Original Study

Vitamin D Deciency among Adolescent Females with Polycystic

Ovary Syndrome
Mandakini Sadhir MBBS 1,*, Alvina R. Kansra MD 2, Seema Menon MD 3
1
Division of Adolescent Medicine, Department of Pediatrics, Medical College of Wisconsin, Milwaukee, Wisconsin
2
Division of Pediatric Endocrinology, Department of Pediatrics, Medical College of Wisconsin, Milwaukee, Wisconsin
3
Department of Obstetrics and Gynecology, Medical College of Wisconsin, Milwaukee, Wisconsin

a b s t r a c t
Study Objective: Studies have suggested that low vitamin D levels may play a role in the pathogenesis of polycystic ovary syndrome (PCOS).
The aim of our study was to compare 25-hydroxyvitamin D [25(OH)D] levels in adolescent females with and without PCOS.
Design, Setting, and Participants: Retrospective chart review at a tertiary care medical center for female adolescents aged 12-21 years with
serum 25(OH)D measurements within a 5-year period. Participants were categorized as having PCOS or as controls based on National
Institutes of Health PCOS diagnostic criteria.
Main Outcome Measure: Exact logistic regression analysis was done to compare normal ($30 ng/mL) vs low (!30 ng/mL) serum 25(OH)D
levels in the PCOS and control groups.
Results: Two hundred ninety-nine charts were reviewed and 107 participants were included in the study. Of the included participants, 37
were in the PCOS group and 70 were in the control group, with a mean age of 15.2 years. In the PCOS group, 97.2% were obese and vitamin
D deciency was noted among 62.2% females. The mean serum 25(OH)D level was 18.4 and 21.6 ng/mL in PCOS and control groups,
respectively. The difference in mean 25(OH)D levels between the 2 groups was not statistically signicant (P O .05) when controlled for
ethnicity, body mass index percentile, and season.
Conclusion: In our study, there was no statistically signicant difference in mean 25(OH)D levels between PCOS and control groups. The
majority of participants in PCOS group were obese. Further studies in adolescent females with PCOS and normal body mass index could be
helpful in delineating the role of vitamin D in the pathogenesis of PCOS.
Key Words: Obesity, Vitamin D, Polycystic ovary syndrome, Adolescents

Introduction of luteinizing hormone, sex hormone binding globulin,7

Polycystic ovary syndrome (PCOS) is a complex dis-
order affecting the hypothalamic-pituitary-ovarian axis
with an estimated prevalence of 5%-10% in reproductive-
age women.1 PCOS is characterized by chronic anovulation
leading to menstrual irregularities and hyperandrogenism
marked by elevated serum androgen levels and clinical
features such as hirsutism, acne, or alopecia.1,2 PCOS is often
associated with obesity and insulin resistance leading to
metabolic disturbances including impaired glucose toler-
ance, type 2 diabetes mellitus, and dyslipidemia with future
risk of atherosclerosis and cardiovascular disease.3,4 Other
associated health concerns include infertility, endometrial
hyperplasia, and cancer.1 The etiology of PCOS is not fully
understood. There are some data pointing to an association
between vitamin D and PCOS.5
Vitamin D is thought to regulate gene transcription
through vitamin D receptors that are widely distributed in
tissues including ovaries.6 Vitamin D receptorerelated
genetic polymorphisms have been linked to serum levels
The authors indicate no conicts of interest.
* Address correspondence to: Mandakini Sadhir, MBBS, Division of Adolescent
Medicine, Department of Pediatrics, Kentucky Clinic, J415, 740 S Limestone St,
Lexington, KY; Phone: 1 (859) 218-5183
E-mail address: m.sadhir@uky.edu (M. Sadhir).
1083-3188/$ - see front matter 2015 North American Society for Pediatric and Adolescent Gynecology. Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.jpag.2014.11.004
testosterone, and insulin.5,7,8 Current literature suggests
a correlation between low vitamin D levels and insulin
resistance in women with PCOS.10,11 In addition, it has been
postulated that vitamin D insufciency and calcium
dysfunction may lead to menstrual irregularities.9 Limited
evidence supports benecial effect of vitamin D supple-
mentation on insulin resistance, menstrual dysfunction,
and fertility.12e15
The prevalence of vitamin D deciency in women
with PCOS is unknown. Low serum 25-hydroxyvitamin D
[25(OH)D] levels have been reported in women with PCOS
with some inconsistencies.16e18 Studies hypothesize that
vitamin D deciency in PCOS seems to be associated with
obesity.16,17 There is currently insufcient data regarding
association of serum 25(OH)D levels and PCOS in adolescent
females. Vitamin D deciency is common particularly
among obese, non-Hispanic black adolescents,18e20 as well
as in those with certain chronic medical conditions, and
thus screening is recommended in these groups.21 In
addition, different seasons impact serum 25(OH)D levels.20
It is unknown if adolescent females with PCOS have higher
prevalence of vitamin D deciency than do adolescent
females without PCOS. The aim of our study was to com-
pare serum 25(OH)D levels in adolescent females with and
without PCOS.
 M. Sadhir et al. / J Pediatr Adolesc Gynecol 28 (2015) 378e381 379

