Background
Obesity, often defined as an elevated body mass index (BMI), has been implicated in the development and progression of adverse cardiovascular outcomes in the general population. However, the long-term prognostic value of obesity may be reversed in specific populations, especially in patients with chronic illness such as coronary artery disease (CAD) and chronic kidney disease (CKD) [
1]. Recent evidence has indicated that the protective effect of obesity may be conferred only to patients who have a healthy metabolic profile [
2‐
4]. Namely, various metabolic risk factors such as insulin resistance, decreased muscle mass, lipid disorders, and chronic inflammation may influence the relationship between obesity and long-term outcomes in end-stage renal disease (ESRD) patients [
5]. Indeed, obese ESRD patients who were metabolically healthy had a lower risk of mortality than normal weight ESRD patients who were metabolically healthy [
4]. However, there was no mortality benefit in obese but metabolically unhealthy ESRD patients. In this context, there is an important limitation when using BMI to diagnose obesity. Patients with higher fat mass may have less favorable metabolic profiles than those with higher muscle mass, even if their BMI is same. Therefore, some authorities advocate a definition of obesity based on sex-specific values for percentage of body fat (PBF) that correspond to internationally recognized BMI cut-points for defining underweight, overweight and obesity [
6].
Apart from the well-established role of vitamin D in CKD-mineral bone metabolism (MBD), more recent studies have established vitamin D as a representative parameter of cardiometabolic disturbance [
7‐
9]. Low serum levels of 25-hydroxyvitamin D (25(OH)D) have been associated with higher prevalence of cardiovascular disease, diabetes, hypertension, obesity, and dyslipidemia [
10‐
13]. Although the causal relationship between these associations has been unclear, their association may be evidentthe fact that obese population had low vitamin D levels is well-established. In fact, obesity is a risk factor for both vitamin D deficiency [
14] and cardiovascular disease [
15,
16]. A systematic meta-analysis showed a significant inverse association between BMI and vitamin D deficiency status in an adult population, and vitamin D supplementation improved cardiometabolic profiles [
16].
Whether vitamin D levels can modify the link between obesity and cardiovascular risk is unclear. The purpose of this study was to evaluate the potential role of vitamin D for modifying the correlation between obesity, defined as elevated percentage of body fat (PBF), and vascular calcification score (VCS) in incident ESRD patients.
Methods
Study population and data collection
This study consisted of 269 ESRD patients who initiated dialysis between July 2011 and June 2015. We excluded patients who received oral vitamin D supplements in outpatient clinics during CKD management (n = 48), had a previous history of peritoneal dialysis or received kidney transplantation (n = 5), and who had an inserted pacemaker (n = 3), for a total of 213 patients included in the study. This study was approved by the institutional review board/ethics committee of Hallym University Sacred Heart Hospital, Anyang, Korea, and all study procedures are in adherence to the Declaration of Helsinki.
Data collected included age, sex, underlying cause of renal disease, comorbidities, systolic and diastolic blood pressure (BP), and various laboratory parameters. Biochemical analyses of hemoglobin, serum albumin, total cholesterol, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, triglycerides (TG), calcium, phosphate, intact parathyroid hormone (iPTH), 25(OH)D, and 1,25-dihydroxyvitamin D (1,25 (OH)2 D) levels were carried out at the start of dialysis. Levels of 25(OH)D were determined using an electrochemiluminescence method (Roche Cobas 8000 System, Tokyo, Japan). Levels ≥ 30 ng/mL were regarded as normal, while low levels between 10 – 30 ng/mL were considered to indicate vitamin D insufficiency, and very low levels < 10 ng/mL were categorized as vitamin D deficiency. In addition, more decreased 25(OH)D levels < 3 ng/mL were designated as severe vitamin D deficiency. Also, serum high-sensitivity C-reactive protein (hs-CRP) levels were measured using a Behring Nephelometer (BN) II Analyzer (Dade Behring, Newark, Del., USA) by a latex-enhanced immunoephelometric method. All other tests were performed according to the manufacturer’s instructions.
Obesity evaluation and measurement of serum leptin
BMI was calculated by dividing the individual’s dry body weight (kg) by height squared (m2). Body composition data were obtained using a portable whole-body bioimpedance spectroscopy device (Body Composition Monitor; Fresenius Medical Care, Bad Homburg, Germany). The device provided objective data about fat mass, relative PBF, and fat tissue index (FTI) in which fat mass was normalized to the body surface area (m2). Obesity was arbitrarily defined as a PBF >26.8% for men and >36.2% for women in our cohort (sex-specific median value in our cohort). Also, serum leptin levels were measured using an enzyme-linked immunosorbent assay (ELISA) kit (R&D Systems, Wiesbaden, Germany) according to the manufacturer’s instructions.
