Abstract
HIV-infected individuals are at increased risk for several metabolic diseases, including low 25-hydroxyvitamin D [25(OH)D]. Data on the prevalence and risk factors for low 25(OH)D in HIV patients living in the tropics is scarce. Patients ≥ 40 years old on stable antiretroviral therapy were enrolled from March 2009 to July 2011 in Hawai‘i (latitude 21° North). Chemiluminescent immunoassay (DiaSorin) was used to determine plasma 25(OH)D levels. Patients were grouped by whether 25(OH)D was collected in summer (May 1 – September 30) or winter (October 1 – April 30). Of 158 patients enrolled, 88 (56%) and 70 (44%) were enrolled in winter and summer, respectively. There were 57.6% Caucasians and 88% men. Over-all median (quartile1, quartile3) age was 51 (46, 57) years and median 25(OH)D was 32.4 (24.0, 41.0) ng/ml. Forty-three percent (n=68) had 25(OH)D<30.0 ng/ml. Median 25(OH)D levels were 29.6 (22.0, 38.0) ng/ml in winter and 36.9 (25.0, 44.5) ng/ml in summer (P = .01). Median body mass index (BMI) of winter patients was significantly higher (P = .03). By simple linear regression, log-transformed 25(OH)D was significantly associated with winter visit (β = −.0737, P = .01), ethnicity (Caucasian versus non-Caucasian, β = .1194, P < .01), BMI (β = −.0111, P < .01) and current use of zidovudine (β = −.1233, P = .03). In multiple linear regression, only Caucasian ethnicity (β = .1004, P < .01) and BMI (β = −.0078, P = .02) retained statistical significance. Seasonal variation in 25(OH)D was observed but the significance of winter visit was not preserved in the final multivariate model. Ethnicity and BMI were better predictors of 25(OH)D levels than season in the tropics.
Keywords: vitamin D, Hawai‘i, 25-hydroxyvitamin D, HIV, season
Introduction
Low vitamin D is associated with poor health outcomes such as increased risk for falls and fractures, arterial dysfunction, and autoimmunity.1–4 Decreased levels have also been associated with increased mortality rates in some, but not all, studies.5–8 Aside from its role on calcium homeostasis and bone mineralization, vitamin D has also been shown to have protective effects against ultraviolet-induced DNA damage.9
HIV-infected individuals are known to be at risk for several metabolic diseases, including low serum 25-hydroxyvitamin D (25(OH)D).10 Antiretroviral therapy, particularly zidovudine and efavirenz, have been linked to low 25(OH)D levels.11,12 Decreased cutaneous vitamin D production from lack of sun exposure during long winter months is another risk factor that has been extensively studied among patients in high-latitude regions.13,14 However, data regarding predictors of 25(OH)D levels in HIV patients living in tropical areas is scarce.
Hawai‘i is located within the tropical zone (latitude 21° North) and, according to the National Weather Service Forecast Office, has two seasons—summer (between May and October) and winter (between October and April).15 We identified risk factors for low 25(OH)D and analyzed the influence of seasons on 25(OH)D levels using data obtained from patients enrolled into the Hawai‘i Aging with HIV - Cardiovascular (HAHC-CVD) Cohort, a longitudinal study of aging HIV-infected patients on potent antiretroviral therapy intended to examine the role of oxidative stress and inflammation on cardiovascular risk.
Patients and Methods
The current study is a cross-sectional analysis of the baseline data of patients enrolled into the HAHC-CVD Cohort. Entry criteria required patients to have documented HIV infection, be at least 40-years-old, and on stable antiretroviral therapy ≥ 6 months. Patients who were institutionalized or with active malignancy, infection, or AIDS-defining illness at the time of enrolment were excluded.
After informed consent, fasting blood samples were collected, stored in EDTA tubes, frozen at −140°C and forwarded to LipoScience Inc. (Raleigh, NC). Chemiluminescent immunoassay (DiaSorin) was used to determine 25(OH)D levels. Low 25(OH)D was operationally defined as having serum 25(OH)D<30.0 ng/ml, which includes both vitamin D insufficiency and deficiency as previously defined by The Endocrine Society, 2011.16 Weight and height were obtained in the clinic in triplicate during the entry visit and averaged.
