Introduction

Psoriatic arthritis (PsA) is an inflammatory spondyloarthropathy associated with psoriasis [1, 2]. In addition to the phenotype with clinically evident psoriasis, the disease can also appear in patients who have never had psoriasis but with a positive history of psoriasis in a first- or second-degree relative [3,4,5,6].

PsA represents a systemic condition, and metabolic syndrome (MetS) has been reported as a related possible aspect of the heterogeneous phenotypical spectrum [7,8,9,10,11,12]. In the development of MetS, an important role is played by the chronic low-grade inflammatory state and by the white adipose tissue (WAT) [13,14,15,16]. WAT represents a tissue devoted to energy storage and a multifunctional endocrine organ able to secrete several pro- and anti-inflammatory molecules called adipokines, therefore regulating and linking inflammation and metabolic state [15, 16]. PsA has been associated with a high prevalence of MetS and its components [17, 18].

Adipokines, synthetized by WAT and at the same time acting on it, have been considered also in the pathogenesis of the inflammatory processes of PsA [19,20,21,22,23,24,25,26,27,28].

Tumor necrosis factor alpha (TNF-α), a Th1-associated and Th17-associated molecule, represents a pro-inflammatory cytokine mainly produced by monocytes and macrophages with a central role in PsA pathogenesis [19]. On the metabolic side, TNF-α is able to induce adipocyte apoptosis, promote insulin resistance, increase plasma triglycerides, and stimulate lipolysis [20,21,22]. Interleukin-6 (IL-6) promotes the hepatic production of many acute phase reactants, and it is able to induce hepatic insulin resistance [21,22,23,24,25].

Leptin is synthesized by adipose tissue and plays an important role in body energy balance and adipose tissue deposition, and its plasma levels positively correlate with body fat mass [26,27,28]. In PsA, leptin levels have been found higher in women with PsA compared with those with psoriasis, and they positively correlate with homeostasis model assessment of insulin resistance (Homa-IR) and BMI [29]. Leptin levels have been found higher in PsA patients compared with non-PsA control individuals [30]. Leptin concentrations have been found higher in patients with psoriasis compared with controls [31]. High levels of leptin have been found in patients with psoriasis and with MetS [32].

Resistin is another adipocyte-specific pro-inflammatory polypeptide synthesized mostly by macrophages and monocytes from fat tissue, and it is able to induce insulin resistance [33, 34]. Resistin concentrations have been found higher in patients with PsA and psoriasis compared with controls [31, 32, 35, 36].

Visfatin mediates a complex cellular signalling process stimulated by oxidative stress resulting in vascular endothelial inflammation [37]. High levels of visfatin have been found in patients with PsA [35].

Ghrelin is a peptide secreted mainly by the gastric mucosa, able to stimulate appetite, with anti-inflammatory effects [38,39,40]. A strong negative correlation has been reported between the Psoriasis Area and Severity Index (PASI) and serum ghrelin levels [41, 42].

In a previous study [43], we have investigated differences in clinical expression between PsA with clinical evident psoriasis and PsA sine psoriasis. Main results showed a significantly higher prevalence of MetS in PsA patients with clinical evident psoriasis. In addition, BMI resulted as an important determinant in the development of MetS in these patients [43]. We also performed cytokine pattern analysis in these PsA subjects, and data are now available. The main aim of the current study is to investigate possible differences and correlations between adipokines and clinical expression in PsA patients with and without clinical evident psoriasis.

Patients and methods

Clinical characteristics of this cohort have been previously described [43]. Briefly, eighty consecutive PsA patients fulfilling the CASPAR criteria attending the Psoriatic Arthritis Clinic at the University Federico II of Naples were enrolled in the study. Forty-two patients (52.5%) were classified as having PsA with clinically evident psoriasis (group 1) (male, 19 (45.24%); mean age 49.62 ± 12.43 years) and 38 (47.5%) as having PsA sine psoriasis (absence of psoriasis and a positive family history for psoriasis) (group 2) (male, 5 (13.16%); mean age 50.66 ± 9.32 years) [43].

Inclusion criteria were both sexes, age ≥ 18 years, and stable medical conditions. Exclusion criteria were concomitant rheumatic diseases other than PsA, endocrinopathies, renal and liver diseases, alcohol consumption, continuous use of nonsteroidal anti-inflammatory drugs (NSAIDs) and/or steroids, use of biologic DMARDs (bDMARDs), use of two or more csDMARDs, and use of progestins and drugs that may cause dyslipidemia.

