Association of inflammation and serum adipocytokines with depression and anxiety burden in axial spondyloarthritis Authors: Markéta Hušáková, Karel Pavelka, Karolína Kuchaříková, Jan Kotrs, Markéta Kuklová, Hana Hulejová, Ladislav Šenolt, Pavel Mohr, Marta Olejárová Published: 2026 ABSTRACT AbstractTo investigate the relationships between serum adipocytokines, body mass index (BMI), disease activity, and symptoms of depression and anxiety in patients with axial spondyloarthritis (axSpA). Ninety-three patients with axSpA and 100 age- and sex-matched healthy controls were assessed. Serum adipocytokines (resistin, leptin, fetuin-A, visfatin) and ghrelin, BMI, and disease activity (BASDAI, ASDAS-CRP) were measured. Depression and anxiety were evaluated using the Beck Depression Inventory-II (BDI-II) and Beck Anxiety Inventory (BAI). Patients with axSpA exhibited higher BDI-II (6.76 ± 6.81 vs 2.81 ± 3.31, p < 0.001) and BAI scores (6.67 ± 7.71 vs 3.89 ± 3.88, p < 0.05) than controls, with the somatic domain most pronounced. Depressive and anxiety symptoms correlated strongly with disease activity (ASDAS-CRP and BASDAI) but not with BMI or spinal involvement. Leptin and ghrelin reflected BMI, whereas fetuin-A and resistin were independently associated with affective symptoms: doubling of fetuin-A increased BDI-II and BAI scores by ~ 4 and 3 points, respectively, whereas doubling of resistin decreased them by ~ 2.5 points. Resistin was inversely correlated with BASDAI (r = –0.33, p < 0.05) and anxiety (r = –0.42, p < 0.001). Other adipocytokines showed limited associations with clinical or psychological measures. Depressive and anxiety symptoms in axSpA are primarily driven by inflammatory burden rather than BMI or structural disease severity. Fetuin-A and resistin exert opposing effects on affective symptoms, highlighting disease-specific immunometabolic pathways. Early identification of psychological distress and integrated management with anti-inflammatory therapy are warranted.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-51470-6.Keywords: Axial spondyloarthritis, Depression, Anxiety, Inflammation, CRP, AdipocytokinesSubject terms: Biomarkers, Diseases, Immunology, Medical research Introduction Axial spondyloarthritis (axSpA) is characterized by inflammation of spine, peripheral joints, sacroiliac joints (SIJ) resulting in a chronic, painful condition associated with impaired mobility1. Beyond musculoskeletal manifestations, the burden of axSpA encompasses a range of comorbidities, depression and anxiety being particularly prevalent and exerting a negative impact on disease outcomes2,3. The mental burden in patients with axSpA is multifactorial, with biological mechanisms likely contributing alongside psychosocial stress related to disability. Depression is thought to arise from several interrelated mechanisms, including reduced neuroplasticity, impaired neural network communication, monoaminergic dysfunction, dysregulation of the hypothalamic–pituitary axis, and the interaction of neurotrophins, neuronal cells, and signaling pathways with inflammatory mediator4. Furthermore, ongoing research on the gut–brain and adipose–brain axes aims to elucidate the influence of adipocytokines, gastrointestinal (GI) microbiota, and gut-derived peptides on the development of depression and anxiety5. Adipocytokines (leptin, adiponectin, visfatin, resistin) secreted primarily by white adipose tissue regulate food intake and energy balance. Fetuin-A is adipocytokine and hepatokine with adipogenic effect, that participates in several biological function such as bone metabolism, insulin signaling and inflammation6. Ghrelin, a peptide from the gastrointestinal tract counteracts leptin in the eating behaviour thereby contributing to the maintenance of body weight. Beyond metabolism, adipocytokines and ghrelin may act via paracrine and autocrine pathways in the immune and neuronal system and their their role in the pathogenesis of the affective diseases is currently under consideration7. Recently, elevated serum leptin levels have been reported in individuals with recurrent major depressive disorder (MDD)8, MDD with somatic signs of anxiety9, and in patients with type 2 diabetes and comorbid anxiety10. In addition, serum levels of fetuin-A, resistin8, ghrelin11 was found to be higher in people with depression. Although the mechanisms are not yet clear, leptin may cross the blood–brain barrier and bind to receptors that are widely distributed, including those on dopaminergic neurons12, and in experimental models may altered serotonin synthesis with behavioural consequences13. In axSpA, adipocytokines are believed to influence both disease activity and structural progression through their immunomodulatory properties. Leptin, in particular, has been shown to promote the proliferation of Th1 and Th17 cells, activate macrophages, and sustain a pro-inflammatory cytokine network characterised by interleukin (IL)-1β, IL-6, and tumour necrosis factor (TNF)-α.14. In patients with axSpA, lower levels of leptin were observed in those who achieved remission or low disease activity compared to those who did not achieve remission after six months of treatment with TNF