Endocrine problems and low vitamin D – review May 2014

Vitamin D as a potential contributor in endocrine health and disease
Eur J Endocrinol May 28, 2014 EJE-14-0158

PDF is attached at the bottom of this page

2 Giovanna Muscogiuri 1, Joanna Mitri 2, Chantal Mathieu3, Klaus Badenhoop4, Gonca Tamer5,
3 Francesco Orio6'7, Teresa Mezza8, Reinhold Vieth 9'10, Annamaria Colao1, Anastassios Pittas2 4
5 1. Department of Clinical Medicine and Surgery, University "Federico II" Naples, Italy;
6 2. Division of Endocrinology, Diabetes and Metabolism, Tufts Medical Center, Boston, MA, USA;
8 3. Department of Endocrinology, UZ Gasthuisberg, 3000, Leuven, Belgium
9 4. Department of Medicine 1, Division Endocrinology & Diabetology, University Hospital
10 of the Goethe-University Frankfurt, Frankfurt am Main, Germany
11 5. Division of Endocrinology and Metabolism, Department of Internal Medicine, Goztepe
12 Training and Research Hospital, Medeniyet University, Istanbul, Turkey
13 6. Endocrinology, University "Parthenope" Naples, Naples, Italy;
14 7. Endocrinology of Fertile Age, University Hospital "S. Giovanni di Dio e Ruggi d'Aragona" Salerno, Italy
16 8. Endocrinology and Metabolic Diseases, Universita Cattolica del Sacro Cuore, Roma, Italia
18 9. Department of Nutritional Sciences, University of Toronto, Toronto, Canada
19 10. Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Canada 21
22
29 Correspondence and reprint to: Giovanna Muscogiuri, MD - Via Sergio Pansini, 5 - 80131
30 Napoli - Italy; Tel. 0817464983; Fax 0815465443; e-mail: giovanna.muscogiuri at gmail.com.
31
35 Objective: It has been suggested that vitamin D may play a role in the pathogenesis of several
36 endocrine diseases such as hyperparathyroidism, type 1 diabetes, type 2 diabetes, autoimmune
37 thyroid diseases, Addison's disease, and polycystic ovary syndrome. In this review, we debate
38 the role of vitamin D in the pathogenesis of endocrine diseases.
39 Methods: Narrative overview of the literature synthesizing the current evidence retrieved
40 from searches of computerized databases, hand searches, and authoritative texts.
41 Results: Evidence from basic science supports a role for vitamin D in many endocrine
42 conditions. In humans, inverse relationships have been reported between not only blood 25
43 hydroxyvitamin D and parathyroid hormone concentrations but also with risk of type 1
44 diabetes, type 2 diabetes and polycystic ovary syndrome. There is less evidence for an
45 association with Addison's disease or autoimmune thyroid disease. Vitamin D
46 supplementation may have a role for prevention of type 2 diabetes, but the available evidence
47 is not consistent.
48 Conclusions: Although observational studies support a potential role of vitamin D in
49 endocrine disease, high quality evidence from clinical trials does not exist to establish a place
50 for vitamin D supplementation in optimizing endocrine health. Ongoing randomized
51 controlled trials are expected to provide insight into the efficacy and safety of vitamin D in
52 the management of endocrine disease.
53

Introduction

64
65 The main physiologic role of vitamin D is to regulate calcium and phosphorus homeostasis
66 and to promote bone health. However, accumulating evidence from animal and human studies
67 suggests that vitamin D may also be important for a variety of non-skeletal actions that may
68 be important in the pathogenesis of several endocrine diseases. The increased appreciation of
69 the pleiotropic effects of vitamin D and the high prevalence of hypovitaminosis D in the
70 general healthy population has generated very high interest in vitamin D among researchers,
71 clinicians and the lay public. Vitamin D has been implicated in the pathogenesis of several
72 endocrine conditions, including primary hyperparathyroidism, type 1 diabetes (1), type 2
73 diabetes (2,3), autoimmune thyroid (4) adrenal diseases (5) and polycystic ovary syndrome
74 (PCOS). The present review focuses on the reported association between vitamin D status and
75 endocrine diseases and the potential role of vitamin D supplementation in the treatment of
76 endocrine diseaser

