Arch Osteoporos (2013) 8:124 DOI 10.1007/s11657-013-0124-5

ORIGINAL ARTICLE

Vitamin D insufficiency together with high serum levels of vitamin A increases the risk for osteoporosis in postmenopausal women J. M. Mata-Granados & J. R. Cuenca-Acevedo & M. D. Luque de Castro & M. F. Holick & J. M. Quesada-Gómez

Received: 23 December 2011 / Accepted: 7 January 2013 / Published online: 16 February 2013 # International Osteoporosis Foundation and National Osteoporosis Foundation 2013

Abstract Summary Postmenopausal women who were vitamin D deficient and had high serum levels of retinol had an eight times higher risk of having osteoporosis. A high retinol level together with vitamin D deficiency/insufficiency is an additional risk factor for osteoporosis. Purpose The aim of this study was to evaluate the association between vitamin D deficiency/insufficiency and excess of vitamin A intake as an osteoporosis risk factor in healthy postmenopausal women Design The design is a cross-sectional study of 232 healthy postmenopausal women. Methods Bone mass was evaluated by dual energy X-ray absorptiometry. Serum calcium, albumin phosphorus, creatinine, total high-density lipoprotein (HDL) and low-density lipoprotein (LDL) cholesterol, and triglycerides analyzed by standard methods and retinol and 25-hydroxyvitamin D [25 (OH)D] measured by an online solid-phase extraction coupled with high-pressure liquid chromatography–ultraviolet detection. J. M. Mata-Granados : J. M. Quesada-Gómez

Results Prevalence of vitamin D deficiency [25(OH)D< 20 ng/mL] was 70.1 %; 14.3 % had a 25(OH)D<10 ng/mL, and 23.6 % had insufficiency [25(OH)D 21–29 ng/mL]. Prevalence of high serum levels of retinol (>80 μg/dL) was 36.4 %. Among subjects with 25(OH)D <20 ng/mL (n = 152), 60.4 % (n= 92) had serum levels of retinol > 80 μg/dL. Bone density measurements revealed that the risk of osteoporosis was ~8 times higher in women with the highest retinol levels, as compared with women with the lowest retinol levels. In women with 25(OH)D<20 ng/mL, the risk for osteoporosis increased substantially in women who had the highest blood levels of retinol compared to the women with lowest retinol levels. Conclusions Higher retinol levels together with vitamin D deficiency could be a significant additional risk factor for osteoporosis, underscoring the need for improved physician and public education regarding optimization of vitamin D status in postmenopausal women and developing policies to avoid high serum levels of vitamin A. Keywords Osteoporosis . Vitamin D . Vitamin A . Bone mineral density . 25-Hydroxyvitamin D

Department of I+D+I, Sanyres Group, Córdoba, Spain J. M. Mata-Granados : M. D. Luque de Castro Department of Analytical Chemistry, University of Córdoba, Córdoba, Spain J. R. Cuenca-Acevedo : J. M. Quesada-Gómez Mineral Metabolism Unit RETICEF, Endocrinology, Reina Sofía Hospital, Córdoba, Spain M. F. Holick (*) Vitamin D Laboratory, Section of Endocrinology, Diabetes, and Nutrition, Department of Medicine, Boston University Medical Center, Boston, MA 02118, USA e-mail: [email protected]

Introduction Vitamin D plays a critical role in bone metabolism [1]. It has been recognized that vitamin D deficiency when prolonged and severe can lead to osteomalacia, characterized by insufficient mineralization of the newly formed bone matrix: the osteoid. Vitamin D deficiency also causes secondary hyperparathyroidism, increased bone turnover, bone loss, and low bone mineral density (BMD) [1, 2]. In recent years, vitamin D deficiency has been associated with loss of muscle mass,

