INTRODUCTION

The increase in life expectancy of patients with systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA), associated with the development of new treatment regimens in recent decades, may lead to an increase in the prevalence of comorbidities such as coronary heart disease (CHD), heart failure (HF), and type 2 diabetes mellitus (DM) in middle and old age. Primary prevention of this socially significant pathology in the general population is based on timely detection and correction of reversible risk factors (dyslipidemia, obesity, arterial hypertension, and hyperglycemia). However, in immunoinflammatory rheumatic diseases (IIRDs), the significance of some of the above factors is not so clear. For example, in RA, the development of cardiovascular complications (CVCы) was associated with a decrease, but not with an increase, in the level of total cholesterol and low-density lipoproteins (“lipid paradox”) [1], and mortality was associated with low body weight and weight loss (“obesity paradox”) [2, 3].

The pathogenesis of both cardiovascular diseases and carbohydrate metabolism disorders in IIRDs is based on the combined influence of autoimmune chronic inflammation, antirheumatic drugs, and conventional cardiovascular risk factors, the standard assessment of which often gives conflicting results. For example, in IIRDs, the accumulation of fat mass can be either clinically pronounced, raising no doubt about obesity upon examination, or hidden, if it is combined with sarcopenia. Conventional screening methods—determining the body mass index (BMI) and waist circumference (WC)—do not reflect the actual body composition [4–6]. For example, in RA and systemic sclerosis, the proportion of patients with excess total fat mass, according to dual-energy X-ray absorptiometry, was 71.1 and 69.6%, although only 54.4 and 50% of patients had a BMI ≥25 kg/m2, respectively. Pathological combined body composition, including obesity on the one hand and osteoporosis and/or sarcopenia on the other, was detected in 32.4 and 45.7% of cases, respectively [7]. In addition, anthropometric measures cannot be used to assess the existing metabolic disorders caused by excess weight, although some evidence suggests that “metabolic ill-health” in the population is associated with adverse outcomes to a greater extent than BMI [8–10].

In most cases, the term “metabolic ill-health” is understood as the presence of certain signs of metabolic syndrome and insulin resistance (IR). An increase in the level of C-reactive protein (CRP) > 3 mL/L as a marker of subclinical inflammation [10–12] is often considered as an additional criterion. However, in IIRDs, the listed parameters can be distorted due to the presence of active inflammation, autoimmune reactions, the consequences of kidney damage as part of the underlying disease, and the use of a number of antiinflammatory drugs and can quickly change in a short period of time. Giraud et al. [13] demonstrated that 16% of RA patients with normal weight, 35% with overweight, and 15% with obesity had metabolic syndrome; the body composition in non-obese patients did not differ between the metabolic health and ill-health groups.

It is necessary to search for new “early” but more stable biomarkers to identify prognostically unfavorable phenotypes of obesity, overweight, and normal weight. In this aspect, leptin, a cytokine produced by adipocytes, which is responsible for the regulation of appetite and energy metabolism, is of great interest. Currently, it is believed that it also plays an important role in metabolic dysregulation. In the general population, patients with obesity and hyperleptinemia are more likely to have IR, type 2 diabetes mellitus, hypertension, and myocardial hypertrophy [14–18]. Leptin levels directly correlate with carotid intima-media thickness and coronary artery calcification, progression of atherosclerosis, heart failure, and severe coronary artery disease [19–25]. Data on leptin levels in RA and SLE, as well as on its relationship with metabolic disorders, are contradictory [26–29].

The objective of this study was to identify different phenotypes of overweight in women with systemic lupus erythematosus and rheumatoid arthritis based on two main indices—body mass index and blood serum leptin level, as well as clarifying the frequency of various metabolic disorders, arterial hypertension, and cardiovascular complications in certain phenotypes.