Vitamin D homeostasis

78
79 Humans derive vitamin D from cutaneous synthesis (in the form of cholecalciferol D3), from
80 diet (in the form of D3) and from nutritional supplements in the form of D3 or ergocalciferol
81 D2 (6).6 Upon exposure to ultraviolet B radiation (UVB), 7-dehydrocholesterol in the skin is
82 converted to pre-vitamin D3, which is immediately converted to vitamin D3 in a heat-
83 dependent process. After ingestion or synthesis, vitamin D is hydroxylated in the liver to
84 form 25 hydroxyvitamin D (25(OH)D2 or 25 (OH)D3), its major circulating form, which has
85 little biological activity. 25(OH)D is converted in the kidney by 25(OH)D-1alpha-
86 hydroxylase (CYP27B1), to its bioactive hormonal metabolite 1,25 dihydroxy-vitamin D
87 (1,25[OH]2D or calcitriol). The primary action of 1,25(OH)2D is through the nuclear vitamin
88 D receptor, which heterodimerizes with the retinoid X receptor and binds to vitamin D
89 responsive elements near target genes (6,7). The primary action of 1,25[OH]2D is to enhance
90 intestinal calcium absorption and to promote osteoclast function, thereby maintaining calcium
91 and phosphorus homeostasis and bone health. However, the discovery that nearly all tissues in
92 the body express the vitamin D receptor and that several tissues also express CYP27B1,
93 thereby allowing for local production of 1,25OH2D with a paracrine effect, has provided
94 important insights into the pleiotropic effects of vitamin D and its potential role in a variety of
95 extra-skeletal tissues (7), including many that affect endocrine disease.
96

Classification of vitamin D status and vitamin D intake requirements

98
99 Blood concentration of 25OHD is the biomarker used by clinicians and researchers to
100 determine vitamin D status. However, there is no consensus on the 25OHD thresholds for
101 defining vitamin D adequacy. The guidelines by the Institute of Medicine (IOM) and the
102 Endocrine Society differ on classification of vitamin D status (8,9). These differences are
103 explained by the populations targeted by the guidelines and how the evidence was
104 synthesized. Specifically, the IOM guidelines concentrated on the general healthy population
105 and considered only trials and concluded that blood concentration of 25OHD > 20 ng/mL is
106 consistent with favourable skeletal outcomes. In contrast, the Endocrine Society clinical
107 practice guidelines concentrated on people at high risk for vitamin D deficiency and
108 considered both trials and observational (epidemiologic) studies in concluding that blood
109 concentration of 25OHD > 30 ng/mL is desirable for optimal skeletal outcomes without any
110 upper limit that would be concerning for safety. Both guidelines agreed that recommendations
111 will require reconsideration in the future as additional data from on-going randomized trials
112 become available. Variability in vitamin D-binding protein and bioavailable 25OHD
113 concentration may also be important when assessing vitamin D status, especially in certain
114 populations, such as African-Americans. (10)
115
116 The recommended intakes by the two guidelines also differ. The IOM report on dietary
117 reference intakes for calcium and vitamin D recommends 600 international units per day of
118 vitamin D for individuals 9-70 years and 800 international units for those older than 70 years
119 as the recommended dietary allowance (RDA), which is defined as the intake that meets the
120 needs of 97.5% of the healthy population. In contrast, the Endocrine Society clinical practice
121 guidelines conclude that to raise the blood level of 25OHD consistently above 30 ng/mL,
122 intakes of 1500 to 2000 IU/day are required. The IOM report recognized the lack of trials
123 with vitamin D supplementation for non-skeletal outcomes as a major hurdle in establishing
124 recommendations, while the Endocrine Society guidelines applied evidence from
125 observational studies to develop its recommendations and considered blood 25OHD
126 concentration as a clinically important surrogate outcome that correlates with health and
127 disease. 128