Table 1
Vitamin D Status and Mean 25(OH)D Level of Participants in Various Subgroups Based on BMI Percentile, Ethnicity, and Season

Vitamin D Status Decient (#20 Ng/mL) Insufcient (21-29 Ng/mL) Sufcient ($30 Ng/mL) Mean 25(OH)D Level (Ng/mL) P value*

BMI Percentile .172

85th-95th (n 5 14) 28.6% (4) 50% (7) 21.4% (3) 26.0
95th-99th (n 5 49) 53.1% (26) 28.6% (14) 18.3% (9) 21.7
$99th (n 5 44) 66% (29) 29.5% (13) 4.5% (2) 21.7
Ethnicity .001
Non-Hispanic black (n 5 28) 96.4% (27) 3.6% (1) 0 13.6
Hispanic (n 5 28) 57.1% (16) 39.3% (11) 3.6% (1) 19.1
Non-Hispanic white (n 5 51) 31.4% (16) 43.1% (22) 25.5% (13) 25.1
Season .01
Fall (n 5 34) 58.8% (20) 32.3% (11) 8.8% (3) 19.7
Spring (n 5 25) 68% (17) 28% (7) 4% (1) 17.9
Summer (n 5 31) 32.2% (10) 35.5% (11) 32.2% (10) 24.6
Winter (n 5 17) 70.6% (12) 29.4% (5) 0 18.5

* Exact logistic regression to compare vitamin D level !30 ng/mL vs $ 30 ng/mL.