Vascular calcification score
The VCS was determined by plain radiographic film of the lateral abdomen in the standing position, as previously described by Verbeke et al. Briefly, abdominal aortic vascular calcification was graded on a 0 to 3 scale at each segment of the first through fourth lumbar vertebrae based on the severity of calcification [
17]. A score of 0 denoted no aortic calcific deposits; 1 denoted small, scattered calcific deposits less than one-third of the longitudinal wall of the aorta; 2 denoted one-third or more, but less than two-thirds; and 3 denoted two-thirds or more. The anterior and posterior aortas were separately graded and the values were summed, resulting in a total score that ranged from 0 to 24. Particularly, VCS ≥ 7 was regarded as having severe VCS. All radiographic films were read by two experienced radiologists without knowledge of subjects’ clinical history.
Statistical analysis
Statistical analysis was performed using the statistical package SPSS, version 24.0 (SPSS Inc., IL, USA). All variables were expressed as the mean ± SD or median with ranges, unless otherwise indicated. The Kolmogorov-Smirnov test was used to analyze the normality of distribution, and natural log values were used for skewed data including hs-CRP and serum leptin levels. Pearson’s correlation analysis was used to clarify the relationship between measured VCS, PBF, and various parameters including 25(OH)D. With multiple regression analysis, the influences of obesity and vitamin D deficiency on VCS were assessed. Multiple logistic regression analysis was performed to evaluate the determinants of significant vascular calcification development. A level of p < 0.05 was considered significant.
Discussion
In this cross-sectional study of 213 ESRD patients, we found that: 1) at the start of dialysis, most patients (76.6%) had 25(OH)D deficiency and more severely decreased 25(OH)D levels were also frequently observed; 2) obesity, defined as elevated PBF, was closely associated with a higher VCS and low serum 25(OH)D level; and 3) serum 25(OH)D affected the relationship between obesity and the risk of vascular calcification, such that vitamin D deficiency was associated with greater risk of a high VCS, especially with obese population, but not with non-obese patients. These findings suggest that the magnitude and direction of the correlation between obesity and the risk of vascular calcification may depend on an individual’s 25(OH)D level, a possible representative marker of cardiometabolic disturbance. In our knowledge, this is the first report to elucidate the potential role of 25(OH)D deficiency for aggravating the association between obesity and cardiovascular adverse outcome in incident ESRD patients.
With increasing numbers of obese dialysis patients, mortality for these patients was expected to be much higher than that of non-obese patient. However, contrary to the expectation, epidemiologic studies have shown the protective benefit of obesity in the ESRD population [
1]. Although the exact reasons of this reverse epidemiology have been unclear as of yet, poor nutritional status or protein-energy wasting are strong predictors of worse outcomes in the CKD population [
18]. Recent studies have found that the benefit of obesity could be conferred only to metabolically healthy obese population [
2,
3]. In fact, with 4,374 CKD patients, Hanks et al. reported that overweight or obese patients who were metabolically healthy had lower risk for mortality as compared with metabolically healthy normal weight individuals [
4]. However, there was no mortality benefit in metabolically unhealthy obese patients compared to metabolically healthy normal weight patients. It is very important to point out that an individual’s metabolic profile could be a potential and very decisive factor for determining adverse cardiovascular outcomes in patients with reduced kidney function.
In the general population, metabolic risks are often assessed in accordance with the criteria for metabolic syndrome. However, as more than half of ESRD patients have diabetes, which is one important criteria of metabolic syndrome, assessment of an individual’s metabolic risks while considering only traditional metabolic components such as BP, glucose, TG, and HDL levels has an important limitation in the ESRD population. Recently, vitamin D has been regarded as a possible representative marker of cardiometabolic disturbance [
7]. Various observational studies have found a close relationship between vitamin D deficiency and numerous adverse extra-skeletal complications, including cardiovascular disease, metabolic syndrome, autoimmune disorders, neurodegenerative disorders, and cancer [
8]. Moreover, vitamin D physiology may differ between subjects with obesity and non-obese individuals [
19,
20]. A recent meta-analysis identified a significant inverse relationship between BMI and vitamin D status in an adult population [
19]. Correction of vitamin D deficiency improved cardiometabolic profiles in adults with obesity [
20]. However, there is limited data for evaluating the role of vitamin D as a metabolic risk factor in ESRD patients. Moreover, whether vitamin D status could affect the link between obesity and vascular calcification, especially in atherosclerosis-prone ESRD population remains unclear.
First of all, our data showed significantly much higher prevalence of 25(OH)D deficiency in our cohort compared to previous data. Wolf et al. reported that 18% of ESRD patients have 25(OH)D deficiency at the start of dialysis [
21]. However, in our data, the prevalence was 76.6%, and the difference of the prevalence of much lower 25(OH)D levels (<3 ng/mL, severe deficiency) was more obvious. Although various hypotheses may be possible, but racial difference may be one of the important factors for the difference. In fact, Wolf also found significantly lower 25(OH)D levels in black subjects compared to white subjects [
21]. Moreover, the prevalence of vitamin D deficiency or insufficiency has been reported to be significantly higher in ethnic minorities than the white population [
19,
22].