MATERIALS AND METHODS

The observational study included 96 patients: 50 women with RA and 46 women with SLE. Inclusion criteria were a definite diagnosis of RA (according to the criteria of ACR/EULAR (American College of Rheumatology/European Alliance of Associations for Rheumatology) from 2010 [30]) or SLE (according to the criteria of SLICC (Systemic Lupus International Collaborating Clinics/ACR) from 2012 [31]) and availability of informed consent. Exclusion criteria were age under 18 or over 65 years, pregnancy and lactation, history of diabetes, fasting hyperglycemia (venous blood glucose ≥6.1 mmol/L), and/or taking glucose-lowering drugs at the time of examination. The characteristics of patients with RA and SLE are presented in Table 1 and Table 2, respectively.

Table 1. Characteristics of rheumatoid arthritis patients (n = 50)
Full size table
Table 2. Characteristics of systemic lupus erythematosus (n = 46)
Full size table

In all patients, height and weight, waist circumference (WC) and hip circumference (HC) were measured, BMI and WC/HC ratio were calculated, and the concentrations of leptin (DBS kits (Diagnostics Biochem Canada Inc., Canada) for enzyme immunoassay), insulin (Elecsys kit for electrochemiluminescent analyzer Cobas e411 (Roche Diagnostics, United States), glucose, and total cholesterol in fasting blood serum was determined. Blood pressure (BP) was measured, taking into account the need for constant use of antihypertensive drugs (as a marker of hypertension). Information about previous cardiovascular events (myocardial infarction, operations for its revascularization, and stroke) was collected. The HOMA-IR index was calculated using the formula: glucose (mmol/L) × insulin (µU/mL) /22.5 [32]. Hyperleptinemia was diagnosed when leptin concentration was >11.1 ng/mL, IR was diagnosed when HOMA-IR values were ≥2.77.

Three main phenotypes of overweight were distinguished:

1) “classic” (BMI ≥ 25 kg/m2 + hyperleptinemia);

2) “healthy” (BMI ≥ 25 kg/m2, no hyperleptinemia);

3) “hidden,” or “latent” (BMI < 25 kg/m2 + hyperleptinemia).

“Normal weight” was diagnosed at BMI < 25 kg/m2 in the absence of hyperleptinemia.

Statistical processing of data at all stages of the study was performed using parametric and nonparametric statistics. When comparing two independent groups for qualitative characteristics, the χ2 method was used (if necessary, with the Yates correction for low frequencies); for quantitative characteristics, the Mann–Whitney test was used; when comparing three groups, the Kruskal–Wallis method was used. Differences were considered statistically significant at p < 0.05; for multiple comparisons, taking into account the Bonferroni correction, at p < 0.017.

Patients with RA and SLE were comparable in age (p = 0.4), duration of IIRDs (p = 0.2), BMI (p = 0.5), WC (p = 0.4), and the proportion of participants in menopause (p = 0.1); however, the WC/HC ratio was higher in the SLE group (p = 0.03). In RA, the median leptin concentration was 10.8 [5.2; 23.1] ng/mL; in SLE, 29.6 [10.3; 75.5] ng/mL (p < 0.0001). In RA, hyperleptinemia was found in 23 (46%) patients; in SLE, in 34 (74%) patients (p = 0.005). IR was found in 5 (10%) and 10 (22%) patients, respectively (p = 0.2). Insulin and HOMA-IR levels were higher in SLE than in RA (8.26 [2.89, 12.05] vs. 5.32 [3.85, 7.72] ng/mL (p = 0.0007) and 1.90 [1.23; 2.52] versus 1.18 [0.88; 1.78] (p = 0.001), respectively). Concentrations of glucose (p = 0.3) and total cholesterol (p = 0.6) did not differ.

The distribution of overweight phenotypes was as follows: in RA, the “classic” phenotype was diagnosed in 15 (30%), “healthy” in 4 (8%), and “hidden” in 8 (16%) patients; in SLE, in 20 (44%), 0 (0%), and 14 (30%), respectively (Fig. 1). In subgroups under 45 years of age with RA (n = 30), the “classic” phenotype was found in 4 (13%) cases, “healthy” in 3 (10%), “hidden” in 3 (10%) cases; in SLE (n = 30), in 10 (33%), 0 (0%), and 12 (40%) cases, respectively.