Methods using the Centers for Disease Control and Prevention

A retrospective chart review of adolescent females aged
12-21 years was conducted. The study was approved by the
Children's Hospital of Wisconsin's institutional review
board. A total of 299 participants with serum 25(OH)D
levels measured within a 5-year period (2008-2012) were
identied from adolescent medicine, adolescent gynecol-
ogy and pediatric endocrine clinics. Serum 25(OH)D levels
were obtained using liquid chromatographyetandem mass
spectrometry, which is a standard method at our insti-
tutional laboratory. Exclusion criteria included history of
vitamin D deciency; current vitamin D supplementation;
increased risk of vitamin D deciency secondary to current
medication including steroids, antiretrovirals, antiepilep-
tics, or antifungals; or chronic medical conditions such as
chronic kidney disease, osteoporosis, osteopenia, cystic
brosis, or inammatory bowel disease. In addition, those
with incomplete documentation of menstrual pattern,
undocumented ethnicity, or premenarchal state were
also excluded. Chart review was performed to collect
information on menstrual pattern, clinical signs of hyper-
androgenemia such as hirsutism, and serum free testos-
terone level. The participants were then categorized as in
either the PCOS or the control group based on National
Institutes of Health (NIH) PCOS diagnostic criteria (chronic
anovulation characterized by oligomenorrhea, primary or
secondary amenorrhea, clinical or biochemical hyper-
androgenemia with exclusion of other mimicking condi-
tions).2 All participants in the PCOS group had chronic
anovulation with elevated serum free testosterone. Con-
ditions such as hyperprolactinemia, congenital adrenal
hyperplasia, thyroid disease, and androgen-secreting tu-
mors were effectively excluded by collecting information
on prolactin, 17-hydroxyprogesterone, thyroid-stimulating
hormone, and dehydroepiandrosterone sulfate levels.
Data regarding the participant's age, body mass index
(BMI) percentile, ethnicity, serum 25(OH)D levels, and the
season the blood was drawn were recorded in an Internet-
based database (Redcap: www.project-redcap.org). Serum
25(OH)D levels were dened as decient if the levels were
criteria for BMI percentile, participants were categorized as
normal (!85th percentile), overweight (85th-95th per-
centile), obese (95th-99th percentile), and morbid obese
(O99th percentile). The participants were also grouped
by season (spring, summer, fall, winter), ethnicity (black,
white, Hispanic), and BMI percentile (!85th, 85th-95th,
95th-99th, O99th). Serum 25(OH)D levels were compared
between the PCOS and control groups as well as based on
season, ethnicity, and BMI percentile.
Statistical Analysis
It was calculated that the study would have 80% power
to detect difference in mean 25(OH)D levels of 6.0 assuming
a standard deviation of 10 with a sample size of 40 in
the PCOS group and 80 in the control group. Exact
logistic regression analysis was done to compare normal
($30 ng/mL) vs low (!30 ng/mL) serum 25(OH)D levels in
the PCOS and control groups. Effects of season, ethnicity,
and BMI percentile were controlled in the model. All
data management and analyses were carried out using the
Statistical Analysis System, version 9.2 (SAS Institute, Cary,
NC, USA).Two-tailed values of P ! .05 were considered
statistically signicant.
Results
Of the 299 charts reviewed, 160 participants met the
inclusion criteria and were categorized as either PCOS or
control. All participants in PCOS group had BMI percentile
greater than the 85th percentile. Therefore, participants
with BMI percentile of less than 85th in the control
group were excluded (n 5 53) for analysis. The study
sample included 107 participants; 37 were in the PCOS
group and 70 were in the control group. The mean age of
Table 2
Distribution of Participants in PCOS and Control Group Based on Their Vitamin D
Status
Vitamin D Status Control PCOS

