predictors, considering data to be missing at random; the results in the imputed sample were similar in magnitude and significance. Furthermore, given that systemic inflammation may be less pronounced in mild psoriasis (Beygi et al., 2014Beygi S. Lajevardi V. Abedini R. C-reactive protein in psoriasis: a review of the literature.J Eur Acad Dermatol Venereol. 2014; 28: 700-711Crossref PubMed Scopus (76) Google Scholar), we confirmed that, compared with the rest of the UK Biobank, the psoriasis sample in this study presented with higher systemic inflammation, when controlling for confounders of interest (listed under Statistical analysis: confounders and outlying values) and lifetime depression: β(logCRP) 0.18, 95% confidence interval (95% CI) = 0.15–0.20; β(log neutrophil-to-lymphocyte ratio [NLR]) 0.07, 95% CI = 0.06–0.08; β(neutrophil count) 0.19, 95% CI = 0.15–0.23; P < 0.001 for all markers. For transparency, we report unadjusted current depression coefficients for log(CRP): 0.05 (95% CI = 0.03–0.08), P < 0.001; log(NLR): 0.005 (95% CI = −0.008 to 0.11), P = 0.79; and neutrophil count: 0.10 (95% CI = 0.07–0.14), P < 0.001 (for lifetime depression statistics, see Table 1). Before using treatment as a severity proxy covariate in the main analytical models, we explored its relationship to the inflammatory markers and predictors (depression). We found significant differences among the three treatment groups for neutrophil counts, NLR, and CRP (P < 0.001), using Kruskal-Wallis tests. Mann-Whitney U tests showed that patients on systemic treatments had higher marker levels than patients on topical treatments (Pcrp < 0.001, Pneutrophil count = 0.02, PNLR < 0.001) and untreated patients (P < 0.001; all markers). Patients on topicals had higher neutrophils (P = 0.01) and NLR (P < 0.001), but not CRP (P = 0.61), compared with untreated patients. Regarding the predictors, although there was no significant difference in lifetime depression prevalence (P = 0.24) among the three treatment groups, there was a difference in current depression (mean [SD] Patient Health Questionnaire-2 scores for the systemics group: 2.8 [1.3], the topical group: 2.7 [1.3], and the untreated group: 2.6 [1.1], P[Kruskal-Wallis] = 0.026). Finally, we performed a sensitivity analysis to determine whether adjustment for psoriasis severity may have reduced the magnitude of the association of depression with inflammation. We re-ran the models without adjustment for the treatment/severity variable and found similar results in effect size and significance for all three markers and both depression outcomes. In Supplementary Figure S1, we provide a visualization of the data across treatment groups for the significant model (neutrophil count) for the interested reader. Obesity increases both the risk for depression and inflammation and is expected to confound their association. However, the role of obesity may be complex (Capuron et al., 2017Capuron L. Lasselin J. Castanon N. Role of adiposity-driven inflammation in depressive morbidity.Neuropsychopharmacology. 2017; 42: 115-128Crossref PubMed Scopus (103) Google Scholar), and regressing out obesity may partly obscure potential mediating or collider effects. Similar to the sensitivity analysis for the treatment-based measure, we conducted an additional analysis where we removed body mass index as a covariate from our models; of note, effects of depression on log(CRP) became statistically significant (current depression: β = 0.03, 95% CI = 0.01–0.05, P = 0.01; lifetime depression: β = 0.11, 95% CI = 0.01–0.21, P = 0.03), whereas results for NLR did not gain in statistical significance.