Fig. 1.
figure 1

Frequency of overweight phenotypes in patients with rheumatoid arthritis (a) and systemic lupus erythematosus (b). BMI—body mass index, N—normal values.

Full size image

“Healthy” overweight phenotype was detected only in the patients with RA. BMI in this phenotype averaged 27.6 kg/m2, median WC was 97 [87; 102] cm, and WC/HC was 0.9 [0.83; 0.99], which is higher than in the group of patients with “normal weight” (p < 0.001 for all parameters). Age, duration and disease activity according to DAS28 (Disease Activity Score 28), levels of inflammatory markers (erythrocyte sedimentation rate and CRP), cholesterol, glucose, HOMA-IR index, systolic and diastolic blood pressure, and frequency of use of antihypertensive drugs were comparable to those in patients with “normal weight.”

Patients with “classic” and “hidden” overweight phenotypes in RA were older than those in the “normal weight” group and were more likely to be in menopause. The “hidden” phenotype was also characterized by a tendency towards longer duration of RA and more frequent use of GCs and genetically engineered biological drugs (biologics; Table 3).

Table 3. Characteristics of rheumatoid arthritis patients with “classic” or “hidden” overweight phenotypes and with “normal weight”
Full size table

In SLE, differences in three groups (“classic”, “hidden” overweight phenotype, and “normal weight”) were revealed not only in age, but also in the maximum dose of GCs for the entire duration of the disease, as well as a tendency towards shorter duration and greater activity of IIRDs in the last of them (Table 4).

Table 4. Characteristics of systemic lupus erythematosus patients with “classic” or “hidden” overweight phenotypes and with “normal weight”
Full size table

In patients with a “healthy” overweight phenotype, there was no history of taking antihypertensive drugs or cardiovascular events. There were no differences with the subgroup with “normal weight” in the level of systolic and diastolic blood pressure, insulin concentration, total cholesterol, glucose, and HOMA-IR index. Metabolic parameters for the “classic” and “hidden” phenotypes are presented in Table 5.

Table 5. Metabolic parameters in rheumatoid arthritis and systemic lupus erythematosus patients with “classic” or “hidden” overweight phenotypes and with “normal weight”, Me [25th; 75th percentile]
Full size table

The frequency of IR increased from 1 (3%) case at “normal weight” to 4 (18%) cases at “hidden” and 10 (29%) cases at the “classic” phenotype (p = 0.009 when compared with “normal weight”). A gradual increase in the proportion of patients receiving antihypertensive drugs was demonstrated: from 2 (6%) patients with “normal weight” to 10 (45%) patients with a “hidden” phenotype (p = 0.0012 when compared with “normal weight”) and up to 23 (66%) patients with the “classic” phenotype (p < 0.0001 when compared with “normal weight”). Systolic (120 [115; 140] mm Hg) and diastolic (80 [70; 85] mm Hg) blood pressure at the “classic” phenotype turned out to be higher than at “normal weight” (110 [110; 120] mm Hg (p = 0.001) and 70 [65; 80] mm Hg (p = 0.003), respectively).

In two patients with RA and two patients with SLE, a history of one episode of cardiovascular events was registered: one had severe coronary artery disease, for which stenting of the coronary arteries was required, one had myocardial infarction, and two had acute cerebrovascular accident (ACVA). In three cases, women with CVCs had “classic” overweight phenotype, and one SLE patient with stroke had “normal weight” phenotype.

The study demonstrated that, in women under 65 years of age with SLE, despite similar BMI, levels of leptin, insulin, and the HOMA-IR index are higher, and the proportion of patients with hyperleptinemia is greater than in RA. At the same time, no statistically significant differences in the frequency of IR in RA and SLE were detected.

In the foreign and Russian literature, no information on the comparison of leptin concentrations in RA and SLE was found, although there are reports of both increased and similar levels of adipocytokine in these IIRDs compared to healthy controls [26–29]. In a number of studies, the frequency of IR in RA was greater than in SLE [33, 34], and in other studies it was comparable [35, 36]. The identified differences may be related to the demographic characteristics of patients and the method for determining IR. Characteristic features of our cohort were a relatively young age and the absence of obvious disorders of carbohydrate metabolism (DM and fasting hyperglycemia).