inhibitors (TNFi)15. In contrast, fetuin-A, resistin, and visfatin are elevated, with visfatin correlating with Ankylosing Spondylitis (AS) Disease Activity Scores (ASDAS), Bath AS Disease Activity Index (BASDAI), and spinal progression, underscoring the distinct roles of adipocytokines in axSpA16–18. Given the close association between disease activity and mood or anxiety burden in axSpA, we aimed to determine whether adipocytokines (leptin, resistin, visfatin, fetuin-A) and ghrelin, body mass index (BMI), or disease activity influence the psychological burden in these patients. This approach allows us to disentangle potential biological, metabolic, and disease-related contributors to mental health in axSpA. Methods Participants’ characteristics People with axSpA and healthy counterparts (HC), aged between 18 and 65 were included, all of them were participants of RHEU-THYMIA study19. The written inform consent was required from all participants prior all study procedures. The study was conducted in compliance with the Declaration of Helsinki and approved by Ethical Committee of Rheumatology Institute in Prague (number 5704/2021). The following data were gathered from all participants: sex, body mass index (BMI), comorbidities, pharmacotherapy, alcohol abuse and physical activity in accordance with the World heath association (WHO)’s weekly recommendations and five dimensional EuroQol questionnaire (EQ-5D) for quality of life20. The below exclusion criteria were applied: fibromyalgia (FM), other autoimmune disorders, chronic decompensated diseases e.g. type 2 diabetes, chronic lung, kidney and heart disease, malignancy, drug abuse, current infection, injuries or surgical care within the last 6 weeks, pregnancy, breastfeeding, and pharmacological therapy included corticosteroids, opioid analgesics, any hormonal therapy including hormonal contraceptive or hormonal substitutive therapy, anti-seizure or anti-Parkinson’s disease drugs, monoclonal antibodies, cytokine analogues or recombinant proteins for any reason other than axSpA, or any experimental therapy. People without any antidepressive or anxiolytic drug were included in this cohort. All axSpA patients were required to fulfil the Assessment of Spondyloarthritis International Society (ASAS) classification criteria21. Based on sacroiliac joint (SIJ) radiographic findings, patients were classified as having r-axSpA or nr-axSpA. Because the data of modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS) was not available, axSpA patients were categorically divided (SIJ limited/Spine involvement) regarding the presence of any spinal syndemophyte. Characterisation of axSpA included current and previous treatments, disease duration, and the presence of inflammatory back pain (IBP), arthritis (in any of 44 joints), enthesitis (in any locations defined by the Maastricht AS enthesitis criteria)22, Bath AS metrology index (BASMI) and ASDAS with C-reactive protein (-CRP)23. Patient-reported outcomes (PROs) included the Bath Ankylosing Spondylitis Disease Activity Index and the Bath AS Functional Index (BASFI). Blood sampling and analyses Blood samples were collected in the morning (08:00–12:30) following overnight fasting. Due to the potentially long travel required for some participants, a light breakfast—consisting of unsweetened tea and up to 20 g of bread or a roll—was permitted, provided it was consumed no earlier than two hours before sampling. The serum levels of CRP using were analysed on the same day. All samples were immediately centrifuged, after which the serum was stored at -70 °C until analysis. All biomarkers were analysed after completion of sample collection, and assays were performed in batches to minimise potential bias. Serum levels of leptin, resistin, visfatin, and ghrelin were measured using the Bio-Plex Pro Human Diabetes 10-Plex Panel (Bio-Rad, CA, USA) according to the manufacturer’s instructions. The inter-assay and intra-assay coefficients of variation were ≤ 25% and ≤ 15%, respectively, and assay specificity was < 10% cross-reactivity between analytes.The serum levels of Fetuin A were analysed by commercially available RayBio Human Fetuin A SpeedELISA Kit, according to the manufacturer’s protocol (RayBiotech, Norcross, GA, United States). The minimum detectable dose of Human Fetuin A was determined to be 0,31 ng/ml. Absorbance at a primary wavelength of 450 nm was detected using a Sunrise ELISA reader (Tecan Group Ltd., Zürich, Switzerland), and the data were quantified using Kim version 5.43.01 software (Daniel Kittrich—Software Production, Prague, Czech Republic). Mental health assessement The validated Czech versions of the Beck Depression Inventory (BDI-II) and the Beck Anxiety Inventory (BAI) were utilised24–27. Both inventories are self-reported instruments designed to capture respondents’ subjective perceptions of depressive or anxiety symptoms, and each consists of 21 items rated on a four-point Likert scale (0–3). The BDI-II includes somatic, cognitive, and affective signs of depression, while the BAI covers anxiety symptoms such as dizziness, trembling, facial flushing, catastrophic thinking, and fear of death or loss of control. BDI-II scores indicate the