Vitamin D and Primary Hyperparathyroidism

130
131 Ionized calcium is the most tightly regulated analyte in the circulation. This fine regulation is
132 achieved through an interplay between parathyroid hormone (PTH), calcitonin and
133 1,25(OH)2D. Parathyroid hormone is the major stimulator of renal CYP27B1, which
134 increases biosynthesis of 1,25(OH)2D. In turn, PTH is down-regulated both by 1,25(OH)2D
135 and ionized calcium.
136
137 The seemingly inactive vitamin D metabolite, 25(OH)D, is an important regulator within
138 parathyroid tissue. Parathyroid cells take up vitamin D binding protein along with its
139 25(OH)D, which is the mechanism that provides parathyroid tissue with better access to
140 circulating 25(OH)D than most other tissues. Furthermore, the parathyroid glands possess the
141 enzyme, CYP27B1, which produces 1,25(OH)2D for local paracrine regulation. The
142 combined effect of efficient access to circulating 25(OH)D and 1,25(OH)2D plus the local
143 production of 1,25(OH)2D is suppression of both PTH secretion and parathyroid cell
144 proliferation (11). 145
146 Larger parathyroid adenomas respond poorly to feedback by calcium or 1,25(OH)2D;
147 consequently, in primary hyperparathyroidism, 1,25(OH)2D levels often correlate positively
148 with circulating 25(OH)D (12). If vitamin D supply is low, then primary
149 hyperparathyroidism can remain latent, known as "normocalcemic primary
150 hyperparathyroidism" (13). Hypercalcemia develops once the 25(OH)D concentration
151 increases and elevated PTH stimulates renal CYP27B1 relentlessly, which generates
152 1,25(OH)2D in proportion to the supply of 25(OH)D. This relationship highlights a
153 fundamental aspect of the vitamin D system: its operation under first-order reaction kinetics,
154 namely, the yield of the product (1,25OH2D) is proportional to the supply of the substrate
155 (25OHD). Therefore, the enzymes of the vitamin D system need to modify their function
156 according to the supply of 25(OH)D. Depending on severity, primary hyperparathyroidism
157 can disrupt the adaptation, resulting in elevated 1,25(OH)2D and increased intestinal calcium
158 absorption. While this form of hypercalcemia, promoted by the underlying parathyroid
159 adenoma, is not strictly a manifestation of vitamin D toxicity, it is a form of hypersensitivity
160 to higher doses of vitamin D that is important to consider, given how common parathyroid
161 adenomas are (14). 162
163 In healthy persons, the reference (normal) range for serum PTH is known to decline as serum
164 25(OH)D levels increase. Therefore, the theoretical plateau in PTH, as 25(OH)D increases,
165 can be used as a determinant in establishing adequacy of vitamin D status. That relationship
166 breaks down in primary hyperparathyroidism, because with disease progression, PTH
167 becomes an unregulated driver of 25(OH)D metabolism into 1,25(OH)2D, a powerful
168 hypercalcemic hormone. Ongoing research (table 1) will further clarify the effect of vitamin
169 D supplementation on hyperparathyroidism.
170

Vitamin D and type 1 diabetes (T1DM)