124, Page 2 of 8

muscle weakness, and falls [3]. Therefore, vitamin D deficiency is a major determinant for osteoporosis fracture risk [1, 2]. There continues to be debate about the optimal blood level of 25-hydroxyvitamin [25(OH)D]. The Institute of Medicine defined vitamin D deficiency as a 25(OH)D< 20 ng/mL as it relates to bone health [4]. The consensus of the Endocrine Society’s Endocrine Practice Guidelines on Vitamin D also defined vitamin D deficiency as <20 ng/mL and defined vitamin D insufficiency as 21–29 ng/mL and vitamin D sufficiency as a 25(OH)D serum level of 30– 100 ng/mL [5]. Vitamin D deficiency is recognized to be a worldwide problem. More than 60 % of postmenopausal women have deficient 25(OH)D serum levels [6, 7], even in sunny countries including countries in the Middle East, Australia [1], and Spain [8], and it is a public health concern. It is also known that an excess of retinol intake can result in adverse skeletal effects. An excess of vitamin A both in vitro and in vivo suppresses osteoblast activity and promotes osteoclast formation thereby stimulating bone resorption and inhibiting bone formation. Vitamin A has been reported to antagonize some actions of vitamin D and potentiate the vitamin D deficiency/insufficiency effects on the skeleton [9–14]. Thus, excess vitamin A intake has the potential to accelerate bone loss, decrease bone mineral density, and increase fracture risk. Several observational reports have suggested that longterm intake of a diet high in vitamin A [15–18] and high serum retinol levels [18, 19] may contribute to the development of osteoporosis and increase osteoporosis fracture risk. Normal reported serum retinol levels seem to be highly regulated within a range of 0.7 to 2.8 μmol/L. Severe vitamin A deficiency is <0.35 μmol/L, and moderate deficiency is from 0.35 to 0.7 μmol/L. Serum levels above 2.4 μmol/L of vitamin A are considered to increase fracture risk [20–23]. Recently, it was reported that 50 % of Spaniards who had a serum 25(OH)D < 20 ng/mL had high levels of retinol (>68.6 μg/dL) [20]. Therefore, we assessed the serum 25 (OH)D and retinol levels and their impact on BMD in a cross-sectional study of 232 postmenopausal Spanish women.

Methods and materials Subjects A total of 232 healthy postmenopausal Spanish women from a Junta de Andalucía population screening program for breast cancer were recruited after giving informed consent. The study was approved by the Investigation Committee of the “Reina Sofía” Hospital (Córdoba) Spain. All the women were Caucasian, white, ambulatory, and in general good health. They did not have renal, hepatic, gastrointestinal,

Arch Osteoporos (2013) 8:124

or thyroid diseases or any other secondary causes for either low BMD or low serum fat soluble vitamin levels. None of them had been treated with calcium or vitamin supplements, hormone replacement or antiresorptive therapy, nor did they take thiazides, steroids, or other medications that might affect BMD and vitamin A levels. The women were recruited between April 1st and May 30th, in order to minimize the impact of seasonal changes of vitamin D status on bone mineral density [21], the dual energy X-ray absorptiometry (DXA) measurements and blood collections were done in the same week. Venous blood samples were drawn from an antecubital vein—without stasis in the morning after an overnight fast. After collection, the samples were centrifuged and the supernatants were collected. An aliquot was immediately used for biochemical measurements, and several fractions were frozen at −80 °C until use. Measurements Osteoporosis risk factors were assessed by a detailed autoadministered questionnaire [22]. Morning fasting samples of venous blood were taken, and serum calcium, albumin phosphorus, creatinine, total high-density lipoprotein (HDL) and LDL cholesterol, and triglycerides were analyzed by standard methods as well as the samples were divided into aliquots and frozen at −80 °C until analysis. Vitamin A, measured as retinol and 25(OH)D, were quantified by a method based on automatic online solid-phase extraction system for cleanup/preconcentration coupled with high-pressure liquid chromatography and ultraviolet detection as previously described [23]. The intraassay coefficients of variation (CVs) of retinol and 25(OH)D are 1.06 and 0.83 % and the inter-assay CV 2.4 and 1.8 %, respectively [23]. The methods were validated using standard reference material (SRM 968c, National Institute of Standards and Technology, Gaithersburg MD, USA). SRM 968c is used to validate methods that determine fat-soluble vitamins, carotenoids, and cholesterol in human serum or plasma. SRM 968c provides certified concentration values for alltrans retinol (vitamin A), α-tocopherol (vitamin E), and reference concentration values for 25(OH)D3. We also participate in Vitamin D External Quality Assessment Scheme of Charing Cross Hospital of London. Bone mineral density was evaluated by DXA performed in the lumbar spine (L1–L4) (LS) and total hip (TH) using a Hologic® QDR 1000 bone densitometer (Waltham, MA, USA). A sample of 1,331 women was used to establish the mean peak of BMD in healthy Spanish women and to calculate the T-score [24]. According to the World Health Organization (WHO) criteria, we considered that osteoporosis was present when the T-score measured at the LS or TH was equal to (or less than) −2.5 and osteopenia (low bone mass) as a T-score −1 to −2.5. [25].