following categories: no depression (0–13), mild (14–19), moderate (20–28), and severe (29–63) depression. BAI scores indicate the following categories: no anxiety (0–9), mild (10–18), moderate (19–29), and severe (30–63) anxiety27,28. Statistical analysis For descriptive statistical analysis continuous variables were presented as means with 95% confidence intervals (CI) or medians with upper and lower quartiles while discrete variables as frequencies and percentages. Healthy controls were randomly frequency-matched to patient group by age and sex. The t-test was used for univariate subgroup comparisons while the Wilcoxon exact or Mann–Whitney tests for non-parametric data. Correlations were performed using Spearman’s correlation coefficient. A p-value of less than 0.05 was considered statistically significant. To explore biomarker level contributions towards various outcomes (ASDAS-CRP level, BDI-II or BAI scores) we have employed linear regression modelling. Given the exploratory rather than predictive goal all models were constructed using all measured biomarkers. Mild heteroskedasticity was detected (using LOWESS curve shape diagnostics in scale-location plots as attached) thus we have refitted them with weights computed as the inverse of the squared residuals as a commonly used heuristic (WLS). All biomarkers were log₂-transformed prior to modelling to account for non-linearity and to enable interpretation of effects as the change associated with a doubling of the biomarker level.Analyses were conducted using Python libraries: statsmodels (v0.14.1) and scikit-learn (v1.3.2)29,30. Coefficient visualization was done using forestplot (v0.3.3), and data processing with pandas (v2.1.4) and numpy (v1.26.3)31. Results Sociodemographic and axSpA clinical variables A total of 93 patients with axSpA and 100 age- and sex-matched healthy controls (HC) were included. Baseline characteristics are presented in Table 1. The groups differed in EQ-5D score and city population. Stratified by sex, the EQ-5D score was lower in axSpA patients than in HC in both sexes (Supplementary Table 1). No sex differences were observed in disease activity or other patient-reported outcomes (PROs) among axSpA patients. Males had a lower proportion of nr-axSpA than females (28.8% vs. 55.9%) and a higher proportion of r-axSpA (71.2% vs. 44.1%; p < 0.05). Similarly, hip and spine involvement were more frequent in males than in females (33.9% vs. 11.8%, p < 0.01; and 49.1% vs. 11.8%, p < 0.0001, respectively). Table 1Demographic data and axSpA features.FeatureAxSpAHCn = 93n = 100Age, median (IQR)41.00 (34.00—48.00)44.50 (31.00—52.25)Sex—Female/Male, %36.60/63.4036.00/64.00BMI, median (IQR)25.69 (23.46—28.69)26.81 (23.31—29.80)Marital status—1/2, %29.00/71.0038.00/62.00City population—< 100,000/ > 100,000, %66.70/33.30*46.00/54.00*Work status 1/2/3, %87.10/9.70/3.2083.00/6.00/11.00Comorbidities—0/1–2/ > 2 tracked dis./, %54.80/37.60/7.5056.00/36.00/8.00Physical activity according to WHO°—YES, %68.8067.00Alcohol intake according to WHO°°—Lower/Equivalent/Higher, %79.60/19.30/1.1080.00/15.11/5.005D-EQ, median (IQR)0.80 (0.69—0.85)**1.00 (0.80—1.00)**CRP mg/l2.84 (1.29—7.20)NAASDAS-CRP < 1.4 inactive disease, %26.90NAASDAS-CRP active disease: 1.4–2.0/2.1–3.4/ > 4.0, %37.60/24.70/10.80NABASDAI, median (IQR)2.60 (1.40—4.20)NABASFI, median (IQR)1.30 (0.30—3.10)NABASMI, median (IQR)2.40 (2.00—3.40)NAHLA B27 positive, %89.30NADisease duration since diagnosis, median (IQR)7.25 (3.33—10.04)NAnr-axSpA/r-axSpA38.70/61.30NAaxSpA SIJ limited/ spine involvement, %64.50/35.50NACurrent IBP/arthritis/enthesitis, %41.90/8.60/18.30NAMedication—bDMARDs/csDMARDs, %22.60/7.50NAMedication—NSAID on demand/regularly, %67.70/2.20NABMI—body mass index, Marital status: 1—married/partner, 2-single,divorced,widowed, WHO—World Health Organisation, 5D-EQ five dimensional quality of life; Work status: 1-full-time/part-time; 2:sick-leave/unemployed/dis.pension; 3:student/parental-leave/pension; IBP—inflammatory back pain, CRP—C reactive protein, ASDAS-CRP—Ankylosing spondylitis (AS) disease activity score with CRP, BASDAI—Bath AS disease activity index, BASFI -Bath AS functional index, BASMI—Bath AS Mobility index, SIJ—sacroiliac joint, enthesitis was evaluated according to Maastricht AS enthesitis criteria, DMARDs—disease modifying drugs, b—biologic, cs—conventional syntetic, NSAID—non-steroidal anti-rheumatic drugs.°: WHO Physical activity: 30 min of moderate-intensity physical activity 5 times a week or 20 min of high-intensity physical activity 3 times a week.