Low vitamin D level was associated with metabolic syndrome and high leptin level in subjects with nonalcoholic fatty liver disease: a community-based study

Nonalcoholic fatty liver disease (NAFLD) is defined as fat accumulation in the liver greater than 5% by weight in the absence of excessive alcohol consumption [1, 2]. NAFLD is reported to be associated with other systemic diseases, such as cardiovascular diseases, insulin resistance (IR), obesity, dyslipidemia and metabolic syndrome (MS) [1‐4]. The prevalence of NAFLD is increasing and has been estimated to be between 10 and 30% worldwide [5, 6]. Vitamin D deficiency is also a worldwide condition and is present in approximately 30–60% of the general adult population [7, 8]. Vitamin D levels have been reported to be inversely related with fasting glucose concentrations [9, 10], adiposity [11, 12], lipid [13], and blood pressure [14‐16]. Fasting blood sugar, adiposity (central obesity), lipid and blood pressure are the component of MS criteria. Previous studies have revealed that vitamin D deficiency can significantly increase the risks of IR and MS [8, 17, 18]. Vitamin D may influence hepatocytes and non-parenchymal hepatic cells (hepatic stellate cells, Kupffer cells) in NAFLD via metabolic, anti-inflammatory and anti-fibrotic effects [19, 20]. Several inflammatory cytokines, such as interleukin (IL)-6, tumor necrosis factor alpha (TNF-α) and IL-1β, and adipokines, such as adiponectin and leptin, may participate in these reactions [19, 20]. Because NAFLD and vitamin D deficiency engender similar risks for cardiovascular disease, insulin resistance and MS in epidemiologic studies, there have been many reports debating a potential association between NAFLD and vitamin D deficiency [19‐23]. However, most studies have lacked information on serum cytokines or adipokines and had small sample sizes. The hypothesis in this study was that vitamin D influences MS and NAFLD via inflammatory cytokines [such as TNF-α and highly sensitive C reactive protein (HS-CRP)] and adipokines (such as adiponectin and leptin). The primary aim in this study was to analyze the potential association between vitamin D level, MS and NAFLD from community-based cohort data. The secondary aim was to reflect that the association between serum vitamin D levels and levels of adiponectin, leptin, HS-CRP and TNF-α.

From August 2013 to August 2016, a community-based survey for MS was performed in the Wanli, Ruifang and Anle districts in the north-eastern region of Taiwan. The Institutional Review Board of the Chang-Gung Memorial Hospital approved this research (IRB No: 103-3886C). All participants agreed to the study conditions and provided written informed consent before enrollment in this study. Informed written consent was obtained from all subjects.

Because the majority of MS develops in people with middle age [24], the inclusion criteria were: subjects aged > 30 years who underwent abdominal ultrasonography (US) and completion of blood testing. Exclusion criteria were:

All subjects participated in a demographic survey, physical examination and blood testing. Blood tests included complete blood cell counts, liver and renal biochemistry tests, fasting glucose and insulin levels, vitamin D, adiponectin, leptin, TNF-α and HS-CRP levels and lipid profiles. The demographic survey assessed the past history of systemic diseases, such as hypertension, diabetes mellitus, hyperlipidemia, hematologic disorders, autoimmune diseases, medication history, and family history. The physical examination evaluated heart rates, blood pressures, body weights, body heights, and waist girths (circumferences). Body mass index (BMI) (kg/m²) was calculated as weight (kg) divided by squared height (m). Subjects were asked to fast overnight before giving blood samples. Blood samples were analyzed within 4 h after collection to determine complete blood cell counts, biochemical parameters, and antibody titers. The serum samples were stored in tubes at − 80 °C following centrifugation (3000 rpm at 4 °C for 30 min). The assays for adiponectin and leptin were performed using stored serum samples.

