Psoriatic arthritis (PsA) is a chronic arthritis associated with psoriasis. The prevalence is .58% (95% CI: .14-.49) in our setting (Spain) according to data from the EPISER registry1.

It is a very heterogeneous disease, with multiple facets in addition to musculoskeletal manifestations2. PsA is accompanied by several cardiovascular and metabolic comorbidities that increase the risk of a major cardiovascular event (MACE)3,4. However, the occurrence of MACE is not exclusively due to common risk factors such as hypertension, dyslipidaemia, diabetes mellitus or obesity. The inflammatory nature of the disease also plays an important role5. Acute phase reactants, such as C-reactive protein (CRP) or erythrocyte sedimentation rate (ESR),which are used in the follow-up of chronic arthritis, have been associated with arterial wall stiffness6 and the increase in cardiovascular risk7,8.

Obesity is highly prevalent in people with psoriasis and with PsA, compared with health individuals4. Being overweight causes low grade systemic inflammation and is associated with higher levels of CRP and ESR9, which may lead to errors when assessing obese patients.

When considering the interaction between obesity and disease, there is evidence that obesity is a negative predictor of both the effectiveness and persistence of an adequate response to biological and non-biological treatment in patients with PsA10. However, this interaction is bidirectional: proinflammatory cytokines involved in the onset and maintenance of immune-mediated disease also promote metabolic syndrome11. Therefore, there are studies that suggest that treatment of inflammatory arthropathies could improve the metabolic profile and cardiovascular risk of this type of patient12, although the evidence is not very clear, at least in PsA13.

Our study is twofold in its aim. On the one hand, we want to test whether there are interactions between the patient’s metabolic status, inflammatory parameters and the disease assessed by different tools (Disease Activity in Psoriatic Arthritis [DAPSA], Minimal Disease Activity [MDA] and HAQ index). On the other hand, we want to assess whether metabolic or cardiovascular diseases have any association with the reduction of inflammatory burden by PsA treatment.

We conducted a descriptive cross-sectional study of 160 patients with PsA from a rheumatology department of a tertiary hospital in north-eastern Spain during a period from June 2018 to March 2019. Outpatients of both sexes over 18 years of age, diagnosed with PsA according to the Classification Criteria for Psoriatic Arthritis (CASPAR), were included. Patients with and without disease-modifying treatments were included. No patients were excluded. An additional consultation was performed with all patients for data collection.

The study procedures included the collection of sociodemographic, clinical, analytical and radiographic variables. For the prescription of systemic medications, as well as for the assessment of therapeutic targets, the updated recommendations for the management of PsA given by the Spanish Society of Rheumatology were followed.

According to the dominant joint pattern in the last 5 years of follow-up, patients were diagnosed as having oligoarticular forms of PsA if they had a maximum of 4 inflamed joints; polyarthritis if they had 5 or more inflamed joints, and the ASAS criteria for axial spondyloarthritis were followed for the definition of axial forms. The presence of dactylitis or enthesitis was also collected, whether present at the time of data collection or recorded in the previous medical history. Distal interphalangeal joint involvement was recorded as another feature of PsA, but not as a specific joint pattern.

Both the DAPSA index, which focuses on joint involvement, and the MDA, a binary measure of disease status covering important domains of PsA (including joints, skin and entheses), have been shown to have construct validity and are easy to perform in clinical practice. We used them to assess both disease activity (DAPSA) and treatment goals (MDA and DAPSA remission) in the present work. Patients were considered to be in MDA when they met at least 5 of the following criteria: painful joint count ≤ 1; swollen joint count ≤ 1; Psoriasis Area and Severity Index ≤ 1 or body surface area ≤ 3%; patient’s visual analogue pain scale score ≤ 15; patient's overall visual analogue disease activity scale score ≤ 20; HAQ score ≤ .5; and number of painful entheses ≤ 1. The DAPSA score was calculated by summing the number of tender and swollen joints, pain on the visual analogue scale, patient global assessment and CRP. Clinical DAPSA was calculated without the contribution of CRP. DAPSA and clinical DAPSA ≤ 4 scores identified a state of remission. We used the Spanish version of the HAQ disability questionnaire to assess patients’ quality of life.