°°: WHO Recommendation for alcohol consumption: 2 standard units for women and 3 standard units for men; standard units:hard alcohol—0.04 l shot 1.3 units, beer 10° 0.33 l 1 unit, beer12° 1.5 units, wine 1dcl 1 unit.P significance * < 0.01, ** < 0.001(CI -0.25—-0.14). Demographic data and axSpA features. BMI—body mass index, Marital status: 1—married/partner, 2-single,divorced,widowed, WHO—World Health Organisation, 5D-EQ five dimensional quality of life; Work status: 1-full-time/part-time; 2:sick-leave/unemployed/dis.pension; 3:student/parental-leave/pension; IBP—inflammatory back pain, CRP—C reactive protein, ASDAS-CRP—Ankylosing spondylitis (AS) disease activity score with CRP, BASDAI—Bath AS disease activity index, BASFI -Bath AS functional index, BASMI—Bath AS Mobility index, SIJ—sacroiliac joint, enthesitis was evaluated according to Maastricht AS enthesitis criteria, DMARDs—disease modifying drugs, b—biologic, cs—conventional syntetic, NSAID—non-steroidal anti-rheumatic drugs. °: WHO Physical activity: 30 min of moderate-intensity physical activity 5 times a week or 20 min of high-intensity physical activity 3 times a week. °°: WHO Recommendation for alcohol consumption: 2 standard units for women and 3 standard units for men; standard units:hard alcohol—0.04 l shot 1.3 units, beer 10° 0.33 l 1 unit, beer12° 1.5 units, wine 1dcl 1 unit. P significance * < 0.01, ** < 0.001(CI -0.25—-0.14). Serum levels of adipocytokines reflect body constitution rather than disease activity of other features of axSpA Table 2 summarizes serum biomarker concentrations in axSpA patients (including subgroup analyses) and HC. Among the assessed biomarkers, only visfatin differed between groups. No sex differences were observed within the axSpA cohort. However, male axSpA patients had higher visfatin levels than healthy males (449.11 ± 514.68 vs. 234.59 ± 457.60; p < 0.05). Table 2Serum levels of adipocytokines and selected axSpA features.Ghrelin pg/mlLeptin pg/mlResistin pg/mlVisfatin pg/mlFetuin-A ng/mlaxSpA, n = 93412.67 (± 336.97)2951.48 (± 3216.19)8293.09 (± 3197.61)402.10 (± 492.44) *728.60 (± 268.45)HC, n = 100469.77 (± 448.44)3518.16 (± 4105.06)7755.36 (± 2919.01)243.42 (± 437.63)789.97 (± 289.41)axSpA featuresASDAS-CRP < 1.4 (n = 31)425.60 (± 266.93)1734.65 (± 1653.54)**8349.41 (± 3264.04)460.48 (± 553.62)716.54 (± 261.33)ASDAS-CRP ≥ 1.4 (n = 62)406.21 (± 368.90)3559.90 (± 3623.16)**8180.45 (± 3110.16)372.90 (± 460.87)752.74 (± 285.02)nr-axSpA (n = 36)513.07 (± 382.13)*3188.53 (± 3969.85)8204.81 (± 2539.92)369.44 (± 449.38)762.45 (± 254.49)r-axSpA (n = 57)349.26 (± 291.05)*2801.77 (± 2661.738348.85 (± 3571.77)422.72 (± 520.63)707.23 (± 276.97)BMI < 25530.21 (± 395.85) **1255.98 (± 1242.12) ***8265.20 (± 2923.44)456.03 (± 508.02)715.56 (± 291.20)BMI ≥ 25327.79 (± 259.33) **4176.01 (± 3634.02) )***8313.24 (± 3408.79)363.14 (± 481.89)738.02 (± 253.14)Ankylosing spondylitis disease activity score with C reactive protein, SIJ—sacroiliac joint; Data reported as mean (SD).The intergroup differences of each biomarker, P significance: *0.05 **0.01 ***0.001.Addional data: Ghrelin nr-axSpA/r-axSpA.(CI -312.71—-14.91), BMI (CI -347.49—-57.35); leptin BMI (CI 1856.67—3983.39); ASDAS (CI 737.31—2913.19), visfatin (CI 25.99—291.35). Serum levels of adipocytokines and selected axSpA features. Ankylosing spondylitis disease activity score with C reactive protein, SIJ—sacroiliac joint; Data reported as mean (SD). The intergroup differences of each biomarker, P significance: *0.05 **0.01 ***0.001. Addional data: Ghrelin nr-axSpA/r-axSpA. (CI -312.71—-14.91), BMI (CI -347.49—-57.35); leptin BMI (CI 1856.67—3983.39); ASDAS (CI 737.31—2913.19), visfatin (CI 25.99—291.35). In axSpA, leptin and ghrelin were associated with BMI, while only leptin was additionally associated with higher disease activity. Patients with nr-axSpA had higher ghrelin levels, whereas no differences were observed between patients with SIJ-limited and spinal involvement disease. Correlation analysis showed that leptin levels were positively correlated with BMI (r = 0.5901; p < 0.0001) and negatively correlated with male sex (r = − 0.5173; p < 0.001) (Supplementary Table 2). No associations were found between leptin and ASDAS-CRP or BASDAI. Resistin was negatively correlated with BASDAI (r = − 0.3324; p < 0.05) and showed a positive trend with physical activity (r = 0.265; p = 0.06). No significant associations were observed between the remaining adipocytokines and clinical features of axSpA. Patients with active axSpA tended to have a higher BMI than those in remission (median 26.98 [95% CI 24.62–29.25] vs. 24.69 [95% CI 22.25–27.21]; p = 0.06). Consistently, patients with spinal involvement had a higher BMI than those with SIJ-limited disease (median 27.43 [95% CI 25.08–29.61] vs. 24.79 [95% CI 22.21–28.08]; p < 0.01), while no difference was observed between nr-axSpA and r-axSpA. The primary analysis focused on disease activity. Linear regression (Table 3) was therefore performed to assess associations between disease activity and adipocytokines or BMI. Although fetuin-A reached statistical significance, a doubling of serum fetuin-A was associated with a decrease in ASDAS-CRP of 0.03. In contrast, a doubling of CRP was associated with an increase in ASDAS-CRP of 1.20. Table 3Linear regression model for ASDAS-CRP.coefSEtP0.025–0.975Age0.01110.0012.5060.0140.002–0.020Sex—male-0.15600.102-1.5230.131-0.360–0.048Leptin0.03060.0530.0570.566-0.075–0.136Resistin-0.12120.117-1.0340.304-0.354–0.112Visfatin0.01410.0210.6780.499-0.027–0.055Fetuin-A-0.30010.139-2.1570.034-0.577–0.024Ghrelin-0.07550.068-1.1130.269-0.210–0.059CRP1.20020.05222.970 < 0.0011.096–1.3004BMI-0.02100.012-1.7750.079-0.044–0.003Intercept3.19721.4172.2570.0270.382–6.013CRP—C reactive protein, ASDAS-CRP—Ankylosing spondylitis disease activity scores with CRP, BMI—body mass