Vitamin D Deficiency in HIV Infection: Not Only a Bone Disorder

Hypovitaminosis D is a worldwide disorder, with a high prevalence in the general population of both Western and developing countries. In HIV patients, several studies have linked vitamin D status with bone disease, neurocognitive impairment, depression, cardiovascular disease, high blood pressure, metabolic syndrome, type 2 diabetes mellitus, infections, autoimmune diseases like type 1 diabetes mellitus, and cancer. In this review, we focus on the most recent epidemiological and experimental data dealing with the relationship between vitamin D deficiency and HIV infection. We analysed the extent of the problem, pathogenic mechanisms, clinical implications, and potential benefits of vitamin D supplementation in HIV-infected subjects.


Introduction
Human immunodeficiency virus type-1 (HIV) is a global health problem that has infected 60 million people and caused 25 million deaths worldwide. To date, it has been estimated that more than 33 million people, including 2 million children, live infected by HIV. However, even if the problem is far from a definitive solution, highly active antiretroviral therapy (HAART) has profoundly changed the natural history of HIV infection dramatically reducing AIDS-(acquired immune deficiency syndrome-) related morbidity and mortality [1]. Nevertheless, at least until now, HAART cannot eradicate HIV [2,3]. Increased life expectancy exposes HIVinfected subjects both to chronic adverse drug reactions and to age-related morbidities, including neurocognitive disorders, cardiovascular and metabolic disease, renal and bone diseases (i.e., osteopenia/osteoporosis), and cancer [4][5][6]. Many of these appear to occur earlier in HIV patients compared to the general population. Key factors explaining premature age-associated non-AIDS-related events in patients receiving HAART are chronic inflammation and immune activation [7,8]: plasma levels of several inflammatory and coagulopathy biomarkers, such as interleukin-6 (IL-6), highly sensitive C-reactive protein (hsCRP), and D-dimer are higher and correlate with outcome in HIV infection [9,10].
Considering the potential role of vitamin D in many of these chronic illnesses, the scientific community focuses attention on the possible impact of its deficiency on the HIVinfected population. In this review, we first briefly describe vitamin D metabolism and its biological functions; then, we focus on the most recent epidemiological and experimental data dealing with the relationship between vitamin D deficiency and HIV infection. We analyse the extent of the problem, pathogenic mechanisms, clinical implications, and the potential benefits of vitamin D supplementation among HIVinfected subjects. We researched the PubMed database for the period from 1980 through January 31, 2015, using the keywords "HIV, " "vitamin D, " "neurocognitive disorders, " "cardiovascular disease, " "metabolic disease" (i.e., diabetes and metabolic syndrome), "renal disease, " and "cancer. " Articles 2 BioMed Research International presenting original data as well as reviews were included in our analysis.

Prevalence of Hypovitaminosis D in HIV-Infected Subjects
Hypovitaminosis D is a worldwide disorder, with a high prevalence in the general population of both Western and developing countries. It has been estimated that more than 1 billion people suffer from either 25(OH)D (25-hydroxyvitamin D) deficiency or insufficiency. According to the results of the National Health and Nutrition Examination Survey (NHANES), 25(OH)D deficiency and insufficiency are at 79% among adults [11]. Thus, like the general population, it is not surprising to find high rates of hypovitaminosis D among HIV-infected subjects. The overall estimated prevalence in people living with HIV and 25(OH)D deficiency is high, ranging from 70.3 to 83.7% (Table 1). Eckard et al. conducted an investigation on hypovitaminosis D and the possible variables associated with this pathological framework in HIV-infected pregnant women and their infants compared to healthy controls. It was found that 25(OH)D concentrations in serum cord blood were <30 ng/mL in 100% of subjects from both groups. The only variables associated with higher serum 25(OH)D concentrations were white race and non-Hispanic ethnicity [12]. These data agreed with previous observations asserting that vitamin D deficiency not only contributes to HIV disease progression and mortality in HIV-infected pregnant women, but also increases the overall risk of mother-to-child transmission by 46% [13] and the risk of death in newborns by 61% during follow-up [13]. While most infants born to HIV-infected mothers in the US will not acquire HIV infection, in utero ART (antiretroviral therapy) exposure may increase their cancer risk later in life. Thus, maternal and therefore infant 25(OH)D deficiency should not be disregarded [14].

Risk Factors for Vitamin D Deficiency in HIV-Infected Subjects
In the setting of HIV infection, 25(OH)D deficiency may be affected by both HIV-related and -independent risk factors; however, it is often challenging to differentiate the direct impact of HIV infection from the effect of traditional risk factors which may be normal or overexpressed in HIVpositive cohorts.