172
173 Type 1 diabetes is one of the first endocrine disorders where a potential role for vitamin D
174 was reported. Type 1 diabetes is characterized by an autoimmune destruction of the insulin
175 producing pancreatic islet beta-cells, rendering patients dependent on insulin administration
176 for survival (15). Potential effects of vitamin D deficiency on T1DM are multiple, including
177 alterations in the innate immune system, such as impaired macrophage function, but also
178 dysfunction of the beta-cell itself (16). Caution, however, is warranted when postulating a
179 direct effect of vitamin D deficiency on immune or beta-cell function in vivo, as vitamin D
180 deficiency leads to decreased calcium concentration, with calcium being a crucial ion both for
181 immune function and insulin secretion. In non obese diabetic (NOD) mice, the principal
182 animal model of T1DM, severe vitamin D deficiency increases the risk for developing
183 diabetes (17), but absence of any effect on T1DM presentation in vitamin D receptor
184 knockout NOD mice suggests a redundancy of the vitamin D system in the pathogenesis of
185 T1DM (18). In vitro, the active form of vitamin D, 1,25(OH)2D, directly protects beta-cells
186 from the destructive effects of inflammatory cytokines and limits the inflammatory profile of
187 macrophages (19,20).
188
189 In NOD mice, administration of high doses of 1,25(OH)2D from early life onwards lowers
190 incidence of T1DM (21, 22), highlighting vitamin D as a promising intervention in preventing
191 T1DM. To replicate the immune and metabolic effects of the active form of vitamin D seen in
192 the mouse model, very high doses are required that would induce hypercalcemia and
193 hypercalciuria and potentially bone decalcification (23). Synthetic analogues of 1,25(OH)2D
194 with immune modulatory function but lesser effects on calcium and bone have been
195 developed to overcome such an obstacle. In NOD mice, such analogues can prevent
196 progression of the disease (24,25), which has been postulated to be due to the direct beta-cell
197 protective effects of 1,25(OH)2D combined with the blocking of inflammation, together with
198 the regulator T lymphocytes, which may be partly a direct effect on T lymphocytes, but also
199 via an effect on the central antigen presenting cells, the dendritic cells (26). In vitro, the
200 presence of 1,25(OH)2D or an analogue results in the generation of dendritic cells with
201 specific characteristics, such as less IL12 secretion, less CD80/CD86 expression, less MHC II
202 expression and most importantly less stimulation of effector T cells and specific generation of
203 regulator T cells (27, 28). Thus, a second possible avenue to exploit the potential beneficial
204 immune modulatory effects of vitamin D is the auto-transfer of ex vivo 1,25(OH)2D (or
205 analogue) -exposed dendritic cells generated from peripheral blood monocytes from patients
206 with T1DM. Upon transfer back into patients, these dendritic cells should be able to induce
207 regulator T cells and shift the immune system from attack towards tolerance towards the beta-
208 cell. Clinical trials exploring this potential therapeutic avenue are underway.
209
210 In population-based studies, low vitamin D concentration especially in early life, has been
211 associated with a higher risk for T1DM (1). Lower concentrations of 25(OH)D were reported
212 in North Indian (29), Italian (30), Swedish (31), and British (32) children or young adults
213 with newly diagnosed T1DM compared to controls. An increased prevalence of vitamin D
214 deficiency in children and adolescents with T1DM compared with non-diabetic individuals
215 was also observed in American (33), Australian (34) and Qatari (35) populations. Of interest,
216 there are also reported associations between polymorphisms of genes involved in the vitamin
217 D system and metabolism and type 1 diabetes risk islet autoimmunity risk (36,37).
218
219 Several observational studies have found that supplementation (based on self-reported data)
220 with vitamin D in early life is associated with a lower risk of T1DM in later life (38,39). In a
221 retrospective case-control study in Norway, intake of cod-liver oil by children during infancy
222 did not prevent T1DM, though there was a trend towards an inverse association (40). More
223 recently, the ABIS study in Sweden reported that the use of vitamin D-containing
224 supplements during pregnancy was associated with reduced development of autoantibodies to
225 GAD or IA-2A in offspring of T1DM parents at 1 year, but not at 2.5 years (41). In small
226 intervention studies, data on the effect of vitamin D supplements in patients with established
227 T1DM have been disappointing. For example, a study in Europe showed that administration
228 of 0.25 |ig 1,25(OH)2D3 was safe but failed to reduce loss of beta-cell function, even in
229 patients with high C-peptide at diagnosis (42).
230
231 In summary, based on observational studies, vitamin D deficiency (defined as 25OHD < 12
232 ng/mL) should probably be avoided in individuals at high risk for developing T1DM,
233 specifically in early life. However, whether supplementation with high dose vitamin D or its
234 analogs have a therapeutic role in prevention or treatment of T1DM is presently under
235 investigation (Table 2).
236

Vitamin D and Type 2 Diabetes (T2DM)