Arch Osteoporos (2013) 8:124

Statistical analyses Descriptive statistics are presented as means± SD or as percentages. Serum retinol levels were evaluated both as a continuous variable and as a categorical variable, in quintiles. Separate analyses were performed for osteoporosis using the WHO criteria (T-score≤−2.5). Group means were compared using the two-tailed nonpaired Student’s test and one-factor analysis of variance. The level of significance comparing two means (using a Student t test) was set at the 0.05 confidence level, though results with p<0.01 were also identified. Chi-square test at 95 % of confidence level was used in the logistic model to calculate the association between each parameter and response variable. We considered two separate categorical models: a univariate and a multivariate logistic model. Age, age of menopause, body mass index (BMI), serum calcium, 25(OH)D, weight, and height were included as continuous variables in the multivariate model. Smoking habit at baseline was categorized as no-consumption or consumption. Alcohol consumption at baseline was categorized as no-consumption, consumption less than 1 day for week, between 2 or 3 days for week, 5 days for week, or daily. Physical activity was evaluated in three categories: walking less than 30 min per day, between 30 and 60 min per day, and more than 1 h per day. General linear multivariate models were considered to evaluate the influence of these variables on bone mineral density as continuous variable. Analysis of variance with lack of fit test was used to determine whether the model is adequate to describe the observed data. Statistical significance was defined as a p value less than 0.05. Both, in logistic regression and continuous multivariate models, the estimates remained similar when also included total, HDL and LDL cholesterol, triglycerides, creatinine, calcium, phosphate, and 24-h urinary calcium. Consequently, these variables were omitted from the reported analyses. Statistical analyses were carried out using the Statgraphics Plus software version 5.10 from Manugistics (Rockville, MD, USA).

Results The characteristics of the subjects are shown in Table 1. According to the WHO criteria [24], 78 patients (37.7 %) were osteoporotic (T-score≤−2.5), 125 patients (49.2 %) were considered low bone mass (osteopenic; T-score −1 to −2.5), and 19 patients (13.1 %) had a normal BMD. The average and standard deviation of bone mineral density from Spain reference population used to calculate the Tscore in this study was not statistically different of National Health and Nutrition Examination Survey III. Therefore, there is not any misclassification using the WHO T-score

Page 3 of 8, 124

cutoff for osteoporosis. The prevalence of vitamin D deficiency [25(OH)D<20 ng/mL] was 70.1 %, of which 14.3 % had a 25(OH)D <10 ng/mL, 23.6 % were insufficient with a 25(OH)D of 20.1–29.9 ng/mL, and only 29.9 % had a level of 25(OH)D >30 ng/mL. The prevalence of high serum levels of retinol (higher 80 μg/dL) was 36.4 %. In the group with normal serum levels of 25(OH)D, 41.3 % had high serum levels of retinol, and in the group with serum levels of 25(OH)D between 20.1 and 29.9 ng/mL, 44.3 % had high levels of retinol. Among subjects with 25(OH)D below, 20 ng/mL (n=152), 60.4 % (n=92) had serum levels of retinol higher than 80 μg/dL. Correlations between TH BMD and age (r=−0.435; p< 0.00001) and BMI (r=0.155; p=0.01) were found. Correlations between LS BMD and age (r=−0.269; p<0.0001) and body mass index (r=0.210; p=0.001) were also observed. There were significant linear correlations between 25(OH)D and LS BMD (r=0.212; p=0.002), TH BMD (r=0.148; p= 0.02), retinol (r=0.180; p=0.008), and calcium (r=−0.222; p=0.001). The mean serum retinol levels (95.2 μg/dL) were higher in osteoporotic postmenopausal women compared to postmenopausal women who had osteopenia and a normal BMD (86.9 μg/dL) (p<0.05). The mean serum level of 25(OH)D was lower in the osteoporotic group (18.9 ng/mL) compared to women who were osteopenic or had a normal BMD (21.0 ng/mL)(p<0.01). Multivariate logistic regression A multivariate logistic regression analysis revealed that the serum 25(OH)D level (p=0.001), age (p=0.0001), height (p= 0.0003), serum retinol level in quintiles (p<0.001), and Table 1 Characteristics of the postmenopausal women under study Parameter Age (years) Age menopause (years) Age of menarche (years) Weight (kg) Height (cm) Body mass index (kg/m2) Phosphate (mg/dL) Calcium (mg/dL) Creatinine (mg/dL) Total cholesterol (mg/dL) LDL cholesterol (mg/dL) HDL cholesterol (mg/dL) Triglycerides (mg/dL) 25(OH)D (ng/mL) Retinol (μg/dL)

Total group (n=229) 57.4±6.4 47.9±5.3 12.8±1.5 72.2±11.8 154.6±6.8 30.25±5.04 3.6±0.5 10.0±0.4 0.9±0.1 224±35 140±31 61±13 112±57 17.5±7.2 77.2±24.6

