Vitamin D deficiency is a major public health problem worldwide and is often reported in people with liver diseases. According to literature, the prevalence of vitamin D deficiency among patients with chronic hepatitis C (CHC) infection is >60 % and can further vary based upon the geographical region(Reference Jin, Chen and Sheng1–Reference Melo-Villar, Lampe and de Almeida3).
Nonetheless, despite its continuous improvements, routine assessment of vitamin D status still faces several challenges that are often neglected in clinical practice. One of these challenges deals with the lack of optimal and clinically relevant 25-hydroxyvitamin D [25(OH)D] concentration cut-off value to define the fall in serum 25(OH)D concentration in patients with CHC(Reference Jin, Chen and Sheng1). Regarding this, except for the Endocrine Society guideline, most recent consensus statements about hypovitaminosis D do not specifically consider diseased populations. Therefore, different cut-off values used in different studies might affect the true prevalence of hypovitaminosis D in patients with CHC.
Another challenge is the influence of various factors on the serum concentration of 25(OH)D, such as age, skin pigmentation, clothing style, sunscreen use and seasonality(Reference Bonelli, Buonocore and Aloe4,Reference Mendes, Hart and Botelho5) . Regarding this last aspect, others have already discussed the seasonal variation of serum 25(OH)D concentration during the months with the shortest daylight period. And, as pointed out previously, to achieve a correct diagnosis of vitamin D deficiency, blood sampling seasonality should be regarded as an important preanalytical factor(Reference Bonelli, Buonocore and Aloe4).
However, despite having consistent evidence about the influence of these factors on serum 25(OH)D concentration, to our knowledge, there is a lack of studies on the relevance of these factors on serum 25(OH)D concentration among patients with CHC.
Therefore, the aim of this cross-sectional study was to evaluate the prevalence of vitamin D deficiency in subjects with CHC infection. The secondary aim was to assess whether the season when the blood sample was drawn is associated with the presence of vitamin D deficiency.
Methods
Study population
Medical records of patients attending the – blinded for peer review – were retrospectively reviewed.
Patients were retrospectively enrolled if they met the following criteria: ≥18 years of age with CHC infection and having a 25(OH)D blood test collected at first consultation. Key exclusion criteria included chronic liver disease unrelated to hepatitis C vírus (HCV) infection, co-infection with HIV or hepatitis B virus and history of past and/or current use of vitamin D supplementation.
Study design
A cross-sectional study was conducted on 306 subjects with CHC infection. The following data were gathered from those meeting the inclusion criteria: age, gender, skin colour, presence of any comorbidities, previous treatment experience, severity of liver fibrosis and seasonality of blood sampling. Seasonality was defined according to solstices and equinoxes, that is, winter: June–September; spring: September–December; summer: December–March; and autumn: March–June. Biochemical parameters, including plasma glucose, 25(OH)D, urea, creatinine, aspartate aminotransferase (AST), alanine aminotransferase (ALT), HCV genotype and HCV RNA viral load, were also assessed by commercial laboratory methods. Cases with missing values were deleted listwise, that is, observations with missing values on any of the variables were omitted from the analysis.
Glomerular filtration rate was estimated (eGFR) for each patient using the four-variable Modifications of Diet in Renal Disease (MDRD) study equation(Reference Levey, Bosch and Lewis6). The presence of cirrhosis was defined as a Metavir score of F4 on liver biopsy up to 12 months after first consultation and/or a liver stiffness value >12·5 kPa and/or an AST-to-platelet ratio index (APRI) value >2(Reference Chou and Wasson7,Reference Lin, Xin and Dong8) . In case of conflicting fibrosis results according to these methods, biopsy data took precedence. If biopsy data were not available, the FibroScan result was considered. Otherwise, the APRI result was considered.
Biochemical determinations
In the morning after an overnight fast, venous blood was sampled for measurements of plasma glucose, 25(OH)D, urea, creatinine, AST and ALT following routine biochemical laboratory protocols. Liver transaminases were described as healthy considering 19 and 30 U/l as its upper limit of normality for women and men, respectively(Reference Prati, Taioli and Zanella9).
HCV genotype or subtype was determined using a Real-Time HCV Genotype II assay (Abbott Molecular). Plasma HCV RNA levels were assessed using the HCV COBAS TaqMan Test (version 2.0; Roche Molecular Systems) with a lower limit of quantitation of 12 IU/ml.