Cardiometabolic risk factors (obesity, diabetes, hypertension, dyslipidaemia and smoking) and adverse cardiovascular events (angina, myocardial infarction, stroke and peripheral arterial ischaemia) were recorded for all patients. Cardiometabolic risk factors were only recorded if they were present at the time of the visit. Normal values for CRP and ESR are set at .5 mg/dl and 20 mm/h respectively, according to our laboratory cut-off points.

Descriptive statistics are expressed as percentages, means or medians, and the interquartile range (IQR) is indicated where appropriate. For comparisons between qualitative variables we used the chi-squared test, and for quantitative variables, Student’s t-test, Mann-Whitney U test and Kruskall Wallis H test, depending on the normality or non-normality of the distribution of the variables, considering p < .05 as statistically significant. We also used odds ratio (OR) to measure the association between treatment and different outcomes or events, such as MACE, DAPSA, MDA, CRP and ESR. To establish correlations between BMI, waist circumference, CRP, ESR, DAPSA, HAQ and MDA, Pearson's correlation coefficient was used. Meeting MDA criteria is a qualitative value, so it was recorded as a binary value: 0 = does not meet criteria; 1 = meets criteria, so a negative relationship with this variable indicates an association with meeting MDA criteria. IBM SPSS Statistics 19.0 software was used for data analysis.

Our series included 77 women and 83 men. The mean age was 54 years (IQR: 44.25-64.75) and the mean duration of the disease was 10.6 years (IQR: 4-15). Regarding anthropometric data, the mean BMI of our patients was 28.3 (IQE: 24.53-30.88) and the mean waist circumference was 101.1 cm (IQR: 92.25-108.75) in our series, although we did not differentiate between men and women.

Approximately one third of our series suffered from a cardiovascular risk factor, the most common being high blood pressure (28%), dyslipidaemia (25%) and obesity (33%), followed by a tobacco habit (24%).

The oligoarticular pattern was the most frequent in our series (44%), followed by the polyarticular pattern (26%) and the mixed pattern (23%), with the exclusively axial form being the least frequent (7%). Twenty five per cent of our series had enthesitis and involvement of distal interphalangeal joints. Forty one per cent of the subjects had dactylitis. Erosions were present in 12% of the study population, and 3% presented with uveitis.

The DAPSA median was 6 (IQR: 2-12), whilst the acute phase reactants were negative in the majority of our series: CRP median: .2 (IQR: .1-.6); ESR media: 6,5 (IQR: 3-15). HAQ median .25 (IQR: 0-.6). Sixty six per cent of our sample met with the MDA criteria.

In our series there are 50 patients being treated with synthetic disease-modifying drugs (S-DMARDs): 43 patients with methotrexate, 6 patients with leflunomide and 1 patient with sulfasalazine. There are 89 individuals receiving treatment with biologic therapies (b-DMARDs): 67 patients with anti-TNF (iTNF), 18 patients with ustekinumab and 4 patients with secukinumab. The remaining 21 patients do not receive any specific treatment due to good disease control. All these data are presented in Table 1.

BMI and waist circumference positively correlated with the CRP and the ESR (bilateral significance  < .05), although the degree of correlation is low (Pearson < .4). We found no relationship between BMI or waist circumference and DAPSA or meeting with MDA criteria. There was also a correlation between the CRP and the ESR with DAPSA and fulfilment of the MDA criteria (Bilateral significance: < .05). These data are contained in Table 2.

When assessing the relationship between metabolic diseases, acute phase reactants, DAPSA and HAQ; we found that high blood pressure was associated with higher ESR values (Mann-Whitney U, p = .019) and HAQ (Mann-Whitney U, p = .049), while hyperuricaemia was associated with higher CRP levels (Mann-Whitney U, p = .026). We found no association with the rest of metabolic diseases. These data are contained in Table 3.

