Altered body composition and increased visceral adipose tissue in premenopausal and late postmenopausal patients with SLE

Systemic lupus erythematosus (SLE) is one of the most complicated autoimmune diseases, and it can affect any organ and present with diverse phenotypes. It affects women more frequently than men at a ratio of nearly 9 to 1. The prognosis of SLE patients has markedly improved due to the introduction of immunosuppressive regimens over the last decades. The 5-year survival rate of SLE has exceeded 90% in recent years [1].

However, as life expectancy increases, subsequent complications such as cardiovascular disease (CVD) are becoming a more and more serious clinical problem [1]. Manzi et al. [2] demonstrated that women with SLE (most of them were Caucasian, the rest were African American, American Indian, Asian American, and Eastern Indian.) in the 35–44-year age group were over 50 times more likely to have a myocardial infarction (MI) than age-matched controls. Another southern Sweden prospective study found that the incidence of MI in patients with SLE was nine times higher than that in a control population [3]. Recently, a Danish nationwide study showed that patients with SLE without lupus nephritis (LN) were more than twice as likely to have a MI as the control subjects, and the hazard ratio was 18.3 in patients with LN [4]. The mechanism of CVD in SLE is multifactorial. Traditional CV risk factors, such as age, hypertension, dyslipidemia, and smoking, cannot fully explain the increased CVD occurrence in SLE [5‐7], and disease-related risk factors, such as disease activity, cumulative damage, renal involvement, inflammatory mediators, and medications, have been implicated in increasing the CV risk in SLE [8]. Recently, Seguro et al. showed that SLE was associated with altered adiposity distribution and increased visceral adipose tissues (VAT) in premenopausal SLE patients, and that VAT was correlated with traditional risk factors for CVD [9].

Prospective studies identifying risk factors for incident coronary heart disease had shown that VAT was a predictive factor of coronary heart disease independently of age and body mass index (BMI) [10]. The possible mechanism was justified by more and more studies, which related it to highly lipolytic visceral deposits releasing fatty acids into the portal vein that subsequently travel to the liver. Here, they cause hepatic insulin resistance and lead to hyperinsulinemia and accelerated synthesis of glucose by the liver [11]. Moreover, potential protective factors for CVD and diabetes, such as leptin, adiponectin, and peroxisome proliferator-activated receptor-gamma, were expressed at lower levels in visceral than in subcutaneous adipose tissue [12]. Thus, an assessment of VAT may be necessary for the evaluation of CVD risk in SLE patients.

Unfortunately, there were only two studies examining VAT and its determinants among patients with SLE [9, 13], and the changes of VAT in postmenopausal patients with SLE and in Chinese patients have never been reported. Since the prevalence and disease activity of SLE were found to have geographical variations, SLE patients in Asia may be different from those on other continents. Our study aimed to examine body composition, especially VAT, in a population of Chinese women with SLE, and compare it with age-, sex-, and BMI-matched controls, and then to explore the association between VAT and other clinical parameters in patients with SLE.

Although past studies observed that patients with SLE had increased CV risk, the specific mechanism was still unknown. Here, we focused on body composition, especially VAT, which may have an effect on the CV risk in SLE patients. The significant findings of the present study were that the patients with SLE did have higher VAT volume and mass in both premenopausal and late postmenopausal patients. Moreover, we also found that some disease-related factors correlate with VAT mass, which could be concluded as patients with longer disease duration, poor disease control, and organ damage may have a higher VAT mass. Other CV risk factors, such as hypertension, higher serum uric acid, TC, and TG level, were also related to higher VAT volume.

In most cases, the VAT measurements are performed by computed tomography (CT) or magnetic resonance tomography (MRI). However, because these imaging techniques are expensive, their use is limited in large-scale studies. Recently, some investigators used DXA to provide accurate quantitative assessments of both total and regional adiposity [18, 19]. The comparing study has shown that DXA correlated well with gold standard MRI and CT, and provides a low radiation, efficient, cost-effective option despite its underestimation of VAT to some extent [20, 21]; therefore, we used DXA to evaluate the VAT in this study.

Similar to other studies, we found that our patients had some traditional risk factors for CVD, such as higher serum creatinine and uric acid, and lower HDL-C. Wang et al. [22]reported that the frequency of CVD was high in Chinese patients with SLE, and higher serum creatinine levels and lower HDL-C were the risk factors for CVD. Yang et al. [23]also found that elevated serum creatinine was an independent risk factor for CVD. Daniele Machado et al. [22]reported that the most significant difference of lipid profile between adolescent females with juvenile SLE and healthy controls was lower HDL-C, whereas TC, LDL-C, TG, and non-HDL-C were not different between the two groups, and also suggested low HDL-C might contribute to an increased atherosclerotic risk. Studies demonstrated that the TG/HDL-C ratio was more useful than isolated lipid values, as it more closely reflects the complex interactions of lipoprotein metabolism [24]. Recent studies indicated a high level of the TG/HDL-C ratio had been associated with insulin resistance, obesity, and metabolic syndrome [25, 26]. We actually found that the TG/HDL-C ratio of our patients was higher than that of the control group, which may be associated with higher CVD risk. In 2018, a Brazilian cross-sectional study demonstrated that the TG/HDL-C ratio was higher in dyslipidemic SLE patients than the others and it was correlated with disease activity [27]. Our group of SLE patients had lower serum TC and LDL-C level than controls, which was not consistent with previous studies [23]. The possible reason was that most of our patients had a well-controlled disease and were under minimal glucocorticoid therapy and long-term hydroxychloroquine treatment. A longitudinal study demonstrated that antimalarials could significantly decrease TC and LDL-C in SLE patients [28].