index. Linear regression model for ASDAS-CRP. CRP—C reactive protein, ASDAS-CRP—Ankylosing spondylitis disease activity scores with CRP, BMI—body mass index. Depression in axSpA – the burden across somatic, cognitive, and affective domains Compared with HC, patients with axSpA had higher average BDI-II (6.76 ± 6.81 vs 2.81 ± 3.31, p < 0.001) and BAI scores (6.67 ± 7.71 vs 3.89 ± 3.88, p < 0.05). In categorical analysis, fewer axSpA patients were free of depression (80.60% vs 98.00%) and anxiety (76.30% vs 90.00%), while mild, moderate, and severe symptoms were more frequent in axSpA (depression: 14.00%, 4.30%, 1.10%; anxiety: 19.30%, 3.20%, 1.10%, respectively) than in HC (p < 0.01 for depression, p < 0.05 for anxiety). Across BDI-II domains, somatic scores were highest in both groups; however, all domain scores were higher in axSpA than in HC (somatic: 3.76 ± 3.41 vs. 1.67 ± 2.14; cognitive: 1.49 ± 2.40 vs. 0.52 ± 1.16; affective: 1.51 ± 1.80 vs. 0.59 ± 1.06). Depression and anxiety in relation to disease activity and clinical features Depression and anxiety were associated with disease activity, but not with spinal involvement or BMI. Table 4 shows associations between disease activity, functional status, and symptoms of depression and anxiety in axSpA. All BDI-II domains (somatic, cognitive, and affective) were positively correlated with ASDAS-CRP (r = 0.517, 0.375, and 0.427, respectively; all p < 0.01) and BASDAI (r = 0.579, p < 0.0001; r = 0.430, p < 0.01; r = 0.489, p < 0.001, respectively). CRP levels were also correlated with total BDI-II (r = 0.356; p < 0.01). Table 4Variability of clinical features and serum biomarkers of axSpA in patients with depression and anxiety.Depression (n = 18)Without depression (n = 75)Anxiety (n = 22)Without anxiety (n = 71)Age, median (IQR)39.00 (36.00—47.25)42.00 (34.00—48.00)46.00 (41.25—55.00)**41.00 (33.00—47.00)**Sex—Female, %20.6079.4032.4067.60Sex—Male,%18.6081.4018.6081.40BMI, median (IQR)24.87 (22.24—27.93)26.23 (23.77—28.96)27.76 (24.85—31.38)25.50 (23.28—28.11)5D-EQ, median (IQR)0.58 (0.20—0.66)***0.80 (0.73—1.00)***0.62 (0.52—0.69)****0.80 (0.73—1.00)****CRP mg/l, median (IQR)8.54 (2.23—18.80)*2.52 (1.27—6.14)*6.31 (2.31—9.82)2.50 (1.21—6.41)ASDAS-CRP inactive/active disease (< 1.4/ ≥ 1.4), %27.80/72.20***73.30/26.70***27.30/72.70****76.10/23.90****BASDAI, median (IQR)5.10 (3.60—6.38)****2.40 (1.20—3.80)****5.15 (3.82—6.38)****2.00 (1.15—3.45)****BASFI, median (IQR)3.35 (2.90—6.17)****0.60 (0.20—2.05)****3.15 (1.95—6.42)****0.60 (0.20—2.05)****BASMI, median (IQR)3.00 (2.45—4.20)2.40 (1.80—3.10)2.80 (2.05—3.60)2.40 (1.90—3.10)Disease duration since diagnosis, median (IQR)8.37 (4.19—12.88)6.90 (3.25—9.67)7.49 (4.19—10.03)7.10 (3.30—10.01)nr-axSpA/r-axSpA22.20/77.8042.70/57.3040.90/59.1038.00/61.90Ghrelin pg/ml, mean (SD)466.85 (± 456.64)399.67 (± 303.95)369.78 (± 425.33)425.97 (± 306.93)Leptin pg/ml, mean (SD)2173.23 (± 1813.27)3138.27 (± 3452.82)4023.15 (± 3791.33)2619.42 (± 2968.13)Resistin pg/ml, mean (SD)7775.76 (± 2886.99)8417.26 (± 3273.58)7075.37 (± 2769.40)*8670.42 (± 3244.51)*Visfatin pg/ml, mean (SD)402.91 (± 463.05)401.90 (± 502.22)352.18 (± 457.18)417.56 (± 504.96)Fetuin-A ng/ml, mean (SD)768.82 (± 354.88)718.95 (± 245.31)736.91 (± 293.74)726.03 (± 262.30)P significance * < 0.05, ** < 0.01, *** < 0.001, **** < 0.0001 (CI se bellow).Depression – 5DEQ (CI -0.48—-0.18); CRP (CI 0.38—17.32); ASDAS-CRP (CI 0.21—0.70); BASDAI (CI 1.35—3.22); BASFI (CI 1.63—4.24).Anxiety—(CI -0.43—-0.17); ASDAS-CRP (CI 0.26—0.71); BASDAI (CI 1.94—3.51); BASFI CI 1.56—4.14); resistin (CI -3019.19—-170.91). Variability of clinical features and serum biomarkers of axSpA in patients with depression and anxiety. P significance * < 0.05, ** < 0.01, *** < 0.001, **** < 0.0001 (CI se bellow). Depression – 5DEQ (CI -0.48—-0.18); CRP (CI 0.38—17.32); ASDAS-CRP (CI 0.21—0.70); BASDAI (CI 1.35—3.22); BASFI (CI 1.63—4.24). Anxiety—(CI -0.43—-0.17); ASDAS-CRP (CI 0.26—0.71); BASDAI (CI 1.94—3.51); BASFI CI 1.56—4.14); resistin (CI -3019.19—-170.91). Among BDI-II domains, the strongest association with CRP was observed for the somatic domain (r = 0.425; p < 0.01), whereas associations with cognitive and affective domains were weaker (r = 0.238; p = 0.09 and r = 0.272; p = 0.05, respectively). Overweight patients (BMI ≥ 25) tended to report depression less frequently than those with normal weight (13.0% vs. 28.2%; p = 0.08), while no difference was observed for obesity (BMI > 30). In contrast, anxiety was more frequent in obese patients, with 36.4% reporting anxiety, whereas 87.3% of patients with BMI ≤ 30 were free of anxiety (p < 0.05). Sufficient physical activity was more common in patients without anxiety than in those with anxiety (77.5% vs. 40.9%; p < 0.01). Serum adipocytokines in relation to depression and anxiety in axSpA Among the assessed adipocytokines, only resistin differed between axSpA patients with and without anxiety (Table 4). Accordingly, serum resistin levels were negatively correlated with BAI scores (r = − 0.422; p < 0.001; Supplementary Table 3). No differences in adipocytokine levels were observed between patients with and without depression (Table 4). Associations of fetuin-A and resistin with