All serum 25(OH)D analyses were performed at the same laboratory using Elecsys Vitamin D Total assay (Roche Diagnostics), which is a fully automated electrochemiluminescence protein-binding assay for the quantitation of 25(OH)D3 and 25(OH)D2, allowing quantitative determination of total 25(OH)D. It is standardised against liquid chromatography coupled with tandem MS (LC-MS/MS) calibrated using the National Institute of Standards and Technology reference material SRM 2972(Reference Vogeser10,Reference Phinney11) . The assay has a detection limit of 7·50 nmol/l and a within-run imprecision (CV) of 18·5 %. Notably, a recent study showed a higher performance for the assay in routine laboratory testing: within-run CV ≤7 %, within-laboratory CV <9·5 %, between-laboratory precision CV ≤10·1 %, and functional sensitivity <9·8 nmol/l (at CV 12·9 %)(Reference Wielders, Carter and Eberl12). 25(OH)D concentrations <50 nmol/l were classified as deficiency; those ranging from 250 to 74·9 nmol/l as insufficiency, and between 75 and 250 nmol/l as sufficiency(Reference Holick, Binkley and Bischoff-Ferrari13). As vitamin D inadequacy is particularly important among patients with a chronic illness, the Endocrine Society have recommended attaining a 25(OH)D concentration >75 nmol/l for higher-risk populations, such as CHC-infected patients.(Reference Holick, Binkley and Bischoff-Ferrari13) Therefore, we also categorised 25(OH)D concentration based on this threshold and defined vitamin D adequacy as a 25(OH)D concentration of ≥75 nmol/l.
Statistical methods
Based on previous findings on the overall prevalence of vitamin D sufficiency in the Brazilian population, we assumed that the prevalence of vitamin D inadequacy for a high-risk population would be around 65 % (60–70 %)(Reference Maeda, Borba and Camargo14–Reference Kuchuk, van Schoor and Pluijm16). Therefore, we calculated a required sample size of 243 patients with an allowable error margin of 6 % and a 95 % CI.
Continuous data are presented as mean (sd) or median (interquartile range, IQR) according to the Shapiro–Wilk test of normality. Categorical variables are presented as counts and percentages. Comparisons between groups were performed using Wilcoxon–Mann–Whitney test for continuous variables, and Pearson’s Chi-squared test with Yates’ continuity correction for categorical variables. Comparison of differences in the distribution of values among seasons was conducted using Kruskal–Wallis test, followed by a post hoc Dunn test for multiple comparisons. Univariate logistic regression analysis was performed to explore the unadjusted association between variables and vitamin D deficiency. The variables whose univariate analysis obtained a P-value <0·25 and other variables of known clinical relevance were selected for the multivariate analysis. Statistical significance was assessed at a two-sided P < 0·05. All analyses were conducted using R 3.5.2 (The R Project for Statistical Computing, 2018) in R-studio 1.1.463 (RStudio Inc.).
Results
The general characteristics of studied population are depicted in Table 1. A total of 306 subjects with CHC were included in the study. The majority of subjects were males (162/306; 53 %) with a mean age of 50 ± 11 years. The prevalence of comorbidities was 56 % (169/306); most of the subjects had their first consultation during winter (126/306; 41 %) and had no previous HCV treatment experience (272/306; 89 %). HCV genotype 1 was the most frequent (246/306; 80 %), and mean baseline HCV RNA viral load was 5·9 ± 0·7 log10 UI/ml. The prevalence of vitamin D deficiency and insufficiency was 16 % (50/306) and 23 % (71/306), respectively. The median serum concentration of 25(OH)D was 39 (IQR 17), 62 (IQR 11) and 112 (IQR, 47) nmol/l in patients with vitamin D deficiency, insufficiency and sufficiency, respectively. Subjects with vitamin D deficiency presented a higher prevalence of unhealthy ALT values, despite having no statistically significant difference in the mean ALT value (P = 0·5). A trend was observed towards having a lower HCV RNA viral load for vitamin D-deficient subjects (6 ± 0·7 v. 5·8 ± 0·7 log10 UI/ml, P = 0·06).