Initially, 1922 participants were enrolled. Participants who were viral hepatitis B- or C-positive (199 subjects), frequently drank alcohol (men > 30 g and women > 20 g alcohol consumption per day, 109 subjects), took vitamin D supplement (28 subjects) or steroid treatment (3 subjects), and declined to take abdominal US study (32 subjects) were excluded. According to the abdominal US findings, fatty liver change was found in 835 subjects (544 women) and the other 716 (537 women) subjects with normal US study. After matching age and gender, finally 564 subjects with fatty liver but no viral or alcoholic liver diseases were included in NAFLD group. The other 564 subjects with normal US finding, and normal serum AST, ALT and gamma-glutamyl transpeptidase (GGT) values were included as controls in this study (Fig. 1). The mean age was 57.1 ± 11.9 years in the NAFLD group and 57.5 ± 11.9 years in the control group (P = 0.618). The demographic and characteristic data are listed in Table 1. The prevalence of MS was higher in the NAFLD group than in the control group (202/564, 35.8% vs. 87/564, 15.4%, P < 0.001). Subjects in the NAFLD group also had higher mean values of weight, waist circumferences, systolic/diastolic blood pressure, triglycerides, sugar, AST, ALT, and leptin, but lower mean values of HDL and adiponectin, than the mean values of subjects in the control group. The percentage of subjects exhibiting IR according to the HOMA-IR index was greater in the NAFLD group than in the control group, no matter used the cutoff point of HOMA-IR at 2, 2.5 or 3. The mean value of vitamin D was lower in the subjects with NAFLD (28.5 ± 9.8 ng/ml) than the mean value in control group (29.9 ± 10.2 ng/ml) (P = 0.018).

Table 2 shows the correlation between vitamin D level and other factors, such as age, gender, BMI, adiponectin, leptin, TNF-αand HS-CRP. It reveals that vitamin D level was positively correlated with age and male, but negatively correlated with leptin.

Table 3 shows the logistical regression analysis for the association between vitamin D level and MS, after adjusting for the confounding factors of age, gender, BMI and NAFLD status. Subjects with deficient or insufficient vitamin D levels (< 20 or 20–30 ng/ml) had a higher odds for MS when compared with those with sufficient vitamin D levels (> 30 ng/ml) [deficiency vs. sufficiency, adjusted OR = 1.770 (95% CI = 1.145–2.736), P = 0.010; insufficiency vs. sufficiency, adjusted OR=, 1.673 (95% CI = 1.220–2.295), P = 0.001). When using the quartile method of vitamin D, the result of associations between vitamin D level and MS was similar with the method by deficiency, insufficiency and sufficiency classification. Subjects with vitamin D value located in Q1 (< 22.41 ng/ml) had a higher risk of MS (OR = 2.473, 95% CI = 1.580–3.871, P < 0.001) than those with vitamin D value located in Q4 (> 35.38 ng/ml). Similarly, subjects with vitamin D value located in the Q2 (22.41–28.41 ng/ml) or Q3 (28.41–35.38 ng/ml) also had a higher risk of MS than those with vitamin D value located in Q4. When a subgroup analysis was performed for subjects with NAFLD, subjects with low vitamin D level still had a trend of increasing risk of MS than those with high vitamin D level. Likewise, subjects with vitamin D value located in Q1 had a higher risk of MS (OR = 2.358, 95% CI = 1.386–4.013, P < 0.001) than those with vitamin D value located in Q4.

In view of the association between vitamin D level and NAFLD status, vitamin D insufficiency or deficiency did not increase the risk of having NAFLD when compared with sufficient vitamin D levels [insufficiency vs. sufficiency: adjusted OR = 1.096 (95% CI = 0.839–1.432, P = 0.501); deficiency vs. sufficiency: adjusted OR = 1.238 (95% CI = 0.804–1.775, P = 0.245)].

Table 4 shows the distribution between vitamin D levels and the severity of fatty liver disease by abdominal US. There was no significant difference in the percentage of vitamin D level among the mild, moderate and severe fatty liver by abdominal US (P > 0.05 by Chi-squared test).

When a NAFLD fibrosis score was applied for liver fibrosis evaluation, most subjects with NAFLD were not found to have an advanced fibrosis status. Regarding the NAFLD fibrosis score, 77.0% of subjects had a NAFLD fibrosis score < − 1.455, indicating a low degree of fibrosis (F0–2). Only 6 subjects with a NAFLD fibrosis score > 0.675 (high degree of fibrosis, F3–4) were detected in this study. The mean value of vitamin D in these 6 subjects was 39.0 ± 10.9 ng/ml (range: 20.1 to 52.4 ng/ml). Because of the small number of subjects with advanced liver fibrosis in this study, we could not draw any firm conclusions regarding an association between vitamin D level and fibrosis in subjects with NAFLD.