238
239 Among the multiple associations that have been reported between vitamin D status and
240 chronic diseases, the link between vitamin D and T2DM stands as one of the most promising
241 ones. The potential effect of vitamin D on glycemia appears to be mediated by direct and
242 indirect effects on three different pathways: insulin secretion, insulin sensitivity and systemic
243 inflammation. A direct effect of vitamin D on insulin secretion may be mediated by activation
244 of vitamin D receptors in the pancreatic beta cells. Vitamin D receptor is expressed in
245 pancreatic cells and mice lacking vitamin D receptor have impaired insulin secretion (43). In
246 addition, the direct effect of vitamin D on insulin synthesis is supported by the presence of the
247 vitamin D response element in the human insulin gene promoter (44). Importantly, pancreatic
248 beta cells express CYP27B1, thereby generating 1,25(OH)2D locally, which allows for a
249 paracrine effect of vitamin D. A direct effect of vitamin D on insulin sensitivity may be
250 mediated by stimulating the expression of insulin receptors in peripheral insulin target cells.
251 In addition, vitamin D insufficiency is associated with increased fat infiltration in skeletal
252 muscle, independent of body mass, which is thought to contribute to decreased insulin action.
253 Vitamin D may also decrease the effects of systemic inflammation, known to play an
254 important role in the pathogenesis of T2DM, in several ways, which include directly
255 modulating the expression and activity of cytokines in addition to several other non-cytokine
256 related immune-modulating effects (45). Finally, insulin secretion and insulin sensitivity are
257 both calcium dependent processes (46,47); therefore, vitamin D could affect both pathways
258 indirectly, through alteration in calcium concentration and flux through the cell membranes of
259 the pancreas and insulin responsive tissues. 260
261 The data from observational studies strongly support an association between low vitamin D
262 status and incident T2DM. Recently, two meta-analyses of observational studies were
263 published with nearly identical results. Song et al. reported a 38% lower risk of developing
264 T2DM in the highest reference category of 25OHD compared to the lowest one (relative risk
265 0.62 95% CI 0.54-0.70 (48) while Afzal et al. (49) reported an odds ratio for T2DM of 1.5
266 (95% CI 1.33-1.70) for the bottom versus top reference category of 25OHD.
267
268 Randomized studies have shown inconsistent results. In trials that included participants with
269 normal glucose tolerance or established diabetes at baseline, vitamin D supplementation had
270 no effect on glycemic measures or incident diabetes. It is crucial to note, however, that most
271 studies were underpowered or were post-hoc analyses of completed trials. In addition, the
272 results differed based on the adherence to vitamin D supplementation. For example, in a post-
273 analysis of the RECORD study, while supplementation with 800 IU/day of vitamin D3 did not
274 change the risk of self-reported type 2 diabetes, there was a notable trend towards reduction in
275 T2DM risk with vitamin D3 (odds ratio 0.68; 95% CI 0.40-1.16) among study participants
276 who were highly compliant with supplementation (50).
277
278 Vitamin D supplementation appears to be more promising in patients who are at risk for
279 diabetes. In the Calcium and Vitamin D for type 2 Diabetes Mellitus (CaDDM) study, a 2x2
280 factorial design trial in participants with pre-diabetes, vitamin D supplementation improved
281 disposition index, a measured of beta cell function (51). However, in another trial of non-
282 Caucasians, very high dose vitamin D supplementation had no effect on insulin secretion,
283 insulin sensitivity or incident diabetes in a population with impaired fasting glycemia or
284 impaired glucose tolerance and low vitamin D levels (52).
285
286 In summary, vitamin D appears to have no role in prevention of T2DM in the general
287 population; however, there might be a role for vitamin D for treatment of established T2DM
288 or prevention of T2DM in persons at risk, although the evidence from available trials is
289 inconsistent. There are several ongoing large randomized trials in well-defined populations
290 (D2d [NCT01942694], VITAL [NCT01633177], DDM2 [NCT01736865], Table 2) to test the
291 hypothesis that vitamin D deficiency is a contributor to the pathogenesis of T2DM and to
292 define its role in prevention or therapy of T2DM.
293

Vitamin D and Addison's disease

295
296 Addison's disease is a rare condition resulting from autoimmune mediated destruction of the
297 adrenal cortex and may present as either isolated adrenal deficiency or part of an autoimmune
298 polyendocrine syndrome. Although the etiology of Addison's disease is largely elusive,
299 current concepts point to environmental factors acting as triggers in a background of genetic
300 susceptibility leading to destructive CD8-T-lymphocytic infiltration of the adrenal cortex and
301 characteristic 21-OHase antibody production (53). Although the main genetic susceptibility is
302 identified at the HLA locus (54), other susceptibility genes have been described, including in
303 the vitamin D receptor (VDR) (55) and CYP27B1 (56,57), similarly to other autoimmune
304 endocrine diseases (e.g., T1DM) (32). This shared genetic association led to the assumption
305 that the vitamin D system may be involved in critical pathophysiologic pathways in these
306 immune mediated inflammatory disorders since active 1,25(OH)2D may suppress
307 steroidogenesis by down-regulating CYP21A2 and up-regulating CYP11A1 and CYP17A1.
308 In an adrenal cell model (NCI-H295R line) (58), vitamin D acts not only on the immune
309 system but also on the adrenal tissue itself.
310
311 Whether 25(OH)D concentrations differ between patients with autoimmune Addison's
312 disease and controls is not known and is currently under investigation. However, there is
313 evidence of interaction between vitamin D status and predisposing gene loci, similar to
314 findings in T1DM (34). In summary, preliminary evidence suggests that vitamin D may be
315 important in modifying genetic susceptibility in Addison's disease; however, much remains to
316 be studied on its functional and clinical relevance in humans.
317