HIV-Related Risk
Factor. The relationship between 25(OH)D levels, viral load, and CD4+ T-cell count is not clear cut. Some studies described a positive correlation [15,16], some others failed to demonstrate a significant association [17,18], and, finally, others did not find that vitamin D (in any possible formula) supplementation can increase CD4+ count [19,20]. Different mechanisms have been hypothesized to explain the association between 25(OH)D deficiency and higher severity of HIV disease. First, 25(OH)D deficiency may be a contributory causal agent of the HIV infection itself.
Second , chronic inflammation due to HIV infection and subsequent TNF-overproduction may be responsible for renal  1 -hydroxylase impairment, reducing the PTH (parathyroid hormone) stimulatory effect on the production of the  hormonally active 1,25(OH)2D (1,25-dihydroxyvitamin D). Third, infectious complications as a result of poor immunity require hospital care, which significantly reduces the duration of sun exposure for patients. Lastly, both infectious complications and hospitalization may lead to malnutrition and reduced oral intake of the few foods that contain vitamin D [21,22].  [23,24]. An exception is represented by intravenous drug use, which has not been extensively studied in the general population [25]. Injection drug users (IDUs) often have poor nutritional status and limited/delayed access to healthcare. In addition, intravenous drug use increases the risk for a host of acute and chronic infectious and cardiopulmonary conditions. As a result, this patient population suffers a disproportionate burden of 25(OH)D deficiency, compared to other urban dwelling adults [26,27]

Vitamin D Status and HAART
Recently several in vitro and in vivo studies focused on the impact of antiretroviral drugs on vitamin D metabolism. Both protease inhibitors (PIs) and nonnucleoside reverse transcriptase inhibitors (NNRTIs) have been associated with the impairment of vitamin D metabolic pathways [28][29][30]. PIs, especially darunavir and ritonavir, seem to interfere with vitamin D metabolism by inhibition of vitamin D 1and 25 -hydroxylation both in hepatocyte and in monocyte cultures: reduction of 25(OH)D conversion to its active metabolite may potentially explain the reports of increased 25(OH)D levels in subjects with low 1,25(OH)2D. Regarding NNRTIs, there is an increasing amount of experimental data     associating efavirenz (EFV). Unlike what was just reported for PIs, EFV seems to increase 25(OH)D catabolism and production of inactive metabolites, through the interaction with cytochrome P450 enzymes, some of which may affect vitamin D metabolism (i.e., induction of CYP24A1 [31,32] and reduced transcription of CYP2R1), similar to the effects of antiepileptic drugs [33]. This hypothesis has been supported by several in vivo studies, which described an association between NNRTIs, especially EFV and nevirapine (NVP) use and low 25(OH)D levels ( Table 2). The weakness of most reported studies is the crosssectional design, so that causal relationships cannot be inferred. These data suggest the need for large prospective studies, properly designed to evaluate the specific effects and clinical impact of antiretroviral drugs on vitamin D status.

Association between HIV, Hypovitaminosis D, and Cardiovascular Disease
Several studies have described the association between HIV and increased risk of CVD (cardiovascular disease) [34,35]. HIV infection itself is considered an independent risk factor for atherosclerosis: the prevalence of carotid intima-media thickness (cIMT), atherosclerosis, and myocardial infarction is higher among HIV-positive subjects, occurring earlier compared to uninfected individuals [36,37]. In these patients atherogenesis is enhanced by several factors: HIV-induced chronic inflammation and immune activation (demonstrated by increased levels of proinflammatory cytokines and endothelial activation markers), excess of traditional risk factors (e.g., 2 to 3 times higher prevalence of smoking), and antiretroviral drug-related dyslipidemia, hyperglycemia, central obesity, and lipodystrophy (especially with PIs) [38][39][40].
To make this framework even more complex, 1,25(OH)2D deficiency has been linked to CVD in the general population [41,42]. Vitamin D influences cardiovascular health by suppressing the renin-angiotensin system and stimulating cellular proliferation and differentiation via 1,25(OH)2D binding to vitamin D receptors in the heart, the endothelium, and the vascular smooth muscle [43,44].

cIMT, Brachial Artery Flow-Mediated Dilation, Coronary Artery Calcium (or Calcification), and Coronary Artery
Stenosis. Considering the high prevalence of both hypovitaminosis D and CVD in patients with HIV, the evidence of a relationship between low 25(OH)D and silent and symptomatic atherosclerosis is not surprising. Even though in the general population asymptomatic CVD, as demonstrated by cIMT, brachial artery flow-mediated dilation (FMD, an early marker of endothelial dysfunction), and CAC (coronary artery calcification), has been strongly linked to the occurrence of cardiovascular events and has also been independently associated with 25(OH)D deficiency, only a few studies are available in HIV-infected populations; moreover, none of these studies shows if 25(OH)D repletion might affect cardiovascular outcomes. The clinical characteristics of the populations, the study designs, and the variables included in the analysis of results could explain the differences among the studies [45][46][47] (Tables 3 and 4).