124, Page 4 of 8

Arch Osteoporos (2013) 8:124 P<0.001

a T-score LS BMD

alcohol consumption (p=0.04) were associated with osteoporosis. Odds ratios were 0.94 (CI 0.91–0.98), 1.11 (CI 1.05– 1.17), 0.91 (CI 0.86–0.96), and 4.37 (CI 1.35–14.22) for 25 (OH)D, age, height, and daily alcohol consumption, respectively. The increment of risk of osteoporosis was up to eight times higher in osteoporotic women who had the highest retinol levels (96.1–161.6 μg/dL), as compared with postmenopausal women who had the lowest retinol levels (33.5– 56.6 μg/dL) and odds ratio 8.37 (CI 2.51–27.91). Multivariate analysis of variance

P<0.001 -0,7

P<0.05

-1,1 -1,5 -1,9 -2,3 -2,7 -3,1 1

2

3

4

5

4

5

Quintiles of retinol P<0.001 P<0.01

b T-score Hip BMD

Multivariate analysis of variance (MANOVA) in a multivariate linear model for bone mineral density as continuous variable showed a statistically significant effect of serum retinol levels for both lumbar spine and total hip. The complete models for LS BMD (R=0.575, p=0.0000) and TH BMD (R=0.675, p=0.0000) are shown in Table 2. LS BMD of lumbar spine was decreased in the highest quintile, as shown in Fig. 1a, and a similar pattern to LS BMD model in relation with quintiles of serum retinol is shown for total hip model (Fig. 1b). The Fisher’s least significant difference (lsd) procedure was used to discriminate among the means of quintiles for retinol for a 95 % confidence level. Statistically significant differences were found for lumbar spine bone mineral density between first quintile (−1.38±0.23) and second (−1.81±0.18), fourth (−2.29±0.17), and fifth (−2.62±0.30) quintiles of retinol levels. For the total hip bone mineral density, statistically significant differences were observed between the lower quintile (−0.52±0.18) and third (−0.95±0.10), fourth (−0.96± 0.11), and fifth (−1.62±0.23) quintiles of retinol levels.

P<0.01 -0,2 -0,5 -0,8 -1,1 -1,4 -1,7 -2 1

2

3

Quintiles of retinol Fig. 1 Effect of serum retinol quintiles on BMD T-score. a Lumbar spine, b total hip. Fisher’s least significant difference procedure was used to discriminate among the means of retinol’s quintiles

Serum 25(OH)D and retinol levels Among the patients with serum 25(OH)D levels<20 ng/mL, those with a higher level of retinol (≥80 μg/dL) had a lower Table 2 Multivariate linear regression analysis for lumbar spine and hip bone mineral density LS BMD

Retinol in quintiles Physical activity Height Weight 25(OH)D Body mass index Age Age of menopause Age of menarche

Hip BMD

F ratio

p value

F ratio

p value

2.9 5.1 11.3

0.0233 0.0020 0.0009

3.0

0.0186

16.4 18.3 21.4 6.5

0.0001 0.0000 0.0000 0.0114

47.9 5.9

0.0000 0.0158

65.3 16.1 7.7

0.0000 0.0001 0.0060

bone mineral density in both LS (p<0.05) and TH BMD (p< 0.0001). There were significant univariate linear correlations between serum retinol and TH BMD (r=−0.177; p=0.02) and serum calcium (r=−0.209; p=0.009). The serum PTH level (p<0.0001), age of menarche (p= 0.0004), height (p=0.0003), weight (p=0.001), bone alkaline phosphatase (p=0.0002), serum retinol levels (p=0.04), and BMI (p=0.001) were associated with osteoporosis in multivariate logistic regression for the patients with 25 (OH)D<20 ng/mL. Odd ratios were 1.10 (CI 1.04–1.16), 1.85 (CI 1.27–2.69), 0.38 (CI 0.20–0.72), 2.50 (CI 1.26– 4.99), 1.13 (CI 1.04–1.22), and 0.09 (CI 0.015–0.48) for PTH, age of menarche, height, weight, bone alkaline phosphatase, and BMI, respectively. The increment of risk for osteoporosis for this group was up to 92 times higher for each quintile, as compared with the lowest quintile (odd ratios are shown in Table 3). The complete MANOVA models for LS BMD (r=0.622, p=0.0000) and TH BMD (r=0.633, p=0.0000) for patients with 25(OH)D < 20 ng/mL are shown in Table 4. The

Arch Osteoporos (2013) 8:124 Table 3 Odd ratios for osteoporosis in population with serum levels of 25(OH)D<20 ng/mL, according to the baseline serum retinol level

Page 5 of 8, 124

Retinol quintile

1 2 3 4 5

Median retinol level

(41.8–68.4 μg/dL) (68.6–79.2 μg/dL) (79.8–93 μg/dL) (93.5–115 μg/dL) (115.3–201.9 μg/dL)

Fisher’s lsd procedure was used to discriminate among the means of quintile of retinol at least for 95 % of confidence level. Statistically significant differences (p<0.05) were found for bone mineral density in lumbar spine between first quintile (−0.91±0.27) and second (−2.40±0.23), fourth (−2.67±0.19), and fifth (−2.53±0.41) quintiles of retinol levels and for bone mineral density in total hip were between second quintile (−0.65±0.16) and fourth (−1.15± 0.14) and fifth (−1.16±0.23) quintile of retinol.

