Visceral adiposity in patients with psoriatic arthritis and psoriasis alone and its relationship with metabolic and cardiovascular risk

Abstract

Background

Fat mass distribution, especially in the abdominal visceral region, has been rarely evaluated in patients with PsA or psoriasis (PsO).

Methods

Patients with PsA and patients with PsO alone were evaluated and compared with control subjects (1:1 ratio in each patient group) matched for age, sex and BMI category. Body composition and fat distribution (android and visceral fat) were evaluated by DXA. Anthropometric measurements, disease activity and the systematic coronary risk evaluation (SCORE) cardiovascular risk were assessed. Metabolic parameters (insulin, homeostasis model assessment for insulin resistance), serum adipokines [total and high-molecular-weight adiponectin, leptin, resistin and retinol-binding protein-4 (RBP4)] were measured.

Results

Data for 52 patients with PsA and 52 patients with PsO and their respective paired controls were analysed. Android fat and visceral fat were found to be significantly higher in patients with PsO compared with their controls, while these measurements did not differ between patients with PsA and their controls. By multivariate analysis, after adjusting for age, sex and BMI, visceral fat was higher in PsO patients compared with PsA patients (P = 0.0004) and the whole group of controls (P = 0.0013). Insulin levels and HOMA-IR were increased in both PsA and PsO groups. High-molecular-weight/total adiponectin ratio was decreased in patients with PsO. RBP4 was significantly higher in both PsA and PsO patients. In patients with PsO, visceral fat strongly correlated with SCORE (r = 0.61).

Conclusion

Visceral fat accumulates more in PsO alone than in PsA. Visceral adiposity may be a more pressing concern in PsO relative to PsA.

Trial registration

The ADIPSO study (Évaluation du tissu ADIpeux et des adipokines dans le PSOriasis et le rhumatisme psoriasique et analyse de ses relations avec le risque cardiovasculaire) is a case–control study conducted in Besançon, France, and is registered on ClinicalTrials.gov under the number NCT02849795.

obesity, visceral fat mass, PsA, psoriasis, cardiovascular risk, adipokines, metabolic syndrome

Rheumatology key messages

• Visceral fat accumulates more in patients with psoriasis without arthritis than in those with psoriatic arthritis.

• Visceral fat is strongly correlated with the SCORE cardiovascular risk score in patients with psoriasis.

• Visceral fat seems to be more of a concern in psoriasis than in psoriatic arthritis.

Introduction

Psoriasis (PsO) has a predominantly cutaneous clinical expression but is also a disease with multisystemic implications. Indeed, a large panel of comorbidities is described in patients with PsO, including depression, cancer, metabolic and cardiovascular (CV) diseases, as well as obesity [1]. Axial and/or peripheral inflammatory arthritis can occur in patients with PsO, a clinical feature that is referred to as PsA [2]. Like PsO, PsA may also manifest with extra-articular diseases such as CV and metabolic diseases and obesity.

Obesity is frequently observed in both PsO and PsA compared with the general population [3–8]. In addition, it recently emerged that obesity may be a risk factor for developing PsO [9] and PsA [10]. The link between obesity and psoriatic diseases can be explained by the release from fat tissue of several bioactive proteins, collectively named adipokines [3, 11]. PsO and PsA share common predisposing factors such as obesity, but also obesity-related complications. Indeed, the CV risk is considered to be higher in both PsO and PsA as compared with the general population [12–14]. Obesity is usually defined by the World Health Organization (WHO) using the BMI. However, the BMI gives an approximation of fat percentage. Waist circumference (WC) is a valid measurement for abdominal/android obesity, a fat distribution strongly associated with CV complications [15]. Body composition may be assessed using the reference technique, namely DXA, which quantifies the lean and fat mass content of the whole body and specific anatomical regions such as the android region. And it is well demonstrated that visceral fat, rather than s.c. fat, is a major predictor of adverse CV events [15, 16].

There are limited data on body composition in patients with PsO [17–29] or PsA [20, 21]. In this study, we aimed to evaluate the body composition, notably visceral fat, in patients with PsA or PsO. We also investigated the association of visceral fat with CV risk and metabolic syndrome (MetS). Finally, we evaluated adipokines that may contribute to the inflammatory process of PsA and PsO and their associated metabolic complications.