Patients were grouped by whether 25(OH)D were collected in summer (from May 1 to September 30) or winter (from October 1 to April 30). Group comparisons of continuous variables, including 25(OH)D levels, were tested by two-sided Mann-Whitney test. Chi-square test was used to compare the proportion of patients having low 25(OH)D between seasons. After 25(OH)D levels were normalized by log-transformation and removal of one outlier, simple and multiple linear regression were conducted to identify factors associated with 25(OH)D analyzed as a continuous variable. Dichotomized at 30.0 ng/ml, factors associated with low 25(OH)D were identified using univariate and multivariate logistic regression models. STATA IC 12.0 (College Station, Texas) was used for the analyses. The study was approved by the Committee on Human Studies of the University of Hawai‘i.
Results
Patient Characteristics
Of 158 patients enrolled from March 2009 to July 2011, 88 (56%) and 70 (44%) were enrolled in winter and summer, respectively. The cohort consisted of 57.6% Caucasians, 12.6% Native Hawaiians of mixed or full ancestry, 8.2% Asians, 3.8% African-Americans, 1.9% Native American/Alaskan, and 15.9% of mixed (non-Hawaiian) or unknown ethnicity. The majority (88%) were males. Median age (quartile1, quartile3) was 51 (46, 57) years. Median nadir CD4+ T-lymphocyte count was 140.5 (27, 249) cells/mm3 and median CD4+ T-lymphocyte count at the time of enrolment was 498.5 (341, 661) cells/mm3. No significant difference in age, current CD4 count, or nadir CD4 count between summer and winter patients were found. On the other hand, body mass index (BMI) was significantly different between summer and winter patients (P = .03). Median BMI of winter patients was 26.4 (24.3, 29.8) kg/m2, while median BMI of summer patients was 25.5 (23.4, 27.1) kg/m2. Baseline characteristics of the HAHC-CVD cohort are summarized in Table 1.
Table 1.
Baseline clinical and demographic characteristics of patients enrolled into the Hawaii Aging with HIV-Cardiovascular Cohort Study.
| Summer Patients (May 1 to Sept. 30) n= 70 | Winter Patients (Oct. 1 to April 30) n=88 | P-value | |
| Gender | .23 | ||
| Male | 64 | 75 | |
| Female | 6 | 13 | |
| Ethnicity | .38 | ||
| Caucasian | 43 | 48 | |
| Non-Caucasian | 27 | 40 | |
| Body mass index (kg/m2) | .03* | ||
| Normal (18.5–24.9) | 31a | 32 | |
| Overweight (25.0–29.9) | 34 | 36 | |
| Obese (>30.0) | 5 | 20 | |
| Current use of efavirenz | .16 | ||
| Yes | 34 | 33 | |
| No | 36 | 55 | |
| Current use of zidovudine | .23b | ||
| Yes | 3 | 9 | |
| No | 67 | 79 | |
| 25-hydroxyvitamin D levels | .048* | ||
| <30 ng/ml | 24 | 44 | |
| >30 ng/ml | 46 | 44 |
One patient with BMI of 17.5 was combined into this cell count
P-value was based on Fisher's exact test.
Statistically significant
Group Comparisons of 25-hydroxyvitamin D Levels
Median 25(OH)D of the cohort was 32.4 (24.0, 41.0) ng/ml and mean was 34.0 ng/ml (standard deviation, 14.8). Forty-seven patients (29.8%) had 25(OH)D from 20.0 to 29.9 ng/ml, while 21 patients (13.3%) had 25(OH)D < 20.0 ng/ml. No patient was taking vitamin D as a prescribed medication; however, information regarding the use of over-the-counter vitamin D or multivitamins was not collected. When 25(OH)D was dichotomized at 30 ng/ml, a greater proportion of patients enrolled in winter had low 25(OH)D (P = .048).