Discussion The current prospective, population-based cohort study of healthy postmenopausal women showed that higher serum retinol levels together with vitamin D deficiency were a significant risk factor for osteoporosis. We observed that vitamin D deficiency, 25(OH)D <20 ng/mL, was present in 70 % of our healthy ambulatory population (14.3 %<10 ng/mL), without secondary causes or medications that might affect 25(OH)D levels. These data are consistent with a cross-sectional observational study conducted from the north to the south across Spain, which reported that 63 % of postmenopausal women receiving therapy for osteoporosis and 76 % not receiving treatment had 25(OH)D levels<30 ng/mL [8], similar to other reports around the world [1, 6, 7, 25]. The high prevalence of vitamin D deficiency in this study was consistent across all age

Table 4 Multivariate linear regression analysis for lumbar spine and hip bone mineral density for patients with 25(OH)D levels<20 ng/mL LS BMD

Retinol in quintiles Height Body mass index Age Age of menarche PTH Osteocalcin

Hip BMD

F ratio

p value

F ratio

p value

4.1 13.0 17.9 12.2 5.3 9.9 4.3

0.038 0.0004 0.0000 0.0007 0.023 0.002 0.042

2.0 8.7 31.0 17.7

0.081 0.003 0.000 0.000

11.4 4.3

0.001 0.04

Osteoporosis

μg/dL

No. of women

(95 % CI)

62.9 72.4 85.8 101.2 134.1

29 31 32 31 30

1.00 1.1 (0.3–4.4) 5.1 (1.1–23.9) 24.9 (4.1–153.9) 92.4 (11.1–767.9)

groups and the geographic regions previously studied in Spain and in agreement with other southern European countries that showed a higher prevalence of vitamin D deficiency than in central or northern countries [26]. This paradox could be partially explained because vitamin D intake is generally much lower in southern countries (<200 IU). Moreover, it is taken for granted by the European Mediterranean population that casual exposure to sunlight provides enough vitamin D, while skin synthesis of vitamin D may not compensate for the low nutritional intake because almost all the European Mediterranean countries are located at a high latitude above approximately 37°N where the photosynthesis of previtamin D3 is markedly diminished or absent in late fall, winter, and early spring [1]. The misconception among southern European populations and clinicians in the entire world that diet and sunlight provide enough vitamin D has led to a decline in promoting vitamin D supplementation for the prevention and treatment of osteoporosis [27, 28]. We also observed elevated serum retinol levels in 36 % of our postmenopausal women, which is in agreement with the results of previous prospective evaluations in Spaniards [20, 29]. Moreover, these data are consistent with the high intake of vitamin A reported by a Spain nutritional survey that analyzed the diet and the dietary food habits in a sample of 1,218 Spanish women. A high percentage of Spanish postmenopausal women consumed fortified food (52.7 %) and micronutrient supplements (22.8 %) on a regular basis. Although habitual diets provided enough vitamins and minerals, vitamin A intakes in postmenopausal women were 1,804±1,023 μg/day [30], higher than the 700-μg/day recommended dietary allowances proposal by the Food and Nutrition Board of the Institute of Medicine in 2001. This amount was proposed as needed to ensure adequate stores of vitamin A in the body to support normal reproductive function, immune function, vision, and other health benefits [31]. The intake and serum levels of vitamin A in Spain are almost as high as in Scandinavia [32]. This can be explained because in Spain, in addition to high intake of fatty fish, liver, dairy products, and fortified foods including margarine and milk that contain retinol, other dietary sources of retinoids include carotenoids obtained from fruits and vegetables that are about 33 % equivalent in retinol activity. Over the last several