Patients and methods

Patients

We included patients who had PsA, or PsO without arthritis, between 2014 and 2017. PsA and PsO patients were enrolled from the University Hospital of Besançon, in the outpatient clinic of the rheumatology and dermatology departments, respectively. PsO was diagnosed by a dermatologist based on clinical examination. Patients could be included in the PsO group if they had PsO vulgaris, regardless of the severity of the disease. The diagnosis of PsA was determined by a rheumatologist and based on the Classification Criteria of Psoriatic Arthritis [22]. Control subjects were recruited among the hospital staff (healthy subjects) or among patients referred to the rheumatology department for mechanical back pain. Control subjects were free of PsO and inflammatory disease. Each patient was matched to a single control subject according to age, sex and BMI. Exclusion criteria were an age younger than 18 years or older than 80 years, and patients receiving a biologic agent or systemic CSs at a dosage higher than 10 mg prednisone per day.

Clinical assessments

For all subjects, the following characteristics were recorded: socio-demographic data, anthropometric measurements [BMI (weight/height²), WC (measured at the midpoint between the lowest rib and the iliac crest), waist-to-hip ratio (WHR: ratio of waist to hip circumference; hip circumference was measured at the widest portion of the buttock)], disease duration, comorbidities, CV risk factors, smoking status, current treatment for PsO and PsA (topical treatments, NSAIDs, systemic CSs, systemic agents or conventional synthetic DMARDs). In each group, subjects were categorized according to BMI category (WHO definition): underweight (BMI < 18.5 kg/m²), normal weight (BMI between 18.5 and 24.9 kg/m²), overweight (BMI between 25 and 29.9 kg/m²) and obesity (BMI > 30 kg/m²). In each group, the number of subjects meeting the criteria for MetS was recorded. The criteria of the National Cholesterol Education Program’s Adult Panel III (NCEP-ATP III) for MetS were used [23]. Systematic coronary risk evaluation (SCORE, a CV score of reference in Europe) was used to evaluate CV risk (10-year risk of fatal CV disease) [24, 25]. For patients with PsO, the extent and severity of skin involvement was quantified by the psoriasis area severity index, and quality of life was assessed by the dermatology life quality index. In the PsA group, enthesitis and dactylitis were evaluated using the Leeds Enthesitis Index and dactylitis count, respectively. Disease activity was evaluated by DAS 28 joints-ESR (DAS28-ESR; peripheral joint activity), BASDAI (axial disease activity), and composite psoriatic disease activity index (CPDAI) [26]. HAQ and AS quality of life were also evaluated as measurements of functional impairment and quality of life for axial disease, respectively.

Written informed consent was obtained from each participant, and the study was approved by the local Ethics Committee (Comité de Protection des Personnes CPP Est-II, reference number: 13/411).

Laboratory assessments

Blood samples were obtained from each subject in the morning (8.00 a.m.) after an overnight fast. The blood samples were immediately centrifuged (10 min at 1500 g) and serum was stored (at –80°C) until analysis. Routine laboratory variables that were analysed the day of the visit included inflammatory parameters (ESR, CRP), glycaemia and lipids [total cholesterol, low-density lipoprotein (LDL) and high-density lipoprotein (HDL) cholesterol and triglycerides]. The atherogenic index was calculated as the ratio of total:HDL cholesterol. Analyses performed on frozen serum samples were: circulating IL-6, serum adipokines, insulinaemia as well as serum ghrelin, a gastric peptide involved in appetite regulation. Total adiponectin, high-molecular-weight adiponectin, resistin, retinol-binding protein-4 (RBP4, an adipokine predominantly produced by visceral adipose tissue), leptin and insulin were determined by quantitative sandwich ELISA (R&D systems Europe Ltd, Lille, France). The total ghrelin level was measured by ELISA (Merck Millipore Corp, Billerica, MA, USA). The inter-assay coefficients of variation were 6.8%, 8.6%, 8.2%, 8.6%, 5.4%, 7.4% and 6.2% for total adiponectin, HMW adiponectin, resistin, RBP4, leptin, insulin and total ghrelin, respectively. The lowest level that could be detected by each assay (detection threshold) was 7.8 pg/ml for leptin, 0.89 ng/ml for total adiponectin, 0.98 ng/ml for HMW adiponectin, 0.05 ng/ml for resistin, 30 pg/ml for total ghrelin, 0.3 µU/ml for insulin and 0.63 ng/ml for RBP4. The homeostasis model assessment for insulin resistance (HOMA-IR), calculated as fasting insulin (µU/ml) × fasting glucose (mmol/l)/22.5, was used to estimate insulin resistance [27].

