Background
The prevalence of type 2 diabetes (T2D) is increasing worldwide, including Iran [
1]. It has been shown that diabetes is accompanied by remarkably greater risk for cardiovascular disease (CVD). Accounting for more than 80% of all premature deaths, CVD has been known as the major cause of mortality in T2D [
2].
Diabetes may affect both small and large vessels, leading to micro- and macro-angiopathy, respectively. Hyperinsulinemia and augmented oxidative stress, usually both present in diabetes, are known as two major contributing factors of the long-term complications, including micro- and macro-angiopathy [
3]. It is hypothesized that endothelial dysfunction is the underlying cause of diabetic angiopathy that eventually leads to CVD [
4]. However, a population-based study showed that raised plasma concentrations of endothelial biomarkers may predict diabetes independent of such diabetes risk factors as obesity, insulin resistance and systemic inflammation [
5]. The endothelial function may therefore be the focus of preventive efforts against both diabetes and its fatal complications.
The relationship between vitamin D and T2D has recently been the focus of interest. Vitamin D, mostly known for its calcemic functions, has been shown to have many non-calcemic actions including regulation of gene expression and antioxidant properties [
6]. Vitamin D deficiency has been proposed as an independent risk factor for CVD [
7]. It has been recently reported that the odds ratio for having CVD outcomes in individuals with serum 25(OH)D below 25 nmol/L, compared to those with 25(OH)D ≥37.5 nmol/L, after adjustment for potential confounders was 2.90 (95% confidence interval: 1.67 to 5.12,
P < 0.001) [
8]. Considering the role of endothelial dysfunction in development of CVD, the issue has been raised if vitamin D can influence the endothelia. Vitamin D deficiency has been associated with endothelial dysfunction and lipid peroxidation in non-diabetic adults [
9]. We have recently shown in another study on a separate population that vitamin D intake in subjects with T2D improves glycemic control [
10]. In the current study, it was hypothesized that the effect of improvement of vitamin D status via daily intake of a vitamin D3-fortified Persian yogurt drink (doogh) can affect endothelial biomarkers independent of glycemic status.
Discussion
The high occurrence of poor vitamin D status in our subjects is comparable with other reports from Iran [
10,
17]. While the occurrence of vitamin D deficiency and insufficiency did not change after a 12-week intervention period in the PYD group, a significant decrease of poor vitamin D status in the FYD group indicated both the high efficiency of the dosage used and the high bioavailability of vitamin D3 in the Persian yogurt drink, doogh. All participants completed the intervention period and the compliance of the patients was 100%, as judged by checking the consumption tables, follow-up calls, the subjects' reports and, in the FYD group, the expected rise in serum 25(OH)D. However, the persistence of vitamin D deficiency in the FYD group (4.5%) despite a daily intake of 1000 IU vitamin D for 12 weeks can be a subject of considerable argument. Interestingly, those subjects in the FYD group whose vitamin D status remained undesirable after the intervention did not show any appreciable changes in their glycemic and endothelial biomarkers. A similar observation has already been reported [
18]. Considering the high prevalence of vitamin D deficiency in Iranian subjects either with or without diabetes [
17,
19] and assuming the same prevalence of "non-responders" among the whole population, a huge number of people may not benefit from intake of the usually recommended amount of vitamin D. This individual variability may be, at least in part, explained by vitamin D receptor (VDR) polymorphisms. A significant association between VDR variants and susceptibility to several diseases including cancer [
20,
21], ulcerative colitis [
22], metabolic syndrome [
23] and both types of diabetes [
24,
25] has already been reported. The upcoming results from our study group will soon address the role of VDR genotypes on the variation of outcomes after improvement of vitamin D status in subjects with T2D [
12].
Significant reductions in BMI, WC and FM following 12 weeks of daily intake of 1000 IU vitamin D via a fortified-doogh, are consistent with our previous report [
10]. An inverse association between vitamin D status and adiposity shown in older people [
26] was further confirmed in both children [
27,
28] and adults [
29]. From a mechanistic point of view, intracellular calcitriol has been implicated in both enhancing and inhibiting adipogenesis in cell culture [
30] probably through peroxisome proliferator-activated receptor-gamma (PPAR-γ) and retinoid × receptor (RXR)-involved pathways [
31].
However, some studies have failed to show any additional effect of vitamin D and calcium supplementation during weight loss on body adiposity [
32] and in obese children, poor vitamin D status has been suggested as the consequence, rather than the cause, of adiposity [
33]. Notwithstanding, in that study obese subjects consumed only 0.25 μg (10 IU)/day of vitamin D [
32], a dosage far less than the amount considered sufficient to optimize circulating 25(OH)D [
34]. Moreover, weight reduction may mask the possible influence of vitamin D on adipogenesis. To overcome these problems, both an efficient dosage of vitamin D and a weight maintenance intervention were used in the present study.
Poor vitamin D status has been recognized as an independent risk factor for incident arterial hypertension and it is believed that vitamin D supplementation can reduce SBP by 2 to 6 mmHg [
35]. However, in this study despite the high occurrence of poor vitamin D status among the participants, no significant effect of regular vitamin D intake on blood pressure (either SBP or DBP) was observed. It should be noted that most of our subjects were normotensive as the blood pressure of only eight (15.4%) and nine (17.6%) subjects in the FYD and PYD groups, respectively, was above 135/85 mmHg. Even re-analysis in the subgroups with above normal BP did not show any significant change. The possible effect of vitamin D on BP merits further research.