depression and anxiety in axSpA Linear regression analyses were performed to assess the relationship between adipocytokines and depression or anxiety (Table 5). Male sex was associated with lower BDI-II and BAI scores (by 1.6 and approximately 4 points, respectively). After adjustment for age and sex, only fetuin-A and resistin were associated with depression and anxiety, with opposite effects: a doubling of fetuin-A was associated with an increase of nearly 4 points in BDI-II and 3 points in BAI, whereas a doubling of resistin was associated with decreases of 2.5 and 2.6 points, respectively. Table 5Linear regression model for depression and anxiety.BDI-IIcoefSEtP0.025–0.975Age0.06350.0193.302**0.025–0.102Sex—male-1.58960.466-3.410**-2.516–0.663Leptin-0.81330.231-3.527**-1.272–0.355Resistin-2.58700.533-4.853***-3.646–1.528Visfatin-0.10640.103-1.0350.303-0.310–0.098Fetuin-A3.55180.5486.486***2.463–4.640Ghrelin0.33730.2831.1930.236-0.224–0.899BMI-0.26060.050-5.169***-0.361–0.160ASDAS-CRP3.94380.16823.518***3.611–4.277Intercept8.59696.3721.3490.181-4.067–21.260BAIcoefSEtP0.025–0.975Age0.05330.0232.294*0.007–0.099Sex—male-3.94620.417-9.462***-4.775–3.117Leptin-0.71410.211-3.391**-1.133–0.296Resistin-3.46140.455-7.608***-1.366–2.557Visfatin0.10900.0801.3660.175-0.050–0.268Fetuin-A3.03940.4766.381***2.093–3.986Ghrelin-0.13340.348-0.3830.703-0.826–0.559BMI0.02080.0500.4180.677-0.078–0.120ASDAS-CRP3.64940.22216.475***3.209–4.090Intercept15.57804.8243.229**5.991–25.165CRP—C reactive protein, BMI—body mass index.P significance, *0.05, **0.001, ***0.0001. Linear regression model for depression and anxiety. CRP—C reactive protein, BMI—body mass index. P significance, *0.05, **0.001, ***0.0001. Leptin showed minimal effects, with a doubling associated with changes of less than 0.75 points in both scores. BMI was also associated with BDI-II, but the effect size was small (a one-unit increase corresponded to a decrease of approximately 0.3 points). In contrast, ASDAS-CRP was strongly associated with both outcomes, with a one-point increase corresponding to increases of approximately 4 points in BDI-II and 3.7 points in BAI. Discussion This study is the first to investigate the relationships between serum adipocytokines, BMI, disease activity, and symptoms of depression and anxiety in patients with axSpA, providing new insights into the mental burden of the disease. While higher disease activity was associated with more severe depressive and anxiety symptoms, BMI showed only a marginal influence on either mental health outcomes or disease activity. Consistently, leptin and ghrelin levels reflected BMI, with limited effects on depressive or anxiety symptoms, whereas resistin was inversely associated with both BASDAI and anxiety. In contrast, fetuin-A showed positive associations with depression and anxiety, and, to a lesser extent, disease activity. In our study, symptoms of depression and anxiety were closely associated with the inflammatory burden of axSpA and its impact on quality of life and physical function, but not with disease duration or advanced skeletal involvement. Given the cross-sectional design, no conclusions on causality or directionality can be drawn. These findings are consistent with previous evidence. A meta-analysis reported an association between disease activity and depression and anxiety in axSpA2. More recently, Zhao et al. also demonstrated a close relationship between disease activity and depressive and anxiety symptoms, although no specific subgroups of axSpA patients with differential treatment responses related to mental burden were identified32. Neuroinflammation is one of the proposed mechanisms underlying the development of depression, as it interferes with brain circuitry and may arise from chronic low-grade inflammation characterised by elevated pro-inflammatory cytokines and CRP33,34. Data based on longitudinal national cohorts further indicate that somatic symptoms of depression are associated with elevated CRP levels, rather than with cognitive changes, suicidal ideation, or anxiety33,35. Symptoms such as low energy, fatigue, exhaustion, leaden paralysis, sleep and appetite disturbances, as well as poor attention, executive dysfunction, low mood and motivational anhedonia, are characteristic of immune-derived depression, which may account for up to 30% of MDD cases36. In turn, registry-based studies demonstrate a higher prevalence of mood and anxiety disorders in patients with axSpA compared with the general population3,37,38. These findings suggest a shared mechanism driving depressive symptoms in axSpA, predominantly somatic but also affecting cognitive and affective domains. Whether these symptoms alter neurobiological pathways or disrupt coping mechanisms remains to be clarified. Further research is needed to enable early identification of patients with psychological distress and to support personalised management, integrating psychiatric or psychological interventions with intensive anti-inflammatory therapy to reduce both mental and overall disease burden. Although not yet implemented in clinical practice, individual biological molecules and their combinations have been investigated as potentially informative