Measurements of body composition

A total body scan was performed using a Lunar densitometer (iDXA; GE Healthcare, Madison, WI, USA). Subjects were scanned using standard imaging and positioning protocols according to the manufacturer’s instructions. Body composition was studied from the total body scan, with measurements of fat mass and lean mass. Adiposity (% fat) was defined as the ratio of total fat tissue to the sum of total lean mass and total fat tissue. Fat distribution was evaluated as the relative proportion of fat tissue in the android (abdominal) and gynoid (hip and thigh) regions. For android fat, a region of interest was automatically defined (from the top of the iliac crest to 20% of the distance from the top of the iliac crest to the base of the skull). Visceral adipose tissue was calculated using the Lunar CoreScan software (GE Healthcare, Madison, WI, USA), which estimates specific intra-abdominal adipose tissue, excluding s.c. abdominal fat. The coefficients of variation for adiposity, fat mass and lean mass were 0.63%, 0.59% and 0.45%, respectively. Quality control scans and calibration were performed daily by using the manufacturer’s standards.

Statistical analysis

The sample size for the study was based on data from a personal series [28] and results by Pedreira et al. [20]. We estimated that a sample size of 52 subjects per patient group would provide 80% power to detect a difference of 150 g in android fat mass, with a two-sided significance level of α = 0.05, assuming a mean fat mass of 2300 g in control subjects and standard deviation of 375 g. Patients and controls were described using mean ± standard deviation (s.d.) or mean and 95% confidence intervals for quantitative variables, and as number and percentage for categorical variables. Comparisons between paired cases and controls (patients with PsA and their paired controls on the one hand, and patients with PsO and their paired controls on the other hand) were performed using McNemar’s test for categorical variables and Student’s paired t test for quantitative variables. In order to control α risk inflation, the pre-planned primary analysis of the body composition parameters involved a gatekeeping strategy. In this strategy, variables are tested sequentially. After the first step, tests are only performed if the result of the previous step is significant. The sequence of tested variables was as follows: android fat mass, visceral fat mass, gynoid fat mass, overall fat mass, adiposity, WC, hip measurement, waist to hip ratio, and lean mass. Comparisons of baseline characteristics and exploratory analyses of adipokines using paired Student t-tests were performed without P-value adjustment. Body composition parameter comparisons between the three groups (PsA, PsO and all the controls) were performed using ANCOVA, with adjustment for age, sex and exact BMI, and Tukey–Kramer post-hoc tests. The significance level was set at P < 0.05. Finally, the relationships between CV risk (evaluated by SCORE) and fat mass measurements, metabolic parameters and RBP4 were analysed using Pearson’s correlation coefficient within groups. Statistical analyses were performed by the research methods unit (uMETh) of the clinical investigation center, University Hospital of Besancon, using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA).

Results

Clinical and socio-demographics characteristics of the study population

A total of 196 subjects were analysed, including 52 patients with PsA, 52 patients with PsO alone and 92 controls (a limited number of subjects were used as controls for both PsA and PsO patients for matching reasons) (Table 1). In the PsA group, patients were perfectly matched with their controls. In the PsO group, patients and controls were well balanced for age, gender and BMI. BMI categories were also balanced between patients with PsA or PsO and their respective controls. Eighty percent of patients with PsA and 36.5% of patients with PsO were receiving a conventional synthetic agent/systemic agent at the time of assessment, and they were all biologic-naïve. A low proportion of patients with PsA (15.4%) received systemic CSs (mean daily dose: 5.8 ± 2.5 mg). A quarter (25%) of patients with PsA were in minimal disease activity. Only 3.8% of patients with PsA met the NCEP-ATP III criteria for MetS, whereas the proportion was higher in the PsO group (26.9%). Cardiovascular risk factors were more prevalent in the patient groups compared with their controls (PsA vs controls: 46.2% vs 34.6%; PsO vs controls: 46.2% vs 38.5%). Finally, the SCORE risk calculation gave low values for both PsA and PsO patients (Table 1).