Comparisons of median 25(OH)D levels by groups were conducted. By ethnicity, Caucasians had higher median 25(OH) D compared to non-Caucasians (35.9 ng/ml versus 25.0 ng/ml, P < .001). Obese patients (BMI ≥ 30 kg/m2) had significantly lower 25(OH)D levels compared to non-obese patients (28.3 ng/ml versus 34.0 ng/ml, P=.03). No significant difference in median 25(OH)D levels was found by drug groups (use of efavirenz or zidovudine).
By season of enrolment, patients enrolled in winter had significantly lower median 25(OH)D compared to patients enrolled in summer (29.6 ng/ml versus 36.9 ng/ml, P = .01) (Figure 1). Data was also analyzed by excluding patients enrolled after the end of winter (May) and towards the end of summer/beginning of winter (October) to eliminate patients whose 25(OH)D levels were potentially affected by the season immediately preceding enrolment. Excluding 24 patients who were enrolled in May and October, a significant difference in median 25(OH)D levels between seasons was still found (29.0 ng/ml in winter and 37.3 ng/ml in summer, P < .01).
Figure 1.
Comparison of 25-hydroxyvitamin D levels of HIV-infected patients enrolled in summer and winter in Hawai‘i.
Subgroup analyses of 25(OH)D levels stratified by ethnicity, BMI, and season are presented in Figure 2. Obese patients enrolled in summer had median 25(OH)D level of 31.4 ng/ml, while those enrolled in winter had median level of 25.0 ng/ml (P = .20). Non-obese patients enrolled in summer had median 25(OH)D level of 37.2 ng/ml, while those enrolled in winter had median level of 31.3 ng/ml (P = .04). The median 25(OH) D level of Caucasians enrolled in summer was 38.3 ng/ml versus 33.5 ng/ml among Caucasians enrolled in winter (P = .11). Among non-Caucasians enrolled in summer, median 25(OH) D was 28.6 ng/ml versus 24.9 ng/ml among non-Caucasians enrolled in winter (P = .13).
Figure 2a.
Median 25-hydroxyvitamin D of obese and non-obese patients stratified by season of enrollment.
Figure 2b.
Median 25-hydroxyvitamin D of Caucasian and non-Caucasian patients stratified by season of enrollment.
Factors Associated with 25-hydroxyvitamin D Levels
Linear regression was used to identify factors associated with 25(OH)D analyzed as a continuous variable. Serum 25(OH) D was not normally distributed and was log-transformed. By simple linear regression (Table 2), lower 25(OH)D was associated with winter visit (β = −.0737, P = .01), BMI (β = −.0111, P < .01), and current use of zidovudine (β = −.1233, P = .03). On the other hand, Caucasian ethnicity (β = .1194, P < .01) was associated with higher 25(OH)D levels. No significant association was found with age, gender, or current use of efavirenz. Beta-coefficients from log-transformed models were exponentiated to return an estimate of average change in 25(OH)D levels expressed in ng/ml. For significant variables in univariate linear regression, exponentiated β-coefficients are as follows: BMI, 0.9748; Caucasian ethnicity, 1.3164; winter visit, 0.8439; and current use of zidovudine, 0.7528.
Table 2.
Simple linear regression and univariate logistic regression of factors associated with 25-hydroxyvitamin D.
| Simple Linear Regression | Univariate Logistic Regression | |||||
| Beta coefficient | P-value | 95% CI | Odds Ratio | P-value | 95% CI | |
| Age | −.0015 | .44 | −.0053 to .0023 | 1.02 | .41 | .98 to 1.06 |
| Female gender | −.0148 | .75 | −.1055 to .0758 | .95 | .91 | .36 to 2.50 |
| Body Mass Index (kg/m2) | −.0111 | < .01* | −.0177 to −.0044 | 1.10 | .02* | 1.02 to 1.19 |
| Caucasian ethnicity | .1194 | < .01* | .0627 to .1761 | .27 | < .01* | .14 to .53 |
| Winter visit | −.0737 | .01* | −.1320 to −.0154 | 1.96 | .04* | 1.03 to 3.75 |
| Current use of efavirenz | −.0187 | .54 | −.0785 to .0411 | 1.29 | .43 | .68 to 2.45 |
| Current use of zidovudine | −.1233 | .03* | −.2329 to −.0138 | 1.93 | .28 | .58 to 6.36 |
Dependent variable for simple linear regression models was Log10 25-hydroxyvitamin D; dependent variable for univariate logistic regression was low 25-hydroxyvitamin D dichotomized at 30ng/ml. Abbreviation: CI: confidence interval.
statistically significant.