Vitamin D and Hashimoto's thyroiditis

319
320 Hashimoto's thyroiditis is predominantly a disease of cell-mediated immunity that is
321 manifested by a genetic defect in suppressor T-cell function. Th1 cells secrete various
322 cytokines such as interferon (IFN)-y which induces thyrocytes to express major
323 histocompatibility complex class II (MHC II) surface HLA-DR antigens and renders them
324 susceptible to immunologic attack. Although HLA-DR antigens are not physiologically
325 expressed on thyrocytes, in Hashimoto's thyroiditis, the thyrocytes present HLA-DR antigens
326 on their surface, which may trigger autoimmune process. Activated by T lymphocytes, B
327 lymphocytes produce autoantibodies that react to thyroid antigens (59-62).
328 In Hashimoto's thyroiditis, the autoimmune process may be suppressed at various stages by
329 1,25(OH)2D. At first, vitamin D might suppress dendritic cell-dependent T cell activation,
330 then, it might decrease proliferation of Th1 cells and the synthesis of Th 1 cell cytokines such
331 as IFN-y. Vitamin D may also inhibit expression of MHC II surface HLA-DR antigens on
332 thyrocytes by inhibiting the synthesis of IFN-y, which induces thyrocytes to express those
333 antigens. Furthermore, after being activated by T cells, B cells' ongoing proliferation may be
334 suppressed and B cell apoptosis may be induced by 1,25(OH)2D. In this way, vitamin D might
335 decrease autoantibodies that react with thyroid antigens (59-61).
336 Recently, studies have suggested that low vitamin D concentrations and other conditions
337 which may result with reduced vitamin D function (e.g., certain VDR gene polymorphism,
338 pathologies of vitamin D binding protein and its gene) may increase risk of Hashimoto's
339 thyroiditis (61,63-67). However, additional data are needed to clarify whether there is a link
340 between vitamin D status and Hashimoto's thyroiditis and whether vitamin D
341 supplementation might reduce the risk of Hashimoto's thyroiditis.

Vitamin D and Graves' Disease

343 Graves' disease is an autoimmune thyroid disorder in which TSH receptor autoantibodies
344 cause hyperthyroidism. Given the increasing interest in the role of vitamin D role in
345 determining susceptibility to autoimmune diseases, it has been hypothesized that Graves'
346 disease may also be affected by vitamin D, based upon its ability to modulate the immune
347 system by suppressing the proliferation of activated T cells and enhancing the phagocytic
348 ability of macrophages (68,69).
349
350 Polymorphism in the VDR gene and vitamin binding protein gene have been reported to be
351 associated to Graves' disease's etiology (70,71) probably via a reduction in vitamin D
352 function, which may have an inhibitory effect on regulatory steps within the immune system.
353 The reported effects appear to differ markedly among different ethnicities, e.g. ApaI, BsmI
354 and FokI polymorphisms in the VDR gene are associated with highersusceptibility to Graves'
355 disease in Asian populations, but do not appear to play a role in Caucasian population (72).
356 Further, Feng et al. (73), recently reported that BsmI and TaqI polymorphisms are
357 significantly associated with autoimmune thyroid disorder risk, while the ApaI or FokI
358 polymorphisms are not Women with new onset Graves' disease have decreased 25(OH)D concentration, which is
361 also associated with thyroid volume measured by ultrasonography (74). Furthermore, it has
362 been reported that 25(OH)D concentration is higher in patients who achieve remission
363 compared to those who do not (75). The current evidence to support a role of vitamin D in
364 Graves' disease is preliminary but is worth investigating further in observational and
365 intervention studies.
366

Vitamin D and polycystic ovary syndrome (PCOS)