124, Page 6 of 8

decades, there have been important changes in carotenoid intake from fruit and vegetables in the Spanish population. Although the consumption of fruit and vegetables in Spain is still consistent with a Mediterranean-type pattern, modifications in the consumption of individual fruits and vegetables have caused changes in total and specific carotenoids intake with a considerable increase in α- and β-carotene, together with a decrease in lutein and zeaxanthin intake [33]. These data have significance because the consumption of large amounts of vitamin A and carotenoids can explain the high retinol serum levels found in our population. Although current data are limited, the consumption of large amounts of vitamin A may be associated with a decreased bone mineral density and an increased fracture risk, especially among population with the highest serum retinol levels [15, 19]. Vitamin A is a generic term for a large number of related compounds. The major vitamin A sources in the human diet derives from two major sources, a vitamin A as retinol, retinal, retinoic acid, and retinyl esters (REs) from foods of animal origin or in supplements and provitamin A carotenoids, such as β-carotene, α-carotene, and β-cryptoxanthin, found in plant-derived foods. The intestinal absorption and metabolism of dietary carotenoids has been the subject of several reviews. Carotenoids either are cleaved to generate retinol or absorbed intact; however, REs are completely hydrolyzed in the intestinal lumen, and free retinol is taken up by enterocytes [34]. For these reasons, there are some disagreement on what is the adequate indicator of vitamin A status and its relationship to bone density. In some studies with serum REs, no association was found between this parameter and BMD [35], but only a RE marginal association with osteoporotic postmenopausal women [36]. We chose retinol to evaluate vitamin A status because serum retinyl esters could reflect only a temporary excess rather than long-term vitamin A intake and high level of storage [37]. The osteoporosis risk in this study was higher among postmenopausal women with the highest serum retinol levels. The risk was especially high in women with vitamin D deficiency. Since BMD is the best predictor of risk of osteoporotic fractures, our data support previous reports indicating that hypervitaminosis A, even subclinical, could increase the risk of fracture [15, 19]. High retinol levels affect bone and mineral metabolism through retinoic acid effects, which stimulate osteoclast formation and activity, leading to increased bone resorption, and suppressed osteoblast activity [13]. The deficiency of vitamin D has long been recognized as a major determinant of osteoporosis [38] and fracture risk [39]. But it is important to highlight that an excess of vitamin A reduces the efficacy of vitamin D and potentiates the vitamin D deficiency–insufficiency effects [40]. Accordingly, an adequate bone mineral density may depend on an optimal ratio of retinol to 25(OH)D levels as a consequence of a suitable ratio of vitamin A to vitamin D intake [41].

Arch Osteoporos (2013) 8:124

The hormonally active metabolite of vitamin D (1,25-dihydroxyvitamin D [1,25(OH)2D]) exerts its multiple effects on complexing with nuclear vitamin D receptor (VDR) and heterodimerizes with nuclear retinoid X receptor (RXR) to form VDR/RXR that binds to vitamin D response elements on the promoters of vitamin D responsive genes [1]. Vitamin A is absorbed, then transported by chylomicrons to the liver, and mobilized from liver bound to retinol binding protein or transthyretin; then, after release in tissue target, retinol is oxidized to retinal and to retinoic acid. All-trans retinoic acid exerts its biological effects when bound to any one of several members of the retinoic acid receptor (RAR)/RXR nuclear transcription factor family. Found in nearly every cell, RAR/RXR transcription factors can increase or decrease gene expression by binding to specific DNA response elements [33]. Since RXR are partners for both VDR and RAR, among others, it is possible that increased tissue levels of retinoic acid bind RXR and RAR, thus limiting the availability of RXR to bind to the 1,25(OH)2D–VDR complex. Thus, an increase in serum retinol levels, together with vitamin D deficiency, could be detrimental for bone and mineral metabolism. Alternatively, excessive retinol availability may have toxic effects on osteoblastic function resulting in a decrease in bone mineral density independently of vitamin D status. Therefore, inquiring about vitamin A and carotenoid intake on a measurement of retinol and 25(OH)D should be considered for the evaluation of causes of osteoporosis in postmenopausal women. We must underscore the need for improved physician and public education regarding optimization of vitamin D status in postmenopausal women and potentially to develop policies to avoid high levels of vitamin A intake. Since the therapeutic window for vitamin A is narrow and nutritional supplements contribute significantly to vitamin A intake, the amounts of retinol in fortified foods and vitamin A supplements, currently considered safe, might be harmful. This situation can specially occur in Western countries, where there is a high life expectancy and the prevalence of osteoporosis is increasing. Therefore, these amounts of retinol for health need to be reassessed. Likewise, an evaluation of the possibility of alimentary use of xanthophylls carotenoids as lutein, zeaxanthin instead of retinol as antioxidants should be considered. Limitations of our study include those associated with crosssectional studies; all the subjects studied were attended in our hospital for screening of breast cancer, so our sample could be no representative of the postmenopausal Spanish general population. Moreover, we did not evaluate the osteoporosis fracture risk. Further research is needed to determine the relationship between retinol, 25(OH)D status, and fracture risk. Acknowledgments This study was supported by grants CS 0200/ 2008, P06-FQM-01515, and SAF2005-05254 and teams PAI CTS-413 and PAI FQM-227 of Junta de Andalucía and Sanyres 21, Córdoba (Spain).

Arch Osteoporos (2013) 8:124 Conflicts of interest JMMG, JRCA, MDLC, and JMQG have no conflict of interest. MFH is a consultant for Merck, Sanofi-Aventis, P&G, Quest Diagnostics, and Novartis.