According to our study, SLE patients likely had lower BMC and LM compared with the control group, especially in premenopausal patients. Accelerated rates of bone and muscle loss have reported before in patients with SLE [29, 30]. The inflammatory nature of SLE, decreased physical exercise, malnutrition, vitamin D supplementation, and glucocorticoid therapy have been found to be associated with this phenomenon [31].

So far, only two studies have examined the relationship between visceral fat and SLE. In 2013, Shields et al. [13] used electron beam computed tomography to evaluate descending thoracic aortic perivascular adipose tissue (PVAT) solely, which is a visceral adipose deposit in close proximity to blood vessels. They concluded that the volume of PVAT was greater in female SLE patients than in age- and race-matched controls and associated with calcification in different vascular beds, which could influence CVD. LPC Seguro et al. [9] used DXA to evaluate VAT in premenopausal patients with SLE, and they found that SLE was associated with increased VAT, and the latter was correlated with traditional CV risks. Our result was partly consistent with their result that premenopausal SLE patients with BMI less than 25 had a higher VAT parameter than controls. Furthermore, we found the same change in late postmenopausal SLE patients, but not in perimenopausal patients. The possible reason was that patients with SLE were much younger than controls in the perimenopausal group, and the age factor may offset the effect of the disease itself on VAT. In addition, perimenopausal women have significant individual differences in hormone secretion, which may be another confounding factor affecting VAT.

Our results showed that VAT was related to traditional CV risks, such as age, BMI, menopausal status, hypertension, serum uric acid, serum creatinine, non-HDL-C level, and serum TG level, in both patients and controls. Many studies had verified that the amount of VAT increased with age in both genders and both lean and overweight/obese people [12]. Indeed, the visceral deposit was increasing with BMI increase, whereas the correlation between BMI and adipose distribution was different among individuals [12]. Thus, it was necessary to identify a VAT parameter other than BMI to assess the CV risk. The association between VAT and serum TC, non-HDL-C, LDL-C, and TG levels in the general population was reported before, and was due to higher lipolytic activity and weaker antilipolytic effect of insulin in visceral adipocytes. The mechanism of this phenomenon was complicated, but the different expression and sensitivity of the receptors in visceral and subcutaneous adipocytes was the key factor [12].

In addition, LPC Seguro et al. [9] failed to find that VAT was related to disease duration, SLEDAI score, SLICC/ACR-DI, or current glucocorticoid use. In the present study, we identified several SLE disease-related factors positively associated with increased VAT: SLE disease duration and SLICC/ACR-DI score, whereas low complement C3/C4 was negatively associated. Kravvariti et al. [32] found that disease duration was a determinant of subclinical atherosclerosis progression in SLE patients, whose mechanism was multifactorial and possibly it may have been associated with increased VAT. The basis of the association of complement with VAT was not clear. Because there was no known biologic link between compliment and adipose accumulation, the association may relate to compliment as a disease severity marker.

The previous study demonstrated that long-term low-dose prednisone exposure was associated with increased visceral fat in patients with rheumatoid arthritis [33]. Glucocorticoid could preferentially upgrade lipoprotein lipase expression and activity in visceral adipose tissue, but not subcutaneous adipose tissue. Thereby, a higher rate of free fatty acid delivery from triglyceride-rich lipoproteins might contribute to the visceral fat accumulation [34]. We failed to find glucocorticoid exposure in SLE patients related to VAT; the possible reason was that the majority of patients were under long-term glucocorticoid therapy. And, because the dose of glucocorticoid varied in different periods among individuals, it was difficult to distinguish the influence of glucocorticoid on VAT in a cross-sectional study. According to the related studies, it is still safe to say that clinicians should minimize glucocorticoid use to the possible minimum dose to prevent underlying VAT-associated CV events.

There were some confounding factors that may have association with VAT which we did not explore in this study. A large community-based cross-sectional study demonstrated that participants who adhered to recommended dietary guidelines and physical activity (PA) had lower VAT volumes in White people [35]. Other studies also showed that PA intervention was negatively correlated with VAT in obese/overweight population [36‐38]. Ethnicity was another confounding factor affecting VAT accumulation [39]. Iris A. Lesser et al. [40] found that when ethnic differences in PA were taken into account, there were no longer any differences in VAT between the Chinese and European groups, while VAT remained higher in South Asians than Europeans. Due to this was a preliminary study, we did not take all these factors into account. In the further study, an exercise intervention study is necessary to further elucidate the effects of PA on VAT in ethnic patients with SLE.

Some limitations of our study are the following: First, the sample size of our study was relatively small, and it was a single-center study. Second, the study was cross-sectional in nature, which made it difficult to explain the cause-and-effect correlation. Third, most of our patients had well-controlled disease and low glucocorticoid dose, which could not reflect the diversity of SLE disease. In addition, DXA was not the most accurate method to measure VAT, which may have some influence on the results. Besides, there were some confounders, such as PA and dietary quality, which affected VAT that we could not control. Notwithstanding its limitations, this study does suggest that patients with SLE in China have increased VAT, and it was associated with traditional CV risk factors and related to the disease itself. To fully identify the associations between SLE and VAT distribution, large-scale, long-term cohort studies should be performed to exclude other confounding factors, such as diet and exercise.

We actually found some traditional CV risk factors in SLE patients, such as low HDL-C, high TG/HDL-C, high serum creatinine, and high uric acid. Furthermore, this study suggested the VAT increases significantly in SLE patients, especially in a premenopausal and late postmenopausal period. The increased VAT in patients with SLE was associated with traditional cardiometabolic risk factors, which may contribute to CV risk in SLE populations. Cohort studies are necessary to validate the long-term effect of VAT on CV complications in SLE. Some other confounding factors, such as PA, dietary quality, and ethnicity, should also take into account in the further study.