Multiple linear regression models were used to analyze the association of winter visit, BMI, ethnicity and current use of zidovudine with 25(OH)D. In the final multiple linear regression model (Table 3, Model 5), only Caucasian ethnicity (β = .1004, P < .01) and BMI (β = −.0078, P = .02) remained significant. The exponentiated β-coefficient for Caucasian ethnicity in Model 5 was 1.2601, indicating that 25(OH)D in Caucasians tend to be 1.26 ng/ml higher, on the average, than non-Caucasians. Exponentiated β-coefficient for BMI was 0.9822, indicating a mean decrease of 0.98 ng/ml in 25(OH)D for every 1.0 kg/m2 increase in BMI.
Table 3.
Multiple linear regression and multivariate logistic regression of factors associated with 25-hydroxyvitamin D.
| Linear Regression | Logistic Regression | |||
| Beta coefficient | P−value | aOR | P−value | |
| Model 1 | ||||
| BMI | −.0091 | < .01* | 1.09 | .05 |
| Caucasian ethnicity | .1066 | < .01* | .29 | < .01* |
| Model 2 | ||||
| Winter visit | −.0563 | .06 | 1.71 | .11 |
| BMI | −.0097 | < .01* | 1.09 | .04* |
| Model 3 | ||||
| Winter visit | −.0652 | .02* | 1.90 | .07 |
| Caucasian ethnicity | .1145 | < .01* | .28 | < .01* |
| Model 4 | ||||
| Winter visit | −.0671 | .02* | 1.90 | .05 |
| Current use of zidovudine | −.1091 | .049* | 1.70 | .39 |
| Model 5 | ||||
| Winter visit | −.0467 | .10 | 1.66 | .15 |
| BMI | −.0078 | .02* | 1.07 | .10 |
| Caucasian ethnicity | .1004 | < .01* | .30 | < .01* |
| Current use of zidovudine | −.0913 | .08 | 1.47 | .55 |
Dependent variable for linear regression models was Log10 25-hydroxyvitamin D; dependent variable for logistic regression was low 25-hydroxyvitamin D dichotomized at 30ng/ml.
statistically significant. Abbreviations: aOR: adjusted odds ratio; BMI: body mass index (kg/m2).
Factors associated with low 25(OH)D dichotomized at 30 ng/ml were identified using univariate logistic regression (Table 2). Winter visit (odds ratio (OR):1.96, P = .04) and BMI (OR 1.10, P = .02) were significantly associated with low 25(OH) D. On the other hand, Caucasian ethnicity was protective (OR 0.27, P < .01). In the final multivariate logistic regression model (Table 3, Model 5), only Caucasian ethnicity remained significant (adjusted OR: 0.30, P < .01).