If we check the associations between treatment and the different diseases that make up the metabolic syndrome, we find that the treatment of PsA does not modify any of them. However, these diseases were recorded qualitatively (present or not present). Blood pressure, blood glucose, cholesterol, triglycerides and uric acid values were not collected.

The use of bDMARDs was associated with a lower prevalence of MACE in comparison to sDMARDs (p = .047; OR: .12; 95% CI:.01-.9). We also found a lower prevalence of enthesitis (p = .008; OR: .3; 95% CI:.16-.56), and a higher possibility of having lower DAPSA (p = .03; OR: .28; 95% CI:.11-.72) or meeting with the MDA criteria (p = .001; OR: 3,65; 95% CI:1.63-8.13), and greater probability of having normal CRP (p = .029; OR: .25; 95% CI:.07-.87) and ESR (p = .024; OR: .36; 95% CI:.16-.82) levels in the group which received treatment with bDMARDs compared with the group which received sDMARDs.

Metabolic diseases and cardiovascular risk are two aspects that are often overlooked in the routine management of PsA patients, although it is well documented that these patients tend to have an unfavourable profile in both areas.

Our results indicate that there is a positive correlation between anthropometric measures (BMI and waist circumference) and acute phase reactants (CRP and ESR), but when compared with DAPSA or MDA criteria, there is no such relationship. This indicates that in obese patients we may find elevated acute phase reactants without necessarily being associated with increased disease activity.

If we evaluate the association between treatment and the different metabolic diseases in our patients, we find that treatment of the disease is not associated with a lower prevalence of hypertension, dyslipidaemia, diabetes or hyperuricaemia. Nor are there differences in the prevalence of these metabolic diseases between the different treatment groups (sDMARDs versus bDMARDs). On the other hand, we found a lower prevalence of enthesitis and cardiovascular events in the bDMARDs group compared to the sDMARDs group. These results are consistent with other studies in which enthesitis was associated with increased cardiovascular risk in patients with PsA14. There are also differences in CRP and ESR levels, where we found that treatment with B-CSF is more frequently associated with normal levels of both acute phase reactants compared to those receiving treatment with S-CSF, which may also explain the difference in the prevalence of cardiovascular events between these groups7,8.

Out of all the metabolic diseases included in our study, only hypertension and hyperuricaemia were associated with higher values of CRP, ESR or HAQ, but diabetes or dyslipidaemia were not, probably due to the data collection protocol.

Obesity involves excess adipose tissue, which has classically been regarded as an energy reserve tissue. However, this is a misperception. Adipose tissue acts as an endocrine organ in that it produces anti-inflammatory and pro-inflammatory cytokines, called adipokines15. In situations where calorie supply exceeds the body’s needs, the adipocyte initially hypertrophies to maintain normal adipose tissue function and allows the formation of new adipocytes (hyperplasia). This phenomenon is directly related to the degree of obesity.15 If excessive calorie intake is perpetuated over time, triglyceride accumulation will occur only through adipocyte hypertrophy and not hyperplasia, resulting in impaired regulation of the synthesis of anti-inflammatory and pro-inflammatory cytokines15. In particular, some of the adipokines that are synthesised to a greater extent are tumour necrosis factor alpha (TNF), interleukin-6, monocyte chemoattractant protein 1 and plasminogen activator inhibitor16.

TNFα is a central protein in the pathophysiology of inflammatory arthropathies and the therapeutic target of many biological treatments. It is also a link between immune-mediated diseases, obesity and other associated conditions such as insulin resistance, dyslipidaemia and atherosclerosis10. Evidence suggests that interventions such as diet or physical activity that lead to a decrease in body weight may also decrease TNFα17. There is also an established relationship between obesity and PsA. This relationship is bidirectional: obesity involves the synthesis of cytokines such as TNF or vascular endothelial growth factor, which promote the onset of arthritis in predisposed subjects; in fact, excess body weight is considered an independent risk factor for PsA. On the other hand, proinflammatory cytokines involved in the pathophysiology of PsA favour obesity18.