Characteristics of the study population
Data are presented as n and %, mean and sd, or median and IQR unless otherwise indicated.
Seasonality of blood sampling and serum 25(OH)D concentration
Overall, subjects who have had their blood drawn during summer and autumn showed higher serum 25(OH)D concentrations (P = 0·01, with Kruskal–Wallis test). Post hoc multiple comparisons found a statistically significant difference between spring–summer (P = 0·04), autumn–winter (P = 0·01) and summer–winter (P = 0·009) (Fig. 1). The prevalence of vitamin D deficiency was also shown to be statistically higher in spring and winter (P = 0·007), without statistically significant differences between them.
Mean serum concentration of 25(OH)D by gender and season. 
, males; 
, females
Seasonality of blood sampling as an independent risk factor for vitamin D inadequacy
The relationship between vitamin D inadequacy (i.e. serum 25(OH)D concentration <75 nmol/l) and seasonality of blood sampling was further analysed using a logistic regression model adjusted for the variables whose univariate analysis obtained a P-value <0·25, and other variables of known clinical relevance were selected for the multivariate analysis. In the first model, the presence of vitamin D inadequacy was independently associated with male gender (adjusted OR (aOR) = 0·4, 95 % CI 0·2, 0·7), serum albumin concentrations (aOR = 0·4, 95 % CI 0·2, 0·9) and only with samples being collected in winter (aOR = 2·9, 95 % CI 1, 9·6) and spring (aOR = 2·8, 95 % CI 0·9, 10) (Table 2).
Univariate and multivariate logistic regression for detecting the predictors of serum 25(OH)D concentration <75 nmol/l
In the second model, we sought to examine the predictors of vitamin D deficiency. As presented in Table 3, vitamin D deficiency was independently associated with moderate liver fibrosis (aOR = 0·1, 95 % CI 0·01, 0·8) and only with samples being collected in spring (aOR = 6·2, 95 % CI 1, 120).
Univariate and multivariate logistic regression for detecting the predictors of serum 25(OH)D concentration <50 nmol/l
Discussion
In this study of subjects with CHC infection living in São Paulo, Brazil, almost 40 % had vitamin D inadequacy (i.e. serum 25(OH)D concentration <75 nmol/l). While others have reported similar findings in non-infected subjects, the literature shows a much higher prevalence in HCV-infected subjects(Reference Melo-Villar, Lampe and de Almeida3). In fact, we had expected a higher frequency of vitamin D inadequacy as most of the HCV-infected subjects have cumulative risk factors for hypovitaminosis D.
However, there are several possible explanations for our findings. First, sunlight is an essential determinant of vitamin D status and stimulates the production of vitamin D3 in the skin depending on age, skin pigmentation, clothing style and sunscreen use(Reference Mendes, Hart and Botelho5). Although the last ones were not addressed in this study, most of our subjects were middle-aged men and had light skin colour, which could explain this difference in prevalence.
Yet, as the effect of sunlight on vitamin D metabolism is dependent upon exposure to adequate sunlight wavelength, geographical location and season are considered important determinants of vitamin D status(Reference Bonelli, Buonocore and Aloe4,Reference Mendes, Hart and Botelho5) . Regarding these aspects, in Brazil, seasonal variability of UV-B radiation reaching the earth’s surface decreases proportionally with latitude. Therefore, as this study’s area is located in a low-latitude region (S23°32′), we expected that seasonal variation of UV-B radiation would influence serum concentrations of 25(OH)D. Accordingly, we found that subjects who have had their blood drawn during summer and autumn showed a higher serum 25(OH)D concentration and, thus, a lower prevalence of vitamin D inadequacy during those seasons. To the best our knowledge, so far, no study has been performed in subjects with CHC infection to assess whether the season when the blood sample was drawn is associated with the presence of vitamin D inadequacy.
Nevertheless, recent emerging evidence on the prevalence of vitamin D inadequacy in sunny countries suggests that, despite the abundance of sunlight, this condition might be a public health issue(Reference Mendes, Hart and Botelho5,Reference Maeda, Borba and Camargo14) . However, although they considered the geographical location, none have taken into account the season in which the blood was collected. In this study, we showed that blood sampling seasonality is an important preanalytical factor for the assessment of vitamin D inadequacy in subjects with CHC infection.