Discussion
Ethnicity, BMI, winter visit, and use of zidovudine were factors identified in the current study to be associated with low 25(OH)D levels on univariate analysis. Both ethnicity and higher BMI are well recognized predictors for low 25(OH)D.17,18 Melanin filters ultraviolet-B light, which is necessary for cutaneous synthesis of vitamin D.19 Hence, persons with less skin pigmentation are less likely to have low 25(OH)D. Conversely, those with darker skin color are more susceptible to lower 25(OH)D levels.14,18
Multiple hypotheses have been proposed to explain the association between obesity and low 25(OH)D. It is thought that fat-soluble vitamin D is stored in adipose tissue leading to decreased bioavailability.20 Other behavioral factors such as lower outdoor exercise among obese patients have been implicated to contribute to decreased 25(OH)D in persons with higher BMI.21 More recently, a bi-directional Mendelian randomization study of multiple cohorts from North America and Europe has concluded that obesity is a causal risk factor for the development of vitamin D deficiency.22
The effect of winter is accepted as a confounding variable in studies of vitamin D metabolism in temperate climates. Whether seasonal variation occurs in the tropics is a significant research question for vitamin D studies planned in Hawai‘i and other similar settings. Located at latitude 21° North, Hawai‘i is fully within the tropics, bounded by the Tropic of Cancer (latitude 23.3°) in the northern hemisphere and by the Tropic of Capricorn (latitude 23.3°) in the south. Published studies adjacent to the tropical zone have shown seasonal variations in vitamin D. Among female out-patients in Saudi Arabia (latitude 24.2°N), season was found to be a significant factor in a subgroup analysis of postmenopausal women.23 In three regions of Australia (spanning 27°S to 43°S), season, vitamin D effective daily dose, and simulated maximum daily duration of vitamin D synthesis, each explained about 14% of 25(OH)D variation.24 In northern India (26.8° N), mean serum 25(OH)D levels in rural girls and pregnant women were significantly lower during winter.25 In South Florida, United States (25.46°N), a significant seasonal variation in vitamin D levels among male and female ambulatory patients was found.26
The length of day in Hawai‘i is consistent throughout the year but results in some, albeit minimal, seasonal variations in incoming solar radiations.15 In our cohort, winter visit was found to be significant on univariate analyses but had varying significance on multivariate regression models (Table 3). Ethnicity and BMI seemed to be relatively more important factors affecting 25(OH)D levels as suggested by retained significance (P<.05) in the final multiple linear regression model. Further analysis of patients stratified by BMI (Figure 2a) showed that the effect of season on 25(OH)D is likely to be more profound among non-obese patients.
Previous research have associated efavirenz and zidovudine with low 25(OH)D levels and bone demineralization.11,12 Zidovudine has been implicated to activate osteoclastogenesis, while efavirenz cause increased metabolism of 25(OH)D into inactive compounds by inducing cytochrome P450 enzymes.27 In the current study, the use of zidovudine was associated with lower 25(OH)D levels on simple linear regression but was not significant in the final multiple linear regression model. No association between current use of efavirenz and 25(OH)D levels was found.
Interestingly, the median 25(OH)D of the HAHC-CVD cohort was 32.4 ng/ml with 43% having 25(OH)D levels below 30 ng/ml. This proportion of vitamin D insufficiency is lower when compared to 79.1% of the United States general adult population (National Health and Nutrition Examination Survey (NHANES), 2003–2004 and 2005–2006) and to 70.3% of patients in the SUN cohort (Study to Understand the Natural History of HIV and AIDS in the Era of Effective Therapy) who were found to have 25(OH)D below 30 ng/ml.11 Results of the current study are more consistent with another study among HIV-negative adults in Hawai‘i which reported that 51% had 25(OH)D below 30 ng/ml.28 The relatively lower prevalence of vitamin D insufficiency and deficiency may be due to the substantial availability of sunlight in Hawai‘i.
Our research is limited by the lack of data on over-the-counter vitamin D, multivitamin supplementation, diet, and amount of sun exposure of the study participants. It is also focused on older individuals, which may explain the lack of significant association between age and 25(OH)D levels.
In conclusion, seasonal differences in 25(OH)D levels were observed in our cohort of ambulatory HIV-infected patients in Hawai‘i. BMI and ethnicity were better predictors of 25(OH)D levels than season in multivariate analysis. Future interventional studies to investigate the effect of increasing 25(OH)D levels on morbidity and mortality of HIV-infected patients living in the tropics are warranted.
Acknowledgments
We thank the patients of the HAHC-CVD cohort for their participation in the study, as well as the staff of the Hawai‘i Center for AIDS for facilitating participant recruitment and specimen processing.
Disclosure Statement/Conflict of Interest
This study was supported by NIH grants (R01HL095135, U54RR026136) and U54RR026136. None of the authors identify any conflict of interest.