CRP and ESR are 2 acute phase reactants that are commonly used in the daily practice of the rheumatologist. CRP is associated with poorer blood glucose and blood pressure control and is often higher in smokers7. ESR is an independent predictor of cardiovascular risk, which can be useful for assessing risk even in a non-high-risk population8. Both ESR and CRP are associated with higher BMI and waist circumference7,8. Obesity leads to higher levels of these acute phase reactants, which is partly due to TNF and interleukin-19. This relationship between BMI and acute phase reactants is present in subjects with and without PsA or rheumatoid arthritis (RA)20, indicating that excess CRP in obese individuals does not necessarily imply increased disease activity21. Our results reflect that there is an association between anthropometric data and acute phase reactants, but no such association with disease activity as measured by DAPSA and MDA. Therefore, our results suggest that in obese patients higher CRP and/or ESR values do not necessarily imply higher disease activity.

Obesity is also associated with a worse response to treatment in both RA and PsA, especially if the patient is receiving anti-TNFα therapy22. However, the underlying mechanisms of this phenomenon are not fully understood. One possible explanation is that the production of TNF and interleukin-6 by adipocytes in situations of obesity, together with the disease activity itself, creates an increased inflammatory burden23. There is also a hypothesis that obese patients have a greater volume of distribution, generating a lower peak of the drug in the blood, which is related to lower effectiveness and also facilitates the synthesis of antipharmacological antibodies23.

Given the implications of TNFα in metabolic diseases, it is logical to assume that biologic therapies may have a positive impact on a patient's metabolic profile, although there is no clear evidence in this regard. Some studies have found that individuals treated with infliximab, adalimumab or etanercept show a trend towards a smaller waist circumference, decreased insulin resistance, triglyceride levels and baseline blood glucose levels, and higher HDL levels11,12. These findings appear to be similar in other immune-mediated diseases, such as RA24, psoriasis25 or inflammatory bowel disease26, thus highlighting the role of TNF as an intermediary between inflammatory and metabolic diseases. However, there are other studies that do not find this relationship, at least in patients with PsA13. Methotrexate use has been associated with an improvement in lipid profile and basal blood glucose, but there is no such association with insulin resistance27. In any case, the benefit related to metabolic disease, if any, seems to be more evident in patients treated with anti-TNF than in those receiving methotrexate12. With respect to cardiovascular disease, anti-TNF treatment may reduce the risk of cardiovascular events, although this association is more evident in patients with RA29 than in patients with psoriasis or PsA28,29. Our data show no association between anti-TNFα treatment and metabolic diseases, probably due to the data collection protocol, but they do indicate that the use of bDMARDs could reduce the prevalence of cardiovascular events when compared with those receiving treatment with conventional therapies.

Among the limitations of our study are those inherent to a cross-sectional study. The fact that we did not find a relationship with metabolic status could be due to the fact that we only took into account whether or not the patient was diagnosed with any of the metabolic diseases included in the study. Blood pressure, cholesterol, triglycerides, blood glucose or uric acid values were not recorded. Therefore, there could be some relationship between the metabolic profile and the treatment received; and it was not found due to the data collection protocol. The correlations we have obtained have a low or very low degree of association, probably due to the baseline situation of the patients in terms of disease activity and CRP and ESR values (Table 1).

There is a lower prevalence of enthesitis and cardiovascular events, and lower CRP and ESR levels in the group of patients treated with biologic therapies compared to those treated with methotrexate or other conventional therapies. Our results are consistent with other studies that find enthesitis to be an independent cardiovascular risk factor in PsA. Therefore, our data strengthen the available evidence suggesting that anti-TNF therapy may reduce cardiovascular risk in patients with PsA. This could be due to a greater decrease in inflammatory burden compared to conventional therapies.

Our results show no association between biological therapies and a favourable metabolic profile, although this could be due to the data collection protocol.

The data from our series also indicate that higher CRP and ESR levels may be present in obese individuals, without necessarily implying higher disease activity, so that the assessment of these patients should be more thorough.

The authors have no conflict of interests to declare.