The overall prevalence of vitamin D inadequacy was rather similar to that reported in non-infected postmenopausal women (60–85 years) living in the same geographical location(Reference Arantes, Kulak and Fernandes15). Thus, considering the characteristics of our sample, CHC infection appears to have a significant impact in terms of reducing the age for decline in 25(OH)D and may be related to the degree of liver fibrosis and serum albumin concentration.
Serum albumin revealed to be an independent risk factor for vitamin D inadequacy and deficiency, and that a one unit increase in albumin predicted an increase of 25 nmol/l in the overall serum 25(OH)D concentration (se = 7·03; P < 0·001; data not shown). Indeed, vitamin D-binding protein (DBP) and albumin are known important determinants of the overall circulatory 25(OH)D. About 90 % of total vitamin D is bound with DBP, whereas 9·9 % is loosely bound with albumin, and about 0·1 % is present as free-circulating fraction. Therefore, the availability of 25(OH)D depends not only on the total 25(OH)D concentration but also on the concentration of DBP and albumin and the genetic background of the individual(Reference Jemielita, Leonard and Baker17,Reference Chun18) . For this reason, it seems reasonable that further studies should be performed to assess the impact of extrahepatic hypoalbuminemia on serum 25(OH)D concentrations.
In the literature, there exists clear evidence that the vitamin D system plays a key role in immune modulation(Reference Sassi, Tamone and D’Amelio19). Therefore, by viewing CHC as an inflammatory disease, the nature of this relationship becomes even more important, especially when considering a possible role of the vitamin D system in the regulation of fibrotic process. Indeed, the vitamin D receptor has been shown to regulate hepatic fibrogenesis through the inhibition of hepatic stellate cell activation, down-regulation of the expression of collagen, and antagonism of the SMAD3 pro-fibrotic pathway(Reference Udomsinprasert and Jittikoon20). Moreover, a recent report has shown that CHC subjects with 25(OH)D deficiency might benefit from 25(OH)D supplementation due to its role in the reversion of changes in serum fibrogenic cytokine and enzymes to the pro-fibrolytic state(Reference Komolmit, Kimtrakool and Suksawatamnuay21,Reference Song and Rockey22) . Therefore, from a public health perspective, these data suggest a need for further studies to assess the impact of vitamin D sufficiency on CHC infection with respect to its possible role in the progression of liver fibrosis.
The results of this study should be interpreted with the following limitations in mind. First is the cross-sectional design of this study, which does not provide evidence of a temporal relationship between blood sampling seasonality and vitamin D inadequacy. On the other hand, as the study variables were assessed at the same time, bias owing to variations between independent and dependent variables within the same individual was reduced. Moreover, other confounding factors such as variation in age, gender and skin pigmentation were also controlled. Another limitation was the lack of information on serum 1,25(OH)D concentration, genetic analysis of SNPs of genes encoding DBP and 25-hydroxylase, and other potential confounders such as clothing style and sunscreen use. Despite these limitations, the study provides evidence on the importance of blood sampling seasonality as an important preanalytical factor for the assessment of vitamin D inadequacy in subjects with CHC infection.
In conclusion, this cross-sectional study demonstrated a rather low prevalence of vitamin D inadequacy among CHC-infected subjects in a sunny country. The occurrence of vitamin D deficiency was independently associated with male gender, serum albumin concentration and with samples drawn in winter and spring. These findings reaffirm not only the relevance to consider season as a factor influencing 25(OH)D concentration but also the need to actively screen for hypovitaminosis D in all patients with CHC infection, especially in females and those with low albumin concentration.
Acknowledgements
Acknowledgements: We thank MJR Vaz as well as the nursing staff of the Viral Hepatitis Outpatient Clinic for help with recruiting patients. Financial support: This research received no specific grant from any funding agency, commercial or not-for-profit sectors. Conflict of interest: All authors have no conflicts of interests to disclose. Authorship: All authors had full access to the study data and assume responsibility for data integrity and analytical accuracy. H.P. Jr is the guarantor. The dataset is available from the corresponding author on reasonable request. Ethics of human subject participation: This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects/patients were approved by the Universidade Federal de São Paulo’s Research Ethics Committee. Written informed consent was waived by the Institutional Review Board because of the retrospective nature of the study, and the analysis used anonymous clinical data.