biomarkers for depressive disorders39,40. Given the strong interrelation between psychological symptoms and disease activity, the proposed link between serum resistin and depression—particularly in the context of axSpA—remains complex and requires careful interpretation. The existing literature on resistin can be broadly divided into two partially overlapping research domains: studies focusing on psychiatric populations and depressive disorders, and studies investigating inflammatory conditions such as axSpA. In psychiatric research, resistin has been proposed as a potential biomarker of depression, with several studies reporting positive associations with symptom severity8,39–42. However, these findings are inconsistent, as other studies have failed to confirm such associations or have even reported lower resistin levels in major depressive disorder43–45. In contrast, in axSpA, elevated circulating concentrations of resistin, visfatin and fetuin-A have been consistently reported16–18,46 and these adipocytokines, together with leptin, have been implicated as potential mediators of radiographic progression in r-axSpA18,47,48. Although several studies have described fluctuations in leptin15 and resistin49 during anti-inflammatory treatment, robust evidence supporting a direct relationship between these adipocytokines and contemporaneous disease activity is lacking47,50. In this context, our finding of an inverse association between resistin, anxiety, and BASDAI may reflect the interplay between these two biological frameworks. While psychiatric studies primarily interpret resistin in relation to affective symptomatology, in axSpA it may be more closely linked to inflammatory burden, structural progression, or specific disease phenotypes. Moreover, in our cohort, depressive symptomatology was predominantly characterized by somatic features, and only 35% of patients had predominant spinal involvement, suggesting a heterogeneous clinical profile. This may further contribute to the observed discrepancies, as somatic symptoms of depression overlap with manifestations of chronic inflammatory disease. Taken together, these findings suggest that in axSpA, resistin may not function as a straightforward biomarker of depressive symptoms, but rather reflects a more complex interaction between inflammation, disease phenotype, and psychological burden. Resistin, visfatin, fetuin-A and leptin are predominantly produced by adipocytes, but additional sources—including hepatic tissue and peripheral immune cells—contribute to their circulating levels. Their immunomodulatory capacities can potentiate adipose tissue inflammation51, thereby promoting systemic low-grade inflammation, particularly in individuals with obesity52. Ge et al. reviewed the interaction between obesity and depression, highlighting its bidirectional nature and emphasising the potential involvement of leptin and leptin resistance in the pathogenesis of immune-mediated depression53. Consistent with this notion, a recent meta-analysis reported poorer disease activity outcomes in overweight or obese axSpA patients54. In our study, serum leptin concentrations were higher in overweight individuals and in those with increased disease activity. However, fetuin-A, rather than leptin, demonstrated a stronger association with depression and anxiety scores. Fetuin-A has been described as a modulator of inflammatory responses, including attenuation of excessive immune activation, which may be particularly relevant in the context of chronic inflammatory diseases. We therefore hypothesise that, in axSpA, higher fetuin-A levels may reflect a compensatory anti-inflammatory mechanism that could be linked to the psychological burden of chronic disease. From a clinical perspective, our findings further support the importance of a holistic management approach in axSpA, including attention to body weight and metabolic status, which may influence both inflammatory activity and psychological well-being. Limits of the study Several limitations should be acknowledged. The cross-sectional design precludes assessment of longitudinal changes in body composition, inflammatory biomarkers, or adipocytokines following therapy. Radiographic data were categorical, without detailed scoring of sacroiliac or spinal involvement, and body composition was assessed only by BMI, precluding analysis of visceral fat, total fat, or muscle mass. In addition, several potentially important confounding factors, such as socioeconomic status, educational level, prior psychiatric history, physical activity, and pain-related interference, were not systematically assessed and therefore could not be included in the multivariable analyses. Although composite disease activity indices (BASDAI and ASDAS-CRP) partially reflect fatigue, pain, and overall health status, residual confounding due to unmeasured variables cannot be excluded. Another limitation of this study is the potential influence of ongoing biological therapy (bDMARDs), which may affect both inflammatory activity and psychological outcomes. The another limitations became during sampling and measurements: Although sampling was performed after overnight fasting at a tertiary centre, some