Amelioration of the glycemic status of our subjects after a 12-week intervention period further confirms our previous observations [
10]. Attenuation, but still persistence, of significant differences after controlling for WC and FM might imply both direct (improvement of insulin secretion and insulin sensitivity) and indirect (decrement of FM and weight) effects of cholecalciferol on glycemic status.
Contrary to a previous report [
10], a significant improvement in lipid profile was observed in the FYD group. However, the disappearance of significant between-group changes after controlling for QUICKI implies that the effect of vitamin D3 on the lipid profile may be secondary to its ameliorating effect on glycemic status. Unlike our finding, in a 2-year RCT involving 167 men aged 50 years and older, daily intake of 400 mL low-fat milk fortified with 1,000 mg calcium and 800 IU vitamin D3 did not affect blood lipids [
36]. In contrast, daily intake of 1,200 mg calcium and only 400 IU vitamin D in overweight/obese women during a 15-week weight reduction intervention resulted in lower serum LDL-C, compared to those who just had a weight reducing diet [
37]. Some of these discrepancies may relate to the baseline blood lipids and 25(OH)D of the subjects. It is noteworthy that our subjects in the present study, compared to the previous one, had lower lipidemic status (especially higher for both TG and TC) so that they might be more responsive to a nutritional intervention. The effect of vitamin D on blood lipids and lipoproteins needs to be elucidated by further precisely controlled clinical trials.
Elevated circulating concentrations of endothelin and E-selectin have been attributed to the boosted production of vasoconstrictors and increased adhesion of and permeability to leukocytes, respectively [
38]. On the other hand, a high glucose milieu has been shown to induce over-expression of several matrix metalloproteinases, notably MMP-9, thus promoting matrix degradation, atherogenesis and plaque instability [
39]. It has been recently demonstrated that down-regulation of MMP-9 via insulin treatment will result in reduction of intimal lesions in atherosclerosis-prone diabetic apoE (-/-) mice [
40]. In the current study, endothelial status, as evaluated by serum concentrations of E-selectin, endothelin-1 and a vascular inflammation marker, MMP-9, were improved after daily intake of 1,000 IU vitamin D via fortified-doogh consumption. This was not, however, accompanied by a significant change in urinary excretion of albumin. Soluble E-selectin together with a coagulation factor, XIIa, has been associated with ten-year macrovascular events in subjects with T2D [
41]. Poor vitamin D status-related endothelial dysfunction in diabetes may be partly mediated by nuclear factor κ-B (NF-κB)-linked inflammation [
42] and endothelial progenitor cell depletion [
43]. Persistence of a significant between-group difference of serum endothelin-1 and MMP-9, but not E-selectin, after controlling for QUICKI, FM and WC implies a direct effect of vitamin D3 on the first two endothelial biomarkers but an indirect effect on E-selectin. It has recently been reported that dyslipidemia and high BMI are related to macroangiopathy while the duration of disease and high HbA1c are associated with microangiopathy [
44]. Vitamin D replenishment may, therefore, be considered as an effective preventive measure against development of both macro- and micro-angiopathy by improving lipid profile, BMI and HbA1c. This effect may be comparable with that of the combination of metformin and atrovastatin [
45].
Several observational studies have suggested an association between vitamin D deficiency and vascular disease but only a handful of them have examined the effect of vitamin D on endothelial function. In a small pilot study on 34 T2D subjects with vitamin D insufficiency (25OHD < 50 nmol/L), a single, mega dose of vitamin D2 (100,000 IU) resulted in a significant improvement of endothelial function [
46]. Monthly intramuscular injection of 300,000 IU vitamin D to subjects with subclinical vitamin D deficiency for 3 months had a favorable effect on endothelial function, as evaluated by brachial artery flow-mediated dilatation (FMD) [
9]. In a cross-sectional study, circulating 25(OH)D inversely correlated with MMP-9 and vitamin D status was found to be the only predictor of MMP-9. Following 1-year supplementation of just vitamin D deficient subjects (n = 41), a significant reduction in MMP-9 (-68%) was observed [
47]. Our finding of the suppressive effect of vitamin D on serum MMP-9 in subjects with T2D further supports previous reports. Notwithstanding, all these trials have been conducted on rather small sample sizes and in widely disparate settings with different adducts and doses of vitamin D and outcome measures resulting in poor comparability.
In the current study, a significant decrease in serum concentrations of endothelin-1, E-selectin and MMP-9 even after controlling for QUICKI implies the direct effect of vitamin D independent of glycemic status. In accord with this notion, a recent study demonstrated that decreased vascular endothelial cell expression of VDR and 1-α-hydroxylase due to vitamin D deficiency is associated with vascular endothelial dysfunction [
48].
However, the limitations of this study must be taken into consideration. Since the study was conducted during the cold seasons, the vitamin D status of our subjects does not reflect its status during the entire year. Moreover, the changes observed after a 12-week intervention cannot necessarily be extrapolated to the long-term effects.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
TRN designed and supervised the study, was involved in laboratory analyses and wrote the finalized manuscript. AD and SSB both helped intellectually in finalizing the study design. SSB performed most of the laboratory analyses, wrote the preliminary manuscript and was actively involved in the field work. This study was part of her Ph.D. thesis at TUMS under the supervision of both AD and TRN (grant No: 10533). All statistical analyses were done under the supervision of MRE. AG, together with MZ and SSB, performed HPLC analyses. AK and NS participated in all laboratory investigations. Finally, dietary assessments were performed by HH under AH guidance. All laboratory analyses were performed at the Laboratory of Nutrition Research, NNFTRI. All authors read and approved the final manuscript.