References
- 1.Song GG, Bae SC, Lee YH. Association between vitamin D intake and the risk of rheumatoid arthritis: a meta-analysis. Clin Rheumatol. 2012 Dec;31(12):1733–1739. doi: 10.1007/s10067-012-2080-7. [DOI] [PubMed] [Google Scholar]
- 2.Gatenby P, Lucas R, Swaminathan A. Vitamin D deficiency and risk for rheumatic diseases: an update. Curr Opin Rheumatol. 2013 Mar;25(2):184–191. doi: 10.1097/BOR.0b013e32835cfc16. [DOI] [PubMed] [Google Scholar]
- 3.Shikuma CM, Seto T, Liang CY, et al. Vitamin D levels and markers of arterial dysfunction in HIV. AIDS Res Hum Retroviruses. 2012 Aug;28(8):793–797. doi: 10.1089/aid.2011.0086. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Breer S, Krause M, Marshall RP, Oheim R, Amling M, Barvencik F. Stress fractures in elderly patients. Int Orthop. 2012 Dec;36(12):2581–2587. doi: 10.1007/s00264-012-1708-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lin SW, Chen W, Fan JH, et al. Prospective study of serum 25-hydroxyvitamin D concentration and mortality in a Chinese population. Am J Epidemiol. 2012 Dec 1;176(11):1043–1050. doi: 10.1093/aje/kws285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Blicher TM, Jorgensen HL, Schwarz P, Wulf HC. Low levels of vitamin D are associated with increased mortality in patients attending a university hospital in Denmark. Scand J Clin Lab Invest. 2013 Feb;73(1):24–28. doi: 10.3109/00365513.2012.732238. [DOI] [PubMed] [Google Scholar]
- 7.Trummer O, Pilz S, Hoffmann MM, et al. Vitamin D and Mortality: A Mendelian Randomization Study. Clin Chem. 2013 Jan 14; doi: 10.1373/clinchem.2012.193185. [DOI] [PubMed] [Google Scholar]
- 8.Ford ES, Zhao G, Tsai J, Li C. Vitamin D and all-cause mortality among adults in USA: findings from the National Health and Nutrition Examination Survey Linked Mortality Study. Int J Epidemiol. 2011 Aug;40(4):998–1005. doi: 10.1093/ije/dyq264. [DOI] [PubMed] [Google Scholar]
- 9.Song EJ, Gordon-Thomson C, Cole L, et al. 1alpha,25-Dihydroxyvitamin D(3) reduces several types of UV-induced DNA damage and contributes to photoprotection. J Steroid Biochem Mol Biol. 2012 Nov 16; doi: 10.1016/j.jsbmb.2012.11.003. [DOI] [PubMed] [Google Scholar]
- 10.Griffin AT, Arnold FW. Review of metabolic, immunologic, and virologic consequences of suboptimal vitamin D levels in HIV infection. AIDS Patient Care STDS. 2012 Sep;26(9):516–525. doi: 10.1089/apc.2012.0145. [DOI] [PubMed] [Google Scholar]
- 11.Dao CN, Patel P, Overton ET, et al. Low vitamin D among HIV-infected adults: prevalence of and risk factors for low vitamin D Levels in a cohort of HIV-infected adults and comparison to prevalence among adults in the US general population. Clin Infect Dis. 2011 Feb 1;52(3):396–405. doi: 10.1093/cid/ciq158. [DOI] [PubMed] [Google Scholar]
- 12.Fox J, Peters B, Prakash M, Arribas J, Hill A, Moecklinghoff C. Improvement in vitamin D deficiency following antiretroviral regime change: Results from the MONET trial. AIDS Res Hum Retroviruses. 2011 Jan;27(1):29–34. doi: 10.1089/aid.2010.0081. [DOI] [PubMed] [Google Scholar]
- 13.Lagunova Z, Porojnicu AC, Lindberg F, Hexeberg S, Moan J. The dependency of vitamin D status on body mass index, gender, age and season. Anticancer Res. 2009 Sep;29(9):3713–3720. [PubMed] [Google Scholar]
- 14.Weng FL, Shults J, Leonard MB, Stallings VA, Zemel BS. Risk factors for low serum 25-hydroxyvitamin D concentrations in otherwise healthy children and adolescents. Am J Clin Nutr. 2007 Jul;86(1):150–158. doi: 10.1093/ajcn/86.1.150. [DOI] [PubMed] [Google Scholar]
- 15.National Weather Service Forecast Office, author. Climate of Hawaii. [September 14, 2012]. http://www.prh.noaa.gov/hnl/pages/climate_summary.php.