Recent studies have revealed that vitamin D has a broad effect on immune modulation, cell differentiation, proliferation and inflammation regulation. Possible mechanisms to explain the association between vitamin D level and NAFLD include vitamin D-mediated improvement in insulin secretion and insulin resistance, decreased adipose tissue inflammation, and decreased hepatic inflammation and fibrosis via the regulation of the vitamin D receptor and several cytokines, such as interleukin 6, TNF-α or adiponectin [19, 35, 36]. The hypothesis of the current study was that vitamin D influences the status of MS via inflammatory cytokines (TNF-α and HS-CRP) or adipokines (adiponectin and leptin). In bivariate correlation study, vitamin D level was not associated with TNF-α, HS-CRP or adiponectin levels. But vitamin D level was negatively correlated with leptin. Previous studies reported vitamin D-mediated inhibition of leptin [37]. Another study reported increased leptin level was linked to decreased HDL-C level [38]. In current study, subjects with vitamin D deficiency and NAFLD had increased leptin value but lower HDL-C level. It could partially explain the association between low vitamin D level and MS in subjects with NAFLD. A further study of the potential associations among vitamin D, leptin, MS and NAFLD is warranted [19‐21, 38, 39].

Manson’s et al. reported an over-estimation of vitamin D deficiency and unnecessary vitamin D supplement by applying the cut point of < 20 ng/mL as a deficiency status for general people [40]. They suggested using estimated average requirement (EAR) for the general population. Hence, the definition of vitamin D deficiency might use a different cut point of serum vitamin D value for people in the different areas [40, 41]. Because the real EAR value is unavailable in our area, we used both traditional classification (deficiency, insufficiency or adequate) and quartile divide (Q1–4) methods to evaluate vitamin D status. This analysis revealed a consistent result of MS risk by these two methods.

Participants in NAFLD group have a higher mean BMI and waist but lower vitamin D than those in control group. A decreased release of stored vitamin D from adipose tissue into the circulation and decreased cutaneous synthesized vitamin D3 into circulation induce lower serum vitamin D in obese subjects [14, 42]. Rhee et al. have reported that serum vitamin D levels were lower in the group with NAFLD compared to the control group [43]. Targher et al. also reported a lower serum vitamin D level in those with biopsy diagnosed NAFLD than normal controls (20.4 ± 8.8 vs. 29.8 ± 6 ng/mL, P < 0.001) [44]. Therefore, several published studies have found evidence of a lower serum vitamin D level in subjects with NAFLD compared to those without NAFLD [45]. However, there have been several studies reporting conflicting results concerning vitamin D level and NAFLD status. One Chinese study by Li L et al. [46] and another Korean study by Hong et al. [47] reported no significant differences in vitamin D levels between groups with or without NAFLD. The reasons for these conflicting results regarding a potential association between vitamin D level and NAFLD status may include the different methods and criteria for NAFLD and vitamin D deficiency diagnosis, selection bias in cross-sectional studies and no adjustment for confounding factors, such as BMI. Moreover, polymorphisms of the vitamin D receptor gene and the interaction between downstream products and vitamin D concentration may be associated with the progression and severity of liver diseases [48, 49]. Hence, vitamin D may influence the development or the progression of NAFLD only in certain individuals with specific genotypes.

The current study has similar limitations as previous cross-sectional studies. First, a selection bias may exist in this community-based screening study. Most subjects participating in this study were women and of older age. Second, the definition of hepatic steatosis severity and NAFLD fibrosis scores were made by abdominal US and calculated formulae, respectively, without histological confirmation. The current study was based on a community cohort, and liver biopsy was not feasible or practical. During advanced fibrosis or cirrhosis, steatosis and necroinflammatory reactions may disappear in patients with NASH, and this status is known as “burn-out NASH” [50]. Patients with NAFLD but advanced liver fibrosis may have no sign of fatty liver in the abdominal US because of burning out NASH. In our study, advanced fibrosis (score > 0.674) by NAFLD fibrosis score could be detected in both NAFLD group and normal group. We could not identify whether the subjects with advanced fibrosis in normal group are burn-out NASH or not. Moreover, the prevalence of advanced fibrosis was low in our NAFLD subjects, which was similar with other Asia studies [51, 52]. Because of low number in patients with NAFLD and advanced liver fibrosis, we could not draw any conclusions regarding an association between vitamin D levels and fibrosis in subjects with NAFLD. Third, the test of serum vitamin D was checked all the year without avoid the winter time. Current study was conducted in Keelung City (latitude: 25∘N), where sunshine was adequate all the year. Keelung is unlike some areas in high latitude (above the 37∘N latitude), where skin vitamin D production decreases because of reduced ultraviolet B (UVB) exposure during winter (November to February) [53]. In conclusion, subjects with low vitamin D levels have an increased odds of metabolic syndrome compared to those with sufficient or high vitamin D levels. The distribution of fatty liver severity (as measured by abdominal US) was similar in subjects with NAFLD among the different vitamin D levels.