Table 1

Clinical characteristics and laboratory parameters of disease activity of the patients with PsA or psoriasis and their paired controls

	PsA cases	PaA controls		PsO cases	PsO controls
	Mean (s.d.) or N (%)	Mean (s.d.) or N (%)	Pa	Mean (s.d.) or N (%)	Mean (s.d.) or N (%)	Pa
N	52	52		52	52
Age, years	52.5 (11.7)	52.8 (11.1)	0.69	50.5 (12.8)	50.7 (12.8)	0.76
Gender (M/F)	25 (48)/27 (52)	25 (48)/27 (52)	NC	38 (73)/14 (27)	36 (69)/16 (31)	0.16
Disease duration, years	9.1 (6.7)	—	—	18.1 (13.8)	—	—
BMI, kg/m2	27.4 (5.9)	27.7 (6.5)	0.75	28.4 (5.8)	28.2 (6.1)	0.72
BMI categories, kg/m2			0.11			0.81
Underweight: <18.5	0 0	0 0		1 (1.9)	2 (3.8)
Normal: 18.5–24.99	18 (34.6)	19 (36.5)		13 (25.0)	12 (23.1)
Overweight: 25–29.99	23 (44.2)	18 (34.6)		21 (40.4)	20 (38.5)
Obese: >29.99	11 (21.2)	15 (28.8)		17 (32.7)	18 (34.6)
CV risk factorsb	24 (46.2)	18 (34.6)	0.16	24 (46.2)	20 (38.5)	0.41
Hypertension	15 (28.8)	10 (19.2)	0.20	13 (25)	10 (19.2)	0.44
Diabetes	4 (7.7)	2 (3.8)	0.41	6 (11.5)	4 (7.7)	0.53
Dyslipidaemia	9 (17.3)	9 (17.3)	1	14 (26.9)	12 (23.1)	0.67
Smoking			0.69			0.09
Never	32 (61.5)	35 (67.3)		25 (48.1)	28 (53.8)
Former	11 (21.2)	6 (11.5)		9 (17.3)	14 (26.9)
Current	8 (15.4)	10 (19.2)		18 (34.6)	9 (17.3)
Metabolic syndrome	2 (3.8)	4 (7.7)	0.41	14 (26.9)	4 (7.7)	0.01
SCORE	1.39 (1.89)	—	—	1.53 (1.53)	—	—
Treatments
NSAIDs	6 (11.5)	—	—	0	—	—
CS (systemic)	8 (15.4)	—	—	0	—	—
Topical treatment	3 (5.7)	—	—	4 (7.7)	—	—
MTX	37 (71.2)	—	—	15 (28.8)	—	—
LEF	4 (7.2)	—	—	0	—	—
SSZ	1 (1.9)	—	—	0	—	—
Retinoids	0	—	—	3 (5.8)	—	—
Ciclosporin	0	—	—	1 (1.9)	—	—
Lipid-lowering treatment	9 (17.3)	8 (15.3)	0.76	9 (17.3)	7 (13.5)	0.59
DAS28-ESR	3.5 (1.5)	—	—	—	—	—
BASDAI	4.7 (3.8)	—	—	—	—	—
CPDAI	7.4 (3.3)	—	—	—	—	—
PASI	2.4 (4.1)	—	—	8.4 (4.9)	—	—
Leeds enthesitis index	1.3 (1.5)	—	—	—	—	—
MDA	13 (25.0)	—	—	—	—	—
HAQ	0.75 (0.7)	—	—	—	—	—
ASQoL	8.2 (6.4)	—	—	—	—	—
DLQI	3.9 (5.4)	—	—	10.4 (7.1)	—	—
ESR, mm/h	19.8 (16.6)	7.0 (5.8)	<0.0001	10.8 (8.8)	6.4 (6.5)	0.003
CRP, mg/l	10.5 (11.8)	4.0 (4.8)	<0.001	6.0 (9.0)	4.7 (5.4)	0.38
IL-6, pg/ml	10.1 (15.1)	3.0 (3.0)	0.001	3.6 (3.1)	2.9 (2.4)	0.22

PsO: psoriasis; NC: not calculable; CV: cardiovascular; SCORE: systematic coronary risk evaluation; DAS28-ESR: DAS 28 joints- ESR; CPDAI: composite psoriatic disease activity index; PASI: psoriasis area severity index; MDA: minimal disease activity; ASQoL: AS quality of life; DLQI: dermatology life quality index.

a

McNemar’s test for categorical variables; paired Student’s t test for quantitative variables; ^bmore than one CV risk factor per patient possible.