- 16.Holick MF, Binkley NC, Bischoff-Ferrari HA, et al. Evaluation, treatment, and prevention of vitamin D deficiency: an Endocrine Society clinical practice guideline. J Clin Endocrinol Metab. 2011 Jul;96(7):1911–1930. doi: 10.1210/jc.2011-0385. [DOI] [PubMed] [Google Scholar]
- 17.Crutchley RD, Gathe J, JR, Mayberry C, Trieu A, Abughosh S, Garey KW. Risk factors for vitamin D deficiency in HIV-infected patients in the south central United States. AIDS Res Hum Retroviruses. 2012 May;28(5):454–459. doi: 10.1089/aid.2011.0025. [DOI] [PubMed] [Google Scholar]
- 18.Coney P, Demers LM, Dodson WC, Kunselman AR, Ladson G, Legro RS. Determination of vitamin D in relation to body mass index and race in a defined population of black and white women. Int J Gynaecol Obstet. 2012 Oct;119(1):21–25. doi: 10.1016/j.ijgo.2012.05.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Norman AW. Sunlight, season, skin pigmentation, vitamin D, and 25-hydroxyvitamin D: integral components of the vitamin D endocrine system. Am J Clin Nutr. 1998 Jun;67(6):1108–1110. doi: 10.1093/ajcn/67.6.1108. [DOI] [PubMed] [Google Scholar]
- 20.Wortsman J, Matsuoka LY, Chen TC, Lu Z, Holick MF. Decreased bioavailability of vitamin D in obesity. Am J Clin Nutr. 2000 Sep;72(3):690–693. doi: 10.1093/ajcn/72.3.690. [DOI] [PubMed] [Google Scholar]
- 21.Florez H, Martinez R, Chacra W, Strickman-Stein N, Levis S. Outdoor exercise reduces the risk of hypovitaminosis D in the obese. J Steroid Biochem Mol Biol. 2007 Mar;103(3-5):679–681. doi: 10.1016/j.jsbmb.2006.12.032. [DOI] [PubMed] [Google Scholar]
- 22.Vimaleswaran KS, Berry DJ, Lu C, et al. Causal Relationship between Obesity and Vitamin D Status: Bi-Directional Mendelian Randomization Analysis of Multiple Cohorts. PLoS Medicine. 2013 Feb;10(2):e1001383. doi: 10.1371/journal.pmed.1001383. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Kanan RM, Al Saleh YM, Fakhoury HM, Adham M, Aljaser S, Tamimi W. Year-round vitamin D deficiency among Saudi female out-patients. Public Health Nutr. 2012 Jun 13;:1–5. doi: 10.1017/S1368980012002947. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.van der Mei IA, Ponsonby AL, Engelsen O, et al. The high prevalence of vitamin D insufficiency across Australian populations is only partly explained by season and latitude. Environ Health Perspect. 2007 Aug;115(8):1132–1139. doi: 10.1289/ehp.9937. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Sahu M, Bhatia V, Aggarwal A, et al. Vitamin D deficiency in rural girls and pregnant women despite abundant sunshine in northern India. Clin Endocrinol (Oxf) 2009 May;70(5):680–684. doi: 10.1111/j.1365-2265.2008.03360.x. [DOI] [PubMed] [Google Scholar]
- 26.Levis S, Gomez A, Jimenez C, et al. Vitamin d deficiency and seasonal variation in an adult South Florida population. J Clin Endocrinol Metab. 2005 Mar;90(3):1557–1562. doi: 10.1210/jc.2004-0746. [DOI] [PubMed] [Google Scholar]
- 27.Gutierrez F, Masia M. The role of HIV and antiretroviral therapy in bone disease. AIDS Rev. 2011 Apr-Jun;13(2):109–118. [PubMed] [Google Scholar]
- 28.Binkley N, Novotny R, Krueger D, et al. Low vitamin D status despite abundant sun exposure. J Clin Endocrinol Metab. 2007 Jun;92(6):2130–2135. doi: 10.1210/jc.2006-2250. [DOI] [PubMed] [Google Scholar]