As reported in the previous work [43], no significant differences were found for tender joint count (TJC; 68 joints), swollen joint count (SJC; 66 joints), tender entheseal count (TEC), Health Assessment Questionnaire (HAQ) score, patient pain on visual analogue scale score (pt pain-VAS score), and patient global disease activity (PGA) VAS score.

An increased frequency of female gender was found in patients with psoriasis (45.24%) as compared with patients with sine psoriasis (13.16%), as well an increased frequency of MetS, identified by the use of the NCEP-ACT III criteria [14], in group 1 (40.5%) when compared with group 2 (13.2%) [43].

Serum levels of TNF-α, IL-6, leptin, resistin, visfatin, and ghrelin were measured using the fully automated enzyme-linked immunosorbent assay (ELISA) analyzer Triturus system (GRIFOLS Italy) and R&D Systems Quantikine ELISA Kits. Blood samples were collected after 12 h of overnight fasting and analyzed for these adipokines.

The study was approved by the Local Ethics Committee (CE: NA 126/15), and all patients gave their informed consent.

Statistical analysis

Continuous data were described by median (25th–75th Pctl) and categorical variables as frequencies and percentages. The Shapiro–Wilk test has been used to check normality for each variable. Comparisons between the two categories were made using t tests or Mann–Whitney U tests for continuous variables, as appropriated, and Fisher’s exact test for categorical variables. Linear regression, using standardized coefficients, has been used to assess the relationship between adipokine serum levels and demographic, clinical, and laboratory parameters. Variables associated with adipokine serum levels with p < 0.1 were used in the multivariate analysis. For all the analysis, STATA v.14 has been used.

Results

As reported in the “Patients and methods” section, MetS was more frequent in patients with psoriasis compared with patients without psoriasis. Intriguingly, between groups 1 and 2, results did not show significant differences regarding waist circumference (group 1, 100 (90–107) cm; group 2, 95.5 (85–110) cm; p = 0.45), fasting glucose (group 1, 90 (85–97) mg/dl; group 2, 87.5 (80–95) mg/dl; p = 0.18), HLD cholesterol (group 1, 49 (42–60) mg/dl; group 2, 52 (45–64) mg/dl; p = 0.25), triglycerides (group 1, 113.5 (88–146) mg/dl; group 2, 105.5 (84–140) mg/dl; p = 0.51), and systolic blood pressure (group 1, 127.5 (120–145) mmHg; group 2, 120 (120–130) mmHg; p = 0.076), while in group 1, higher levels of diastolic blood pressure (80 (80–90) mmHg) were found when compared with group 2 (80 (70–80) mmHg, p = 0.01).

Median values of adipokines of the entire cohort are summarized in Table 1.

Table 1 Median values (25th–75th Pctl) of adipokines in the two groups of the study
Full size table

Patients of the two groups were not significantly different for levels of TNF-α, IL-6, leptin, resistin, visfatin, and ghrelin (Table 1).

We then analyzed data using multiple linear regression. In the entire cohort, a positive association has been shown between leptin levels and female gender (β = 0.3, p = 0.001), BMI (β = 0.8, p < 0.0001), TJC (β = 0.23, p = 0.05), and pt pain-VAS score (β = 0.4, p = 0.049). In group 1, serum concentration of leptin was associated with female gender (β = 0.41, p < 0.0001) and BMI (β = 0.6, p = 0.012), whereas in group 2, a positive association was shown between leptin levels and BMI (β = 0.7, p = 0.003) and CRP (β = 0.35, p = 0.012) (Table 2).

Table 2 Multiple linear regression results reaching statistical significance for the outcome of leptin serum levels in the entire cohort of the study, group 1 and group 2
Full size table

With regard to resistin, in the multivariate model, only the association between resistin and IL-6 was found in the overall cohort (β = 0.33, p = 0.002). The association between resistin and IL-6 was confirmed in group 1 (β = 0.46, p = 0.004) but not in group 2.

For ghrelin and visfatin, no association was found in the multivariate model. When using the univariate analysis, the only variable associated with ghrelin serum levels was BMI (β = 0.3, p = 0.007) and no variable was significantly associated with visfatin.

Discussion

Different studies have investigated the role of adipokines in PsA [29,30,31,32, 35, 36]. Until today, the present study represents the first to investigate the difference in the cytokine pattern between PsA patients with psoriasis and sine psoriasis.