Body composition and anthropometric measurements

Significantly higher android fat and visceral fat were observed in patients from the PsO group (P = 0.005 and P = 0.017, respectively) (Fig. 1 and Table 2). In PsA patients, we found no difference in android fat mass and visceral fat mass. Lean mass was comparable between the patients and the controls for each group of psoriatic diseases, as well as total fat mass and adiposity. WC and WHR were also comparable for patients and controls in both PsA and PsO groups (Table 2). Finally, we compared body composition measurements among the three populations, i.e. patients with PsA, patients with PsO and the whole group of controls using ANCOVA analysis, after adjusting for age, sex and exact BMI. This analysis showed that android fat and visceral fat was still significantly higher in PsO patients compared with controls (P = 0.0003 and P = 0.0013, respectively), but also compared with patients with PsA (P = 0.0066 and P = 0.0004, respectively). These fat mass measurements did not differ between patients with PsA and controls (all comparisons not significant).

Fig. 1

Box plots of android and visceral fat mass

Box plots of android and visceral fat mass in patients with PsA (N = 52) and their paired controls (N = 52), and in patients with psoriasis without arthritis (N = 52) and their paired controls (N = 52). Fat mass was measured by DXA. Boxes indicate 25th, 50th and 75th percentiles, whiskers indicate minimum and maximum value excluding outliers, and diamonds denote group mean. PsA: patients with PsA; PsA controls: control subjects paired to patients with PsA according to age, sex and BMI category; PsO: patients with psoriasis alone; PsO controls: control subjects paired to patients with psoriasis according to age, sex and BMI category; * and ** reflect the significance of the difference between patients with psoriasis alone and their paired controls determined by Student paired t-test: *P < 0.05; **P < 0.01.

Table 2

Anthropometric and body composition measurements of the patients with PsA or psoriasis and their paired controls

	PsA		PsA controls			PsO		PsO controls
	Mean	CI 95%	Mean	CI 95%	Pa	Mean	CI 95%	Mean	CI 95%	Pb
Andr'oid fat mass, g	2548.9	(2136.4–2961.4)	2424.5	(2070.7–2778.3)	0.51	3183.0	(2724.3–641.6)	2673.7	(2237.9–3109.4)	0.005
Visceral fat mass, g	1149.9	(918.8–1381.0)	1141.2	(865.3–1417.0)	—	1859.3	(1498.7–2219.9)	1532.8	(1157.1–1908.6)	0.017
Gynoid fat mass, g	4153.8	(3642.9–4664.8)	4183.5	(3822.0–4545.1)	—	4318.6	(3796.7–4840.5)	4105.1	(3650.2–4559.9)	0.47
Total fat mass, kg	27.9	(24.4–31.6)	26.9	(24.5–29.4)	—	30.5	(26.8–34.2)	27.9	(24.8–30.9)	—
% body fat	36.5	(33.6–39.5)	37.7	(34.5–41.0)	—	35.7	(33.3–38.1)	35.8	(32.4–39.2)	—
Waist circumference, cm	91.7	(87.8–95.6)	94.1	(90.5–97.7)	—	99.5	(94.7–104.4)	98.2	(93.6–102.7)	—
Hip measurement, cm	102.4	(99.0–105.7)	102.8	(100.4–105.2)	—	104.2	(100.7–107.6)	103.3	(100.5–106.1)	—
Waist to hip ratio	0.90	(0.87–0.92)	0.91	(0.89–0.94)	—	0.95	(0.93–0.98)	0.95	(0.92–0.98)	—
Lean mass, kg	46.6	(44.1–49.1)	47.0	(44.0–50.0)	—	52.4	(49.7–55.3)	51.3	(48.0–54.6)	—

PsO: psoriasis;

P^a: paired Student’s t test between patients with PsA and their paired controls;

P^##: paired Student’s t test between patients with psoriasis and their paired controls.