Other Risk Factors for CVD in HIV-Infected Subjects.
Other traditional risk factors for CVD, such as insulin resistance and diabetes mellitus, are frequently seen in HIVpositive individuals [48], and, as in the general population [49], an association between vitamin D status and type 2 diabetes, but not with insulin resistance, has been described [50,51].

Association between HIV, Hypovitaminosis D, and HIV Disease Progression
Preclinical experiments have demonstrated that treatment of peripheral blood mononuclear cells with 1,25(OH)2D decreased the cell susceptibility to HIV infection by inhibiting viral entry, modulating expression of CD4+ cell surface antigen, damping viral p24 production, and limiting monocyte proliferation [53,54]. Thereafter, several observational studies have shown a significant association between higher levels of 25(OH)D and rates of immune recovery [15,55,56]. Along these lines, some authors investigated the association between vitamin D and clinical outcomes. Baseline 25(OH)D levels lower than 32 ng/mL were independently associated with progression to more advanced HIV stage among 884 HIV-infected pregnant women in Tanzania, who were followed for a median of 70 months. The women with 25(OH)D in the highest quintile had a 42% lower risk of allcause mortality than the women in the lowest quintile [13].
The same authors demonstrated that 25(OH)D deficiency was also associated with low BMI, oral thrush, acute upper respiratory infections, and severe anemia [57]. However, other studies failed to demonstrate an association between

68.4% had 25(OH)D <30
ng/mL among cART   Authors found an association between hypovitaminosis D and exposure to NNRTIs ( = 0.05) but not to EFV and NVP, probably because of a lack of statistical power of their analysis. However, considering the crude and adjusted coefficients for EFV and NVP in their regression models, the authors suggested a NNRTI class effect, rather than a specific EFV or NVP impact, on vitamin D levels.    An association between vitamin D insufficiency and cIMT, even after adjusting for age, sex, tobacco use, hypertension, and elevated cholesterol, was proved.
The authors found that mean cIMT was 0.13 mm greater in vitamin D insufficient subjects than in normal subjects. Authors found no association between 25(OH)D insufficiency, inflammatory, or endothelial dysfunction markers and cIMT, whereas this was found between cIMT and patient age, impaired fasting glucose, and PI therapy length. 25(OH)D: 25-hydroxyvitamin D; ART: antiretroviral therapy; ccIMT: common carotid intima-media thickness; cIMT: carotid intima-media thickness; US: United States; icIMT: internal carotid intima-media thickness; CVD: cardiovascular diseases; HOMA-IR: homeostasis model assessment of insulin resistance; NNRTI: nonnucleoside reverse-transcriptase inhibitor; NRTI: nucleoside reverse-transcriptase inhibitor.

Association between HIV, Hypovitaminosis D, and Hepatitis C
HCV (hepatitis C virus) infection occurs at a significantly higher rate in HIV-infected persons compared to the general population, and this is especially problematic for resourcelimited settings, where HCV treatment is generally not easily available [59]. HIV has a negative impact on the natural history of HCV, and, compared to HCV monoinfected patients, HIV/HCV coinfected patients have a more rapid progression from chronic active hepatitis to liver-cirrhosis, end-stage liver disease, liver cancer, and death, as well as lower response rate to traditional HCV treatment [60,61]. Male sex, insulin resistance, acquiring HCV at an older age, heavy alcohol consumption, HCV genotype 3, and low CD4+ cell count are the factors contributing to the rapid development of liver fibrosis/cirrhosis among HIV/HCV coinfected patients [62,63]. Other studies of HCV monoinfected patients have shown an independent association of 25(OH)D deficiency with severe liver fibrosis and treatment failure [64]. 1,25(OH)2D effects on the immune system and inflammatory response have been shown to directly inhibit the proliferation and profibrotic effect of hepatic stellate cells [65]. Not surprisingly, liver fibrosis is associated with low serum levels of 25(OH)D during both HBV-and HCV-related chronic hepatitis, in both HIV-coinfected and not-coinfected patients [66]. However, low levels of 25(OH)D have been found in HBV or HCV carriers with minimal or absent liver fibrosis compared to healthy subjects [67].
On the other hand, in HIV-HCV coinfected patients, studies investigating the association between HCV sustained virologic response and vitamin D level have reported varying results, with some studies demonstrating an association [68], whereas other studies do not [69]. Mandorf et al. demonstrated that serum levels of 25(OH)D may predict the response to anti-HCV therapy. Suspicion of such a connection is strengthened by the evidence that cholecalciferol supplementation improves early and sustained virological BioMed Research International 13 response (94% versus 48% in controls and 86% versus 42% in controls, resp.) in HCV genotype 1 patients treated with Peg-IFN/ribavirin [70,71]. The exact mechanism of its antiviral (anti-HCV) effect is unknown, although it was recently shown to amplify the innate antiviral immune response upregulating IFN-and the MxA (an IFN-induced human protein) gene expression and dampening interferon gammainduced protein 10 (IP-10) expression [72].