participants travelled long distances and were allowed up to 20 g of bread no more than two hours before sampling to prevent hypoglycaemia. A limitation of this study is the relatively high inter-assay variability of the multiplex adipocytokine measurements (up to 25%), which may affect the precision of absolute values. However, all samples were processed under identical conditions and analysed in batches, which reduces the risk of systematic bias and supports the validity of between-group comparisons. This variability is inherent to multiplex assay platforms and, given the uniform sample processing and batched analysis, is unlikely to have materially affected the overall interpretation of the results. The use of a healthy control group, while appropriate for comparison of mood scores, does not allow assessment of adipocytokine profiles in individuals with depression or anxiety in the absence of inflammatory rheumatic disease. This limits our ability to determine whether the observed associations are specific to axSpA or reflect broader pathophysiological mechanisms present in the general population. Finally, this was a single-centre study including a predominantly Caucasian population, which may limit the generalisability of our findings. Adipocytokine profiles are known to be influenced by ethnicity, as well as by lifestyle and dietary factors, and therefore our results may not be directly applicable to more diverse populations. Conclusion In this first study to jointly examine adipocytokines, inflammatory activity, and affective symptoms in axSpA, we demonstrate that depressive and anxiety symptoms are closely associated with inflammatory burden rather than with metabolic status or structural disease severity. While leptin and ghrelin largely reflected BMI with limited relevance to either mental health or disease activity, fetuin-A showed clearer associations with depressive and anxiety symptoms. Notably, resistin exhibited an inverse relationship with BASDAI and affective symptoms, contrasting with findings from depressive disorders and suggesting disease-specific immunometabolic pathways. Overall, our results support the concept that affective symptoms in axSpA are closely linked to inflammatory processes. However, given the cross-sectional design, causal relationships cannot be inferred. These findings highlight the importance of recognising psychological distress in axSpA and integrating its assessment into comprehensive patient management. Supplementary Information Supplementary Information. Acknowledgements This work was supported by the Czech Health Research Council (NU21-09-00297), the project (Ministry of Health, Czech Republic) for conceptual development of the research organization 00023728 (Institute of Rheumatology), and by the project BBMRI.cz, reg. No. LM2023033. Author contributions Conceptualization: Markéta Hušáková, Karel Pavelka, Karolína Kuchaříková, Pavel Mohr, Martina Olejárová; Methodology: Markéta Hušáková, Karel Pavelka, Karolína Kuchaříková; Formal analysis and investigation: Markéta Hušáková, Jan Kotrs, Markéta Kuklová, Hana Hulejová, Ladislav Šenolt; Writing—original draft preparation: Markéta Hušáková; Writing—review and editing: Karel Pavelka, Karolína Kuchaříková, Pavel Mohr, Martina Olejárová, Jan Kotrs, Markéta Kuklová, Hana Hulejová, Ladislav Šenolt; Funding acquisition: Markéta Hušáková, Karel Pavelka; Resources: Karel Pavelka, Markéta Hušáková; Supervision: Pavel Mohr, Martina Olejárová. Funding Agentura Pro Zdravotnický Výzkum České Republiky, NU21-09-00297, Ministerstvo Zdravotnictví Ceské Republiky, 00023728, BBMRI.cz, LM2023033 Data availability The datasets generated during the current study are not publicly available due to clinical data but are available from the corresponding author on reasonable request. Declarations Competing interests The authors declare that they have no conflicts of interest regarding the present manuscript. All support received in the previous three years: Research grant dedicated to the Institute of Rheumatology in Prague: MH—NU21-09–00297; KP —NU23-10–00434 Research grants Czech Health Research Council (NIMH), the Czech Research Foundation (NIMH), and the European Brain Council: PM Meeting supports: MH – UCB Honoraria for lecture and speakers fees: KP—AbbVie, Pfizer, Eli Lilly, MSD, Biogen, UCB, LŠ—AbbVie, AstraZeneca, Bristol-Myers Squibb, Eli Lilly, GlaxoSmithKline, Johnson & Johnson, Merck Sharp and Dohme, Novartis, Sanofi, Sandoz, Swedish Orphan Biovitrum, Swixx Biopharma, UCB; PM—Johnson & Johnson, Lundbeck, Richter Gedeon, AstraZeneca, Angelini, Teva, Boehringer Ingelheim, KRka. All other authors also declare that they have no conflicts of interest. Ethical approval All participants agreed to take part in the study and followed the Informed Consent approved by the Ethics Committee of the Institute of Rheumatology in Prague (No. 5704/2021). Footnotes Pavel Mohr and Marta Olejárová Shared senior authors. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Information. Data Availability Statement The datasets generated during the current study are not publicly available due to clinical data but are available from the corresponding author on reasonable request.