Serum adipokines, metabolic and lipid parameters

Serum leptin and serum total adiponectin were significantly elevated in PsA patients, while these adipokines did not differ between PsO patients and their controls (Table 3). The leptin to fat mass ratio, serum HMW adiponectin, HMW/total adiponectin ratio, serum resistin and serum ghrelin levels were comparable between patients and controls in the PsA group. Patients with PsO had a reduced HMW/total adiponectin ratio compared with their controls (P = 0.0003). Finally, serum RBP4 levels were higher in patients compared with their controls in both the PsA and PsO groups (all P < 0.05). Lipid parameters did not differ between patients and their controls in the PsA group. Conversely, serum HDL cholesterol levels were found to be lower, and serum triglycerides higher in PsO patients compared with their controls (P = 0.013 and 0.045). Hyperinsulinaemia was observed in both patient groups (all P < 0.05), with a parallel increase in HOMA-IR. However, in the PsA group, the difference for HOMA-IR was borderline significant (P = 0.058).

Table 3

Serum adipokines and metabolic parameters of the patients with PsA or psoriasis and their paired controls

	PsA		PsA controls			PsO		PsO controls
	Mean	CI 95%	Mean	CI 95%	Pa	Mean	CI 95%	Mean	CI 95%	Pb
Leptin, ng/ml	26.6	(18.2–35.0)	19.4	(14.0–24.8)	0.04	21.3	(14.8–27.8)	17.3	(11.5–23.0)	0.19
Leptin/ fat mass	0.79	(0.60–0.98)	0.65	(0.52–0.77)	0.07	0.61	(0.49–0.72)	0.56	(0.41–0.72)	0.81
Total adiponectin, µg/ml	12.1	(10.2–14.0)	9.8	(8.3–11.3)	0.03	8.2	(6.9–9.4)	8.3	(7.2–9.5)	0.91
HMW adiponectin, µg/ml	6.4	(5.3–7.6)	5.6	(4.5–6.6)	0.16	3.8	(3.1–4.6)	4.5	(3.7–5.3)	0.24
HMW/total adiponectin	0.51	(0.48–0.54)	0.54	(0.51–0.57)	0.28	0.44	(0.41–0.47)	0.52	(0.48–0.55)	0.0003
Resistin, ng/ml	9.3	(8.5–10.1)	8.5	(7.7–9.4)	0.17	8.2	(7.5–9.0)	8.7	(7.7–9.7)	0.48
RBP4, ng/ml	64.4	(59.2–69.6)	55.0	(50.7–59.3)	0.004	65.9	(61.2–70.5)	59.3	(55.1–63.4)	0.044
Ghrelin, pg/ml	735.7	(635.0–836.4)	685.1	(601.8–768.3)	0.47	592.2	(521.2–663.3)	655.6	(587.3–723.8)	0.18
Total cholesterol, mmol/l	5.2	(4.9–5.5)	5.3	(5.0–5.7)	0.52	5.0	(4.7–5.3)	5.2	(4.9–5.5)	0.41
LDL cholesterol, mmol/l	3.1	(2.8–3.4)	3.4	(3.1–3.6)	0.26	3.1	(2.8–3.4)	3.0	(2.7–3.3)	0.85
HDL chlolesterol, mmol/l	1.5	(1.4–1.6)	1.5	(1.3–1.6)	0.72	1.2	(1.1–1.3)	1.4	(1.3–1.6)	0.013
Triglycerides, mmol/l	1.4	(1.2–1.6)	1.3	(1.1–1.4)	0.29	1.9	(1.5–2.3)	1.4	(1.2–1.7)	0.045
Atherogenic index (total/HDL cholesterol)	3.8	(3.3–4.3)	4.1	(3.6–4.6)	0.43	4.4	(4.0–4.8)	4.0	(3.6–4.4)	0.11
Glycaemia, mmol/l	5.0	(4.8–5.3)	5.2	(5.0–5.3)	0.43	5.3	(5.0–5.7)	5.3	(4.9–5.8)	1.00
Insulin µU/ml	57.5	(46.5–68.5)	42.5	(35.1–50.0)	0.0177	68.2	(55.6–80.9)	42.8	(36.4–49.2)	0.0006
HOMA-IR	1.8	(1.5–2.1)	1.4	(1.2–1.7)	0.058	2.4	(1.8–2.9)	1.5	(1.2–1.7)	0.0037

PsO: psoriasis; HOMA-IR: homeostasis model assessment for insulin resistance; HMW: high-molecular-weight; RBP4: retinol-binding protein-4;

P^a: paired Student’s t test between patients with PsA and their paired controls;

P^b: paired Student’s t test between patients with psoriasis and their paired control.