Association between HIV, Hypovitaminosis D, and Tuberculosis
According to the World Health Organization approximately 2 billion people are exposed to M. tuberculosis, 8 million people per year are infected, and 2 million people die as a clinical outcome [73]. HIV is the strongest factor in the development of active TB (tuberculosis), and its spread has fuelled the resurgence of the TB epidemic. It has been proposed that in HIV infection M. tuberculosis escapes the local immune response within the granulomas, decreasing their containing ability and then leading to increased mycobacterial replication, dissemination, and clinical disease [74]. The rise in CD4+ count and improved immune function after HAART initiation partially restore pathogen specific immunity. In the general population, 25(OH)D levels lower than 4 ng/mL were shown to cause a 3-fold probability of having active TB [75], with higher risk of developing MDR M. tuberculosis infection [76]. A cross-sectional study of 174 HIV-infected and 196 HIV-uninfected individuals in Cape Town, South Africa, showed that 25(OH)D deficiency is independently associated with active TB and this association is greater in HIVinfected subjects [77]. A prospective Tanzanian cohort study enrolled 1103 HIV-infected persons initiating HAART in a randomized controlled trial (RCT) of vitamin D-free multivitamin supplementation. Baseline 25(OH)D levels lower than 20 ng/mL, but not 25(OH)D insufficiency, were associated with higher incident smear-positive TB, after a median follow-up of 20.6 months, wasting, and >10% weight loss but not with risk of malaria, pneumonia, or anaemia. Mortality hazard ratio was 2.0 for those with levels below 20 ng/mL versus those with levels above 30 ng/mL over 24 months. Reverse causality (i.e., that vitamin D deficiency occurred as a result of TB) was ruled out in this study by the exclusion of patients who developed TB within 1 month of enrolment. This finding is significant, since TB itself might contribute to vitamin D deficiency by reducing a patient's sun exposure or increasing consumption of 25(OH)D by activated macrophages [78]. Recently a systematic review was conducted to analyse studies published from 1980 to 2006 with data on serum 25(OH)D in pulmonary TB patients and controls. Five out of seven case-control studies, with a total of 531 participants, reported lower serum 25(OH)D in cases compared to controls. Several weaknesses were found: the sample sizes were small, ranging between 30 and 145 participants; some studies did not use culture for diagnosing TB; some studies included extrapulmonary TB; selection of controls was not optimal [79].

Chronic Inflammation.
HIV infection is associated with chronic inflammation (i.e., elevated TNF, IL-6, and CRP) and immune system activation (i.e., increased soluble CD14 and CXCL10), even after achieving full virologic suppression and immune recovery with the use of HAART [80,81]. In this population, elevation of inflammation markers has been shown to be independent predictors of neurocognitive impairment, frailty, cardiovascular events, diabetes and metabolic syndrome, low BMD, malignancies, and all-cause mortality [82][83][84][85]. The same outcomes, including all-cause mortality, were also associated with chronic inflammation in the general population [86]. Thus, there seems to be a considerable overlap in the outcomes associated with 25(OH)D deficiency and chronic inflammation, in both the HIV-infected and HIVuninfected populations.

Malignancy.
Association of vitamin D deficiency with risk of cancer in the HIV-infected population remains to be determined. However, it has already been shown in the general population, including breast cancer (4-fold risk) [87], colon cancer (2-fold risk) [88], ovarian cancer (4-fold risk) [89], and prostate cancer (3-fold risk) [90].