It is known that serum leptin concentration depends on the content of adipose tissue in the body, including in RA, and can potentially be used to assess its amount [37]. The combination of hyperleptinemia with low BMI may indirectly indicate sarcopenia and sarcopenic obesity.

On the other hand, an increase in leptin synthesis reflects morphofunctional changes in adipose tissue, as it is associated with hypertrophy and apoptosis of adipocytes, recruitment of immune cells (primarily M1 macrophages), and the production of proinflammatory cytokines that contribute to the development of IR and epithelial dysfunction [22, 38]. Within the framework of this concept, hyperleptinemia should be considered as the first stage in the formation of “metabolic ill-health,” which subsequently leads to various adverse outcomes. This point of view is confirmed by the results of two studies in RA. In one of them, high leptin levels were associated with an increased risk of death due to cardiovascular causes, with stronger correlations observed in non-obese patients [39]. In another, high leptin level was associated with the risk of various concomitant diseases, including dyslipidemia, hypertension, diabetes, heart failure, as well as overall mortality and mortality due to malignant neoplasms [40].

In this work, we propose to use the determination of leptin in combination with BMI to identify individuals with different overweight phenotypes—“classic”, “hidden”, and “healthy”, reflecting a certain level of metabolic disorders. Slightly less than half of the patients with RA and about a quarter of the patients with SLE had normal weight without hyperleptinemia. Among all the overweight phenotypes, “healthy” turned out to be quite rare, which was found only in some, mostly young, patients with RA. Metabolic parameters at the “healthy” phenotype did not differ from those in women with a BMI of up to 25 kg/m2 without hyperleptinemia. In general, the “classic” phenotype was diagnosed most often both in RA and in SLE, although in the subgroup under 45 years of age it gave way to the “hidden” phenotype. These two phenotypes, compared with the patients who had “normal weight,” were characterized by changes in biochemical parameters (increased concentrations of total cholesterol, insulin, and HOMA-IR index), although not always clinically significant, a higher frequency of hypertension. The “classic” phenotype was also characterized by an increase in IR.

Women with a “classic,” but not “hidden,” phenotype, as well as one patient with “normal weight,” reported CVC events. In the last case, risk factors not associated with obesity (for example, antiphospholipid syndrome, the presence of which was confirmed in this SLE patient) could have played a role in the development of stroke. The fact that “hidden” overweight, unlike “classic” overweight, was not associated with any cardiovascular events confirms that it can be considered as a certain “window of opportunity” for primary prevention.

The detection of metabolically unfavorable phenotypes depended on age and GC use. Low disease activity was of additional importance in SLE. All this apparently led to excess accumulation of fat mass. RA patients with a “hidden” phenotype received biologics somewhat more frequently. It is known that the use of drugs from the group of tumor necrosis factor α inhibitors promotes weight gain due to the accumulation of adipose tissue [41, 42]. The possibility of indirect effects of biologics should also be taken into account. Their prescription for RA, as a rule, is a consequence of a severe course of the disease, which could cause the development of sarcopenia and a decrease in BMI in previously obese patients and, on the other hand, serve as a basis for the use of GCs. The effect of IIRDs therapy on the formation of various overweight phenotypes requires further research.

Thus, the proposed stratification of patients with IIRDs allows us to identify two subgroups of patients— with “classic” and “hidden” overweight phenotypes—who require regular monitoring of lipid profile, blood pressure, insulin levels, as well as prevention of carbohydrate metabolism disorders and cardiovascular events. In this case, basic measures should be aimed, first of all, at reducing the mass of adipose tissue, including a calorie-restriction diet, regular long-term aerobic exercises, and, if they are ineffective, drug therapy. At a “healthy” phenotype, weight maintenance and sarcopenia prevention are recommended. The introduction of appropriate recommendations into the complex of treatment of patients with IIRDs will help reduce the prevalence of obesity-associated diseases such as type 2 diabetes mellitus, coronary artery disease, and heart failure, as well as reduce material costs for their treatment and increase the life expectancy of patients.