References 1. Holick MF (2007) Vitamin D deficiency. N Engl J Med 357:266– 281 2. Lips P (2001) Vitamin D deficiency and secondary hyperparathyroidism in the elderly: consequences for bone loss and fractures and therapeutic implications. Endocr Rev 22:477–501 3. Bischoff-Ferrari HA, Giovannucci E, Willett WC, Dietrich T, Dawson-Hughes B (2006) Estimation of optimal serum concentrations of 25-hydroxyvitamin D for multiple health outcomes. Am J Clin Nutr 84:18–28 4. IOM (Institute of Medicine) (2011) Dietary reference intakes for calcium and vitamin D. Institute of Medicine Committee to Review Dietary Reference Intakes for Calcium and Vitamin D. The National Academies, Washington, DC 5. Holick MF, Binkley NC, Bischoff-Ferrari HA, Gordon CM, Hanley DA, Heaney RP, Murad MH, Weaver CM (2011) Evaluation, treatment & prevention of vitamin d deficiency: an endocrine society clinical practice guideline. J Clin Endocrinol Metab 96(7):1911–1930 6. Holick MF, Siris ES, Binkley N, Beard MK, Khan A, Katzer JT et al (2005) Prevalence of vitamin D inadequacy among postmenopausal North American women receiving osteoporosis therapy. J Clin Endocrinol Metab 90:3215–3224 7. Lips P, Hosking D, Lippuner K, Norquist JM, Eehreb L, Maalouf G et al (2006) The prevalence of vitamin D inadequacy amongst women with osteoporosis: an international epidemiological investigation. J Intern Med 260:245–254 8. Quesada Gómez JM, Mata Granados JM, Delgadillo J, Ramírez R (2007) Low calcium intake and insufficient serum vitamin D status in treated and non-treated postmenopausal osteoporotic women in Spain. J Bone Miner Metab 22:S309 9. Scheven BA, Hamilton NJ (1990) Retinoic acid and 1,25-dihydroxyvitamin D3 stimulate osteoclast formation by different mechanisms. Bone 11:53–59 10. Abu-Hijleh G, Padmanabhan R (1997) Retinoic acid-induced abnormal development of hind limb joints in the mouse. Eur J Morphol 35:327–336 11. Hough S, Avioli LV, Muir H, Gelderblom D, Jenkins G, Kurasi H et al (1988) Effects of hypervitaminosis A on the bone and mineral metabolism of the rat. Endocrinology 122:2933–2939 12. Moore T, Sharman IM (1979) Hypervitaminosis A combined with calcium deficiency in rats. Int J Vitam Nutr Res 49:14–20 13. Oreffo RO, Teti A, Triffitt JT, Francis MJ, Carano A, Zallone AZ (1988) Effect of vitamin A on bone resorption: evidence for direct stimulation of isolated chicken osteoclasts by retinol and retinoic acid. J Bone Miner Res 3:203–210 14. Clark L, Seawright AA, Gartner RJ (1971) Long bone abnormalities in kittens following vitamin A toxicosis in kittens. J Comp Pathol 81:365–371 15. Melhus H, Michaelsson K, Kindmark A, Reinhold B, Holmberg L, Hans M et al (1998) Excessive dietary intake of vitamin A is associated with reduced bone mineral density and increased risk for hip fracture. Ann Intern Med 129:770–778 16. Feskanich D, Singh V, Willett WC, Colditz GA (2002) Vitamin A intake and hip fractures among postmenopausal women. JAMA 287:47–54 17. Michaelsson K, Lithell H, Vessby B, Melhus H (2003) Serum retinol levels and the risk of fracture. N Engl J Med 348:287–294