Relationships between CV risk and body composition measurements or metabolic parameters and correlation between serum adipokines and disease activity

In the PsO group, there was a moderate correlation between android fat and SCORE (r = 0.32), while visceral fat was strongly correlated with CV risk (r = 0.61) (Fig. 2). No similar relationship was observed between SCORE and RBP4, while the correlation with insulin levels and HOMA-IR was moderate (r = 0.52 and r = 0.52, respectively). In the PsA group, SCORE was mildly correlated with RBP4 (r = 0.39). Finally, we analysed the relationships between indices of disease activity and serum adipokines in each form of psoriatic disease. Results showed that serum leptin mildly correlated with parameters of disease activity (DAS28-ESR, CPDAI, ESR or CRP levels) only in the PsA group, even when leptin was corrected for fat mass (Fig. 3). We did not observe similar results for the PsO patients, except for a mild correlation between leptin and leptin: fat mass ratio and ESR. There was no correlation between the different measures of disease activity and the other adipokines, whatever the patient group. However, serum RBP4 was weakly inversely correlated with CRP levels in patients with PsO (r =–0.27).

Fig. 2

Relationships between fat mass serum RBP4, metabolic parameters and SCORE

Relationships between fat mass (android or visceral region), serum RBP4, metabolic parameters and SCORE in the patients with PsA or psoriasis. PsO: patients with psoriasis alone; SCORE: systematic coronary risk evaluation; RBP4: retinol-binding protein 4; HOMA-IR: homeostasis model assessment for insulin resistance; * indicates P < 0.05.

Fig. 3

Relationships between parameters of disease activity and serum adipokines

Relationships between parameters of disease activity and serum adipokines in patients with PsA or psoriasis. PsO: psoriasis; DAS28-ESR: DAS 28 joints-ESR; CPDAI: composite psoriatic disease activity index; PASI: psoriasis area severity index; RBP4: retinol-binding protein 4; * indicates P < 0.05.

Discussion

In this study, we found that patients with PsO are characterized by visceral fat accumulation, whereas the amount of fat in this region did not differ between patients with PsA and their controls. Studies on adiposity in PsO generally refer to anthropometric measurements such as BMI and/or WC or WHR as indicators of visceral fat deposit, but these indices do not accurately reflect the visceral fat mass. Two studies evaluated body composition in patients with PsO using bioelectrical impedance [17, 18]. In the first, visceral fat was found to be increased in patients with PsO compared with controls [17]. However, among the treated patients, 21% had a biologic agent, a drug class that could influence body composition, especially visceral fat mass [29]. In the second study, no specific measurements of visceral fat mass were reported [18]. Another study in patients with PsO used the DXA technique for body composition assessment, but again visceral fat was not specifically examined [19]. Finally, visceral fat tissue was evaluated in patients with PsO using CT [30]. The results showed that patients with PsO had a higher visceral fat mass compared with their controls and that this fat deposit was associated with the presence of MetS. Our results are in line with this study, showing a similar excess of visceral adiposity in patients with PsO. DXA is considered to be a reliable technique for body composition, and currently available DXA software accurately quantifies visceral adiposity [31]. There was a strong agreement between CT and DXA measurements across people of both genders and across a wide range of BMI values [32]. Data on body composition in patients with PsA are scarce. One study in post-menopausal women with PsA found a higher total body fat amount in the patients compared with the controls with an android distribution, but the intra-abdominal region was not specifically examined [20]. In a second study, an excess of adipose tissue was observed in patients compared with their controls, with a high prevalence of abdominal fatness [21]. However, 20% of these patients had a TNF inhibitor, which could have biased the results. In our PsA patients, we did not find evidence of changes in body composition, especially in android or visceral fat mass. Discrepancies between our results and those of previous studies may be explained by the patient characteristics and/or the selection of the controls. Indeed, we chose our controls with characteristics as close as possible to the patients. In addition, we only selected biologic-naïve patients.