In the present study, we found that PsA patients with psoriasis and sine psoriasis are not significantly different for mean levels of TNF-α, IL-6, leptin, resistin, visfatin, and ghrelin.

With regard to leptin, in the entire PsA cohort, results from multiple linear regression showed a positive association between its levels and female gender, BMI, TJC, and pt pain-VAS score. In patients with psoriasis, serum concentrations of leptin were associated with female gender and BMI. Among patients without psoriasis, we found a positive association between leptin levels, BMI, and CRP.

Leptin is a pleiotropic cytokine able to sustain inflammatory responses, by stimulating several cellular activities [26,27,28, 44, 45]. Moreover, evidences are increasing on a possible role of leptin as a modulator of pain [46,47,48] and its possible role in influencing disease activity in inflammatory arthropathies [49,50,51]. In particular, in a recent Mexican study, in RA, circulating leptin correlated with clinical activity in a manner independent of the patient’s BMI [51].

Actually, in the overall cohort, serum leptin levels were correlated with pain aspects, as represented by TJC and patient pain-VAS score, which are frequently considered markers of inflammation [52]. In patients without psoriasis, we found that leptin levels were also correlated with the inflammatory marker, CRP. Correlation of leptin with CRP could be explained at the light of the fact that leptin activates monocytes and macrophages to sustain inflammatory response [28].

Our study confirms also results from other studies in which higher leptin levels have been found in PsA women with clinically evident psoriasis, and these are correlated with BMI [29, 30].

Different from RA, in which leptin has been recently found not correlated with BMI [50], in our PsA population, this adipokine was found to be associated with this anthropometric measure. This aspect suggests the need to clarify leptin’s possible role in determining differences in the metabolic status of these two diseases.

Moreover, our data showed that in PsA patients with psoriasis, serum concentrations of leptin were associated with BMI. These are consistent with the results of Kyriakou and coworkers, who found high levels of leptin in PsA patients with evident psoriasis and MetS [31].

Further studies are needed to confirm a possible role of this adipokine on metabolic aspects in PsA and as marker of pain and inflammation.

Our study shows that resistin is associated with IL-6, and this association is confirmed in patients with psoriasis, suggesting a possible combined role of these two adipokines in addressing a phenotype associated with cutaneous manifestations. The complex role of resistin in PsA remains still not clarified; a recent study has demonstrated that a 6-month treatment with the anti-IL-17 agent, secukinumab, has little influence on the levels of resistin within the first 6 months of therapy. However, this study has shown that resistin might exert a different influence between males and females from a metabolic perspective. In particular, a significant decrease in the levels of resistin (about − 24%) was found on male PsA patients [53].

Other studies have reported that resistin levels were increased in patients with severe cutaneous psoriasis [31, 32, 35, 36] and were correlated with RA clinical activity in a manner independent of the patient’s BMI [51].

Furthermore, results from our study are in line with other evidences describing that resistin levels can be regulated by the pro-inflammatory cytokine IL-6 [34].

Larger studies could clarify the possible role of resistin and IL-6 serum levels, as potential markers of cutaneous involvement, also useful to discriminate PsA with clinical evident psoriasis from PsA sine psoriasis.

Not coincidentally, in our study, despite IL-6 serum levels did not reach a significant difference between the two groups, IL-6 concentrations were higher in PsA patients with clinical evident psoriasis as compared with those in PsA sine psoriasis patients.

Limitations of our study are represented by its cross-sectional nature. Therefore, we cannot suggest any causal relationship.

Previous studies already showed that MetS and pro-inflammatory cytokine pattern are increased in patients with both PsA and psoriasis compared with healthy individuals and are increased in PsA when compared with psoriasis alone [29,30,31,32, 35, 36, 41]. For this reason, we aimed at focusing on patients with arthritis.

Strengths of our study are represented by its relatively large sample size, especially when considering sine psoriasis patients and the extensive description of clinical characteristics of the two groups.

In summary, our results show mainly a strict interplay between leptin, female gender, BMI, and inflammatory activity in overall PsA patients, as well as between resistin and IL-6 mainly in patients with psoriasis.

These findings highlight the importance of further studies on larger cohorts. They could be useful in understanding the complex pattern of adipokines which in PsA patients underlines the heterogeneous phenotypical expression of these subjects. Further studies are also advocated for clarifying possible role of these adipokines as laboratory findings or as disease mediators in addressing the different phenotypes of the disease.