Management of Hypovitaminosis D in HIV-Positive Individuals
10.1. Screening. The main arguments in favor of routine screening of vitamin D in HIV-infected patients include the potential optimization of skeletal, metabolic, and immunologic parameters with vitamin D supplementation. The arguments against routine screening include assay variability and costs, lack of a clear target range, absence of proven supplementation benefits, apart from the benefits connected with osteoporosis as in the general elderly population, limited randomized clinical trial data in HIV-infected patients, inability to distinguish the effects of vitamin D and calcium supplementation on bone, potential harm from some supplementation approaches, and increased pill regimen (possible reduction of patient compliance).
The European AIDS Clinical Society most recent guidelines suggest vitamin D status evaluation in patients with a history of low BMD or fracture, those with high risk of fracture, or those with other vitamin D deficiency associated factors (e.g., persons receiving some antiretroviral drugs, including EFV). Vitamin D replacement is recommended when 25(OH)D is lower than 10 ng/mL; for values ranging between 10 and 20 ng/mL, supplementation is recommended only for patients with osteomalacia, osteoporosis, or increased PTH [92]. McComsey et al. developed recommendations for bone disease in HIV infection, addressing vitamin D deficiency as well. They recommend 50,000 IU of cholecalciferol weekly for 8 to 12 weeks and then monthly thereafter or 2,000 IU daily for 12 weeks and then 1,000 to 2,000 IU daily thereafter. 25(OH)D levels after replacement should be measured. They recommend supplementation to achieve 25(OH)D greater than 32 ng/mL [93].

Supplementation.
In the general population, recommendations regarding vitamin D supplementation are mostly derived from studies on bone health. Several large RCTs found beneficial effects of vitamin D plus calcium on BMD and fracture risk [94]. Meta-analyses showed that vitamin D (cholecalciferol) plus calcium association is superior to the use of a single drug in fracture prevention [94]. Unfortunately, the evidence for vitamin D use in clinical outcomes beyond skeletal health (i.e., on falls, CVD, diabetes, metabolic syndrome, immune response, and cancer) is inconsistent with and insufficient to base general recommendations.

Vitamin D Supplementation Dosage.
In the general population, current recommended vitamin D oral supplementation is 800-1,000 IU cholecalciferol/day, plus calcium 1000 mg to 1200 mg daily. Serum 25(OH)D levels generally increase by approximately 1 ng/mL for every 100 IU of cholecalciferol intake. Few data from small cohorts are available on the efficacy of cholecalciferol repletion in HIV-infected subjects [95]. However, in this subgroup, it seems that a daily intake of at least 1,000-2,000 IU/day may be needed to overcome vitamin D deficiency [29].

Safety of Vitamin D Supplementation. Groleau et al. demonstrated that supplementation with high vitamin D(3)
doses and the concomitant increased serum 25(OH)D level did not correlate with increased whole blood lead concentration in HIV-infected children and young adults. Vice versa, the more robust increase in serum 25(OH)D after 12 weeks of vitamin D(3) supplementation for participants enrolled during winter and spring was accompanied by a decrease in whole blood lead concentration [96]. Animal studies show an inverse relationship between calcium intake and lead levels. This inverse relationship was also found in pregnant women, and calcium supplementation during pregnancy was associated with reductions in blood lead. Overall, the above data provide safety information when considering higher dose vitamin D intervention [97].

Extraskeletal Effects of Vitamin D Supplementation.
There are only a few studies investigating the effect of cholecalciferol supplementation on other cardiovascular, metabolic, and immunological outcomes in the HIV-infected population. In an RCT involving 45 subjects with 25(OH)D lower than 20 ng/mL, 12-week supplementation with daily oral cholecalciferol 4,000 IU produced an increase of approximately 5 ng/mL in 25(OH)D level compared to placebo but did not result in a statistically significant change in brachial artery FMD. Moreover, in the study group, insulin resistance increased from baseline but it was not statistically different from the placebo arm; similarly, baseline inflammatory and coagulation markers (i.e., CRP, IL-6, sTNFR-1, ICAM, vascular cell adhesion molecule (VCAM), D-dimer, and fibrinogen) did not significantly change between the groups. These results could partly be attributed to the modest increase in 25(OH)D (5 ng/mL) in subjects receiving cholecalciferol supplementation [98]. In an RCT involving 52 mostly virologically suppressed vertically infected youths aged 8 to 26 years with 25(OH)D lower than 30 ng/mL, Giacomet et al. showed that 12-month supplementation with cholecalciferol 100,000 IU every 3 months resulted in reduction of antiinflammatory T-cell phenotype (i.e., decrease in T H 17 : T reg ratio) at 3 months. This effect was no longer seen at 12 months. No significant change in baseline CD4+ cell count was observed between the treatment and placebo arms [99]. patients, higher 25-hydroxyvitamin D concentrations were not related to hepatitis C virus treatment responses but were associated with ritonavir use, " The American Journal of Clinical Nutrition, vol. 98, no. 2, pp. 423-429, 2013.