The adipokinome may be viewed as a bridge between obesity and systemic inflammatory diseases [33]. Adipokines have been consistently studied in PsO [34, 35]. Collectively, these studies showed elevated serum levels of pro-inflammatory adipokines such as leptin or resistin [36]. Studies evaluating total and HMW adiponectin in PsO usually found decreased levels [34, 36]. These results may be related to the anti-inflammatory properties of adiponectin, but also to its metabolic function by promoting insulin sensitivity [37]. In our PsO patients, we found no abnormal values for serum leptin, resistin, or ghrelin, while the HMW/total adiponectin ratio was reduced. HMW/total adiponectin has been linked to metabolic disturbances, and a reduced ratio is able to predict insulin resistance [38]. These results are in keeping with the metabolic profile of our PsO patients, who displayed high insulin levels and HOMA-IR. Results on serum adipokines in patients with PsA are more limited. Collectively, serum leptin, total adiponectin and resistin were found to be elevated compared with a control population [39–42]. However, in two studies, a control group was lacking and leptin was not corrected for fat mass [39, 42]. In addition, specific adiponectin isoforms were not evaluated [39, 41]. In our PsA patients, we observed higher values for serum total adiponectin without a parallel change in HMW adiponectin, a result that has been previously reported in different immune-mediated diseases [37]. Finally, we also evaluated RBP4 in our patients. RBP4 is secreted by adipocytes and hepatocytes, with visceral fat being its main source [34]. RBP4 plays a key role in insulin resistance and contributes to the development of MetS [43]. Previous studies evaluating RBP4 in patients with PsO yielded controversial results, which may be explained by the patient characteristics, but also by the selection of controls [34, 36, 44]. In our patients (PsA and PsO), serum RBP4 levels were significantly increased, highlighting the link between fatness, visceral fat accumulation and metabolic disturbances.

Psoriatic diseases have an increased risk of developing cardiometabolic diseases compared with the general population [1, 45]. MetS is highly represented in both PsO [46] and PsA [47]. It has previously been reported that obesity and MetS are more prevalent in patients with PsA compared with patients with PsO alone [39, 48]. However, in these previous studies, visceral fat and anthropometric measurements of abdominal adiposity were not available [48], or more obese patients were selected in the PsA group [39]. In our series, the prevalence of MetS was low among PsA patients and moderate in the PsO group. As stated above, we matched patients and controls according to BMI category. In addition, a proportion of our patients received anti-hypertensive and/or lipid-lowering drugs. Our results also showed that the CV risk as measured by SCORE strongly correlated with visceral fat and insulin resistance in patients with PsO. Surprisingly, RBP4, an indicator of insulin resistance, did not correlate with SCORE. This may be explained by the small sample size. Our data clearly showed a high frequency of visceral adiposity in patients with PsO compared with patients with PsA. The differential frequency of obesity among PsA and PsO patients depends on the definition that is used. Since an excess of visceral fat is an important predictor of metabolic and atherosclerotic diseases [49], our results suggest that (visceral) obesity may differently influence the CV risk in PsO relative to PsA. In this sense, one might consider that visceral adiposity is a more serious concern in PsO compared with PsA. However, our patients were receiving systemic agents or conventional synthetic DMARDs, specifically MTX, which has been shown to have a favourable impact on CV outcomes. This may have dampened the CV risk of these patients [50].

In conclusion, we found that visceral fat accumulates more in patients with PsO without arthritis, than in patients with PsA. Together with metabolic disturbances such as insulin resistance and MetS, HMW/total adiponectin is also unbalanced in PsO. In addition, RBP4 could be a relevant adipokine for evaluating visceral obesity in psoriatic diseases. These findings highlight the relationships between visceral obesity, metabolic disturbances and the CV risk in patients with PsO, and to a lesser extent in patients with PsA.

Acknowledgements

Thanks to Fiona Ecarnot, PhD, EA3920, Cardiologie, CHU de Besançon, Besançon, France, for her help in preparing the manuscript, and to Mrs Sylvie Cour, INSERM CIC-1431, Centre d’Investigation Clinique Biothérapie, CHU de Besançon, Besançon, France, for her help in control subject recruitment. E.T. was the main investigator of the study. He designed the study together with G.D. and E.M.. E.T., D.W., B.A., J.G., O.M. and X.G. contributed to PsA patient recruitment, while F.A. was responsible for PsO patient recruitment. Data analysis and interpretation was done by M.D. Authors G.D. and C.L. were responsible for laboratory assessments. The manuscript was primarily drafted by E.T. All authors have read and approved the final version of the manuscript.

Funding: E.T. received a grant from the CHU de Besançon, France, to conduct this study. The sponsor of the trial was the CHU de Besançon, France.

Disclosure statement: The authors have declared no conflicts of interest.

Data availability statement

The data used and analysed during the present study are available from the corresponding author on reasonable request.