368
369 Accumulating evidence from several studies suggest that vitamin D may be involved in
370 several features of PCOS, such as infertility, hirsutism, insulin resistance and cardiovascular
371 risk (76,77). Wehr et al reported that women with normal ovulation had higher vitamin D
372 levels than women with PCOS (77). In addition, 25(OH)D deficiency was found to be
373 associated with lower rates of follicle development and pregnancy after stimulation in PCOS
374 women (78). Vitamin D supplementation may improve reproductive function in women with
375 PCOS by restoring normal menstrual cycles (79,80). Women with PCOS and hirsutism have
376 lower 25(OH)D levels than BMI matched controls (77, 81), which may be explained by an
377 association of vitamin D with androgens or SHBG (82,83). Vitamin D deficiency seems to
378 also have an impact on insulin sensitivity in PCOS women, as evaluated by HOMA-IR
379 (76,77,80). However, a more accurate evaluation of insulin sensitivity by hyperinsulinemic
380 euglycemic clamp in PCOS women did not confirm such an association (84). In addition to
381 insulin resistance, vitamin D deficiency in PCOS women has been associated with
382 cardiovascular risk factors, such as high total cholesterol, systolic and diastolic blood
383 pressure, C reactive proteins and triglycerides, and low HDL cholesterol (77). Small
384 uncontrolled intervention studies of vitamin D supplementation in women with PCOS have
385 shown improvements in fasting and stimulated glucose and dyslipidemia (triglycerides and
386 HDL) (80,85).
387 An inverse association between vitamin D status and metabolic and hormonal disturbances
388 has been reported in PCOS. However, due to the variability of the PCOS phenotype and the
389 heterogeneity of available studies, it is difficult to draw any conclusions. Ongoing randomized
390 trials in well-defined populations will help define the role of vitamin D in PCOS (Table 1).
391

Limitations in the study of vitamin D

393
394 The inverse association between vitamin D status and endocrine disease in observational
395 studies may be confounded by several factors. Most importantly, good vitamin D status is
396 generally a marker of good health, as high 25(OH)D concentration is associated with young
397 age, normal body weight and a healthy lifestyle, including good dietary and exercise habits.
398 Similarly, a lower vitamin D status may reflect chronic illness, which prevents outdoor
399 activities and sun exposure. Importantly, vitamin D is rarely ingested in isolation, more often,
400 it is ingested as part of a specific food (e.g. milk), a food group (e.g. dairy) or as part of a
401 health dietary pattern (e.g. Mediterranean diet). Therefore, additional nutrients co-ingested
402 with vitamin D (e.g. fish or fortified dairy products) may have independent or synergistic
403 effects on cardiometabolic disease or, alternatively, foods rich in vitamin D may replace other
404 foods that increase risk of cardiometabolic disease (e.g. fortified milk replacing soda). Nearly
405 all available observational studies used single measurements of blood 25(OH)D as a proxy of
406 vitamin D status, which may not reflect vitamin D status over long periods as risk factors for
407 vitamin D deficiency increase with time (aging, declining physical activity, etc). Therefore,
408 inaccurate assessment of the exposure (vitamin D status) and uncontrolled or residual
409 confounding may explain the results of the observational studies, which needs to be
410 confirmed in controlled trials. An additional challenge in the study of vitamin D is that there
411 is no consensus on the 25(OH)D thresholds for vitamin D adequacy.
412

Conclusions

415 Several observational studies have reported an association of low 25(OH)D concentration
416 with endocrine diseases. However, due to paucity of intervention studies, a causal link
417 between vitamin D deficiency and endocrine diseases is far from proven, thus no guidance
418 can be provided for or against recommending vitamin D supplementation for prevention or
419 therapy of endocrine conditions, outside of the current recommendations by the Institute of
420 Medicine for the general populations (600 to 800 IU/day depending on age and gender).
421 Ongoing and future trials are expected to provide answers to whether vitamin D
422 supplementation holds promise for endocrine health and disease.

425 Source of Funding: By research grants R01DK76092 and U01DK098245 (to A. G. Pittas)
426 from the National Institute of Diabetes and Digestive and Kidney Disease, the Office of the
427 Director, National Institutes of Health, and the National Institutes of Health Office of
428 Dietary Supplements and 1-14-D2d-01 (to A.G. Pittas) from the American Diabetes Association.

Summary tables are in PDF