But, consistent with previous data, 25(OH)D levels showed a strong relationship with various markers of obesity. Particularly, PBF had the highest correlation with 25(OH)D deficiency. Although BMI also had a negative association with 25(OH)D, the power was much less than PBF. Rather, adipokine levels, such as leptin, and inflammatory parameters, such as hs-CRP, showed a stronger relationship with 25 (OH)D levels than BMI. Interestingly, there was no significant relationship between 25(OH)D levels and other well-known metabolic components such as BP, HDL, TG, and glucose. Our data suggest that serum 25(OH)D could be a potential candidate marker for assessing metabolic risk, independent to the well-established metabolic syndrome components. Although low vitamin D levels in obesity has been reported in several previous data [
15,
19,
23], its causes are not fully elucidated yet. The link between increased PBF and low vitamin D levels may involve chronic low grade inflammation associated with adipose tissue, as our data shows. Supporting this, recent data showed that total body vitamin D store are significantly greater in obese women than normal weight control, suggesting that the enlarged adipose mass in obese individuals serves as a reservoir for vitamin D and that the increased amount of vitamin D required to saturate this depot may predispose obese individuals to inadequate serum 25(OH)D [
24].
As expected, PBF was closely associated with VCS. Jensky et al. reported that a 1 SD increment in abdominal or visceral fat was associated with a 1.6- and 1.5–fold increase, respectively, in risk for the presence of thoracic aortic calcium [
25]. Similarly, with non-dialysis CKD patients, Cordeiro et al. showed that abdominal visceral fat is associated with coronary artery calcification scores [
26]. However, the strength of our study is that we further investigated the detailed association according to the presence of 25(OH)D deficiency. Our data shows that higher PBF might confer significantly higher risk of increased VCS when 25(OH)D deficiency is also found. Particularly, for much lower 25(OH)D levels, higher and more severe vascular calcification was observed. However, interestingly, the modifying role of 25(OH)D levels was only relevant with the obese population. In the non-obese group, the effects of low vitamin D levels were not definite. The mechanisms explaining these observations are still unclear. But one possible explanation is that the potential harmful effect of low vitamin D may be more exaggerated in obese patients, as obese patients usually have higher systemic inflammation and advanced atherosclerosis. In fact, several observational data showed that vitamin D supplementation caused a significant reduction in body fat mass compared with placebo [
27]. Similarly, High doses of vitamin D supplementation in diabetic patients was associated with significant decrease in arterial properties [
28]. Maybe, the anti-inflammatory effect of vitamin D is associated with the changes of body fat mass and distribution. Also, obesity associated increased fibroblast growth factor 23 (FGF-23) and resultant vitamin D deficiency could potentiate the risk of vascular calcification. In fact, FGF23 has been shown to be associated with markers of insulin resistance, dyslipidemia and obesity, suggesting the FGF23 could be a cardiometabolic risk factor connecting obesity and higher cardiovascular complications [
29]. Moreover, serum FGF23 levels are independently associated with higher levels of inflammatory markers in patients with CKD [
30], which may play a major role in progression of vascular calcification. In fact, Mirza et al. reported that FGF 23 levels were higher in subjects with the increased adiposity compared with those without, and this potentially represent a novel pathway linking high FGF 23 to an increased cardiovascular risk [
31]. Unfortunately, in this study, we could not measure other cardiometabolic risk factors except vitamin D levels.
There are some limitations in our study. First of all, we could not consider the seasonal variations of vitamin D levels. In general, 25(OH)D levels become lower in winter seasons due to insufficient sun exposure [
32]. However, in this study, as most patients had vitamin D insufficiency or deficiency, the seasonal effect was thought to be minor. Also, a large number of patients (34.2%,
n = 73) had very low 25(OH)D levels (<3 ng/mL, defined as severe vitamin D deficiency), and such low levels were not common in other disease populations. So the results may not extrapolate to other groups. Second, the causal relationship between low vitamin D and obesity is not clear in this study. Although most previous data suggest obesity causes vitamin D deficiency or insufficiency, the reverse relationship is also possible. However, as several previous studies showed that vitamin D supplementation could significantly reduce body fat mass and increase serum vitamin D concentration, in this study, we hypothesized that obesity could induce vitamin D deficiency or insufficiency. Future studies that evaluate the beneficial role of vitamin D supplementation in preventing development of vascular calcification or other cardiovascular events, especially in obese patients, may be needed. Third, we only had a single measurement of 25(OH)D, limiting our ability to establish the impact of vitamin D level changes during the study period. Last, even though we suggested the relationship between increased body fat mass and vascular calcification, we could not evaluate whole atherosclerotic profiles as well as early vascular complications such as endothelial dysfunction (actually, vascular calcification is a rather later step in atherosclerosis). Similarly, the well-established other biomarkers dealing with CKD-MBD, such as FGF23 or serum alkaline phosphatase levels were not measured in present study.