Page 7 of 8, 124 18. Promislow JHE, Goodman-Gruen D, Slymen DJ, Barrett-Connor E (2002) Retinol intake and bone mineral density in the elderly: the Rancho-Bernardo study. J Bone Miner Res 17:1349–1358 19. Opotowsky AR, Bilezikian JP (2004) Serum vitamin A concentration and the risk of hip fracture among women 50 to 74 years old in the United States: a prospective analysis of the NHANES I followup study. Am J Med 117:169–174 20. Mata-Granados JM, Luque de Castro MD, Quesada Gómez JM (2008) Inappropriate serum levels of retinol, alpha-tocopherol, 25 hydroxyvitamin D3 and 24,25 dihydroxyvitamin D3 levels in healthy Spanish adults: simultaneous assessment by HPLC. Clin Biochem 41:676–680 21. Rosen CJ, Morrison A, Zhou H, Storm D, Hunter SJ, Musgrave K et al (1994) Elderly women in northern New England exhibit seasonal changes in bone mineral density and calciotropic hormones. Bone Miner 1994(25):83–92 22. O’Neill TW, Cooper C, Cannata JB, Díaz López JB, Hoszowski K, Johnell O et al (1994) Reproducibility of a questionnaire on risk factors for osteoporosis in a multicentre prevalence survey: the European Vertebral Osteoporosis Study. Int J Epidemiol 23:559– 565 23. Mata-Granados JM, Quesada Gómez JM, Luque de Castro MD (2009) Fully automatic method for the determination of fat soluble vitamins and vitamin D metabolites in serum. Clin Chim Acta 403:126–130 24. Diaz Curiel M, Carrasco de la Peña JL, Honorato Pérez J, Pérez Cano R, Rapado A, Ruiz Martínez I (1997) Study of BMD in lumbar spine and femoral neck in a Spanish population. Multicentre Research Project on Osteoporosis. Osteoporos Int 7:59–64 25. Kanis JA, Melton LJ, Christiansen C, Johnston CC, Khaltaev N (1994) The diagnosis of osteoporosis. J Bone Miner Res 9:17– 1141 26. Lips P, Duong T, Oleksik AM, Black D, Cummings S, Cox D et al (2001) A global study of vitamin D status and parathyroid function in postmenopausal women with osteoporosis: baseline data from the multiple outcomes of raloxifene evaluation clinical trial. J Clin Endocrinol Metab 86:1212–1221 27. Lips P (2007) Vitamin D status and nutrition in Europe and Asia. J Steroid Biochem Mol Biol 103:620–625 28. Jacobs ET, Alberts DS, Foote JA, Green SB, Hollis BW, Yu Z, Martínez ME (2008) Vitamin D insufficiency in southern Arizona. Am J Clin Nutr 87:608–613 29. Olmedilla B, Granado F, Southon S, Wright AJ, Blanco I, GilMartínez E et al (2001) Serum concentrations of carotenoids and vitamins A, E, and C in control subjects from five European countries. Br J Nutr 85:227–238 30. Úbeda N, Basagoiti M, Alonso-Aperte E, Varela-Moreiras G (2007) Hábitos alimentarios, estado nutricional y estilos de vida en una población de mujeres menopáusicas españolas. Nutr Hosp 22:313–321 31. Trumbo P, Yates AA, Schlicker S, Poos M (2001) Dietary reference intakes: vitamin A, vitamin K, arsenic, boron, chromium, copper, iodine, iron, manganese, molybdenum, nickel, silicon, vanadium, and zinc. J Am Diet Assoc 101:294–301 32. Cruz JA, Moreiras-Varela O, van Staveren WA, Trichopoulou A, Roszkowski W (1991) Intakes of vitamins and minerals. Eur J Clin Nutr 45:121–138 33. Granado F, Blázquez S, Olmedilla B (2007) Changes in carotenoid intake from fruit and vegetables in the Spanish population over the period 1964–2004. Public Health Nutr 10:1018–1023 34. Harrison EH (2005) Mechanisms of digestion and absorption of dietary vitamin A. Annu Rev Nutr 25:87–103 35. Ballew C, Galuska D, Gillespie C (2001) High serum retinyl esters are not associated with reduced bone mineral density in the Third National Health and Nutrition Examination Survey, 1988–1994. J Bone Miner Res 16:2306–2312

124, Page 8 of 8 36. Penniston KL, Weng N, Binkley N, Tanumihardjo SA (2006) Serum retinyl esters are not elevated in postmenopausal women with and without osteoporosis whose preformed vitamin A intakes are high. Am J Clin Nutr 84:1350–1356 37. Krasinski SD, Cohn JS, Russell RM, Schaefer EJ (1990) Postprandial plasma vitamin A metabolism in humans: a reassessment of the use of plasma retinyl esters as markers for intestinally derived chylomicrons and their remnants. Metabolism 39:357–365 38. Mezquita-Raya P, Muñoz-Torres M, Luna JD, Luna V, LópezRodríguez F, Torres-Vela E et al (2001) Relation between vitamin

Arch Osteoporos (2013) 8:124 D insufficiency, bone density, and bone metabolism in healthy postmenopausal women. Bone Miner Res 16:1408–1415 39. Quesada-Gómez JM, Alonso J, Bouillon R (1996) Vitamin D insufficiency as a determinant of hip fractures. Osteoporos Int 6:42–47 40. Metz AL, Walser MM, Olson WG (1985) The interaction of dietary vitamin A and vitamin D related to skeletal development in the turkey poult. J Nutr 115:929–935 41. Boucher BJ, Chandra RK, Melhus H, Michaelsson K (2003) Serum retinol levels and fracture risk. N Engl J Med 348:1927–1928