20 ng/mL or less, insufcient if levels were between 21 and Decient (#20 ng/mL) 51.4% (36) 62.2% (23)
Insufcient (21-29 ng/mL) 32.8% (23) 29.7% (11)
29 ng/mL, and sufcient if levels were 30 ng/mL or grea-
Sufcient ($30 ng/mL) 15.7% (11) 8.5% (3)
ter.22 BMI percentile was used as a marker for adiposity;
380 M. Sadhir et al. / J Pediatr Adolesc Gynecol 28 (2015) 378e381
Table 3
Vitamin D Levels in PCOS and Control Groups
Group n Age (y) Mean serum 25(OH)D level (ng/mL)
PCOS 37 14.9 18.46
Control 70 15.5 21.60
* Exact logistic regression to compare vitamin D level !30 ng/mL vs $ 30 ng/mL. P ! .05 considered statistically signicant.
participants was 15.2 years. In the sample, non-Hispanic
whites composed 47.5% of the participants, 26.2% of the
participants were Hispanic, and 26.2% of participants were
non-Hispanic blacks. Vitamin D deciency and insufci-
ency were frequently diagnosed in our study population
(Table 1). Sufcient 25(OH)D levels were found in only 14 of
107 participants (13%). Non-Hispanic black females had a
signicantly lower mean 25(OH)D level compared with
Hispanic and non-Hispanic white females. The majority of
participants with BMI greater than 95th percentile were
vitamin D decient with no statistical difference in mean
25(OH)D levels based on each category of BMI percentile.
Seasonal variation in 25(OH)D levels was noted with
signicantly higher levels during the summer season.
Vitamin D deciency was noted among 62.2% of partici-
pants with PCOS versus 51.4% of participants in the control
group (Table 2).The mean serum 25(OH)D levels was
18.46 ng/mL in the PCOS group and 21.6 ng/mL in the
control group. The difference in 25(OH)D levels between
the 2 groups was not statistically signicant (P 5 .459)
when controlled for ethnicity, BMI percentile, and season
(Table 3).
Discussion
Our study is among the rst to investigate vitamin D
levels in adolescent females with PCOS. While there was no
statistically signicant difference in mean 25(OH)D levels
between the PCOS and control groups, the majority of par-
ticipants with PCOS were obese with low serum 25(OH)D
levels. BMI percentile in adolescents has been found to
correlate with 25(OH)D levels.20 Vitamin D is fat soluble,
and higher sequestration occurs in adipose tissue, resulting
in less bioavailability.22 In addition, obese individuals may
spend less time outdoors with less sun exposure and have
insufcient vitamin D biosynthesis in the skin. Dietary
preferences and inadequate calcium and vitamin D intakes
may also contribute to insufcient vitamin D levels.19,20
PCOS in the adolescent population has proved to be a
diagnostic quandary. Evidence of this are the many PCOS
diagnostic models used in clinical practice, commonly
referred to as the NIH criteria, Rotterdam criteria, and
androgen excess society criteria.23 Recently, revised Rot-
terdam criteria were proposed specic to adolescents that
includes the presence of all 3 major criteria: oligomenor-
rhea or amenorrhea for 2 years after menarche, clinical and
biochemical hyperandrogenism, and increased ovarian
volume of 10 cm3 or greater in 1 ovary.24 There has been
no consensus on the best or most valid diagnostic
approach. PCOS falls under the domain of various pediatric
subspecialties such as endocrine, adolescent medicine, and
gynecology. These specialties often have different ap-
proaches to documenting clinical examination ndings,
Standard deviation Median Minimum Maximum P value*
7.20 17.00 7.00 40.00 .459
9.16 20.00 5.00 51.00
androgen panel evaluation, and use of ultrasound when
evaluating for PCOS and, ultimately, often subscribe to
different PCOS diagnostic recommendations. We chose to
use the NIH diagnosis model based on the clinical practice
of our gynecology and adolescent group. We are comfort-
able that a true difference does exist between our control
group and PCOS group as we strictly adhered to the diag-
nostic criteria. Further, while we did not have age of
menarche as an inclusion/exclusion criterion, the mean age
of our participants in the PCOS and control groups was 15.5
and 14.9 years, respectively, likely eliminating age as a
cofounder and validating our PCOS diagnosis.
Another limitation of the study was that a comparison of
serum 25(OH)D levels involving participants with BMI in
the normal percentile was not done as there were no
normal BMI percentile adolescent females with PCOS in our
study. In addition, sample size in each group was smaller
than the calculated sample size. A future, large-scale study
that includes normal-BMI adolescent females with PCOS
would be helpful to determine if there is an association
between PCOS and vitamin D.
Our study provides more information about serum
25(OH)D levels in adolescent females with PCOS. Vitamin D
deciency in adolescent females with PCOS seems to be
related to obesity. The study ndings are consistent with
known seasonal variation of 25(OH)D levels20 and pre-
valence of vitamin D deciency in non-Hispanic black
adolescent females. It supports screening recommenda-
tions among non-Hispanic black and obese population.21
Acknowledgments
We would like to thank Shi Zhao, MS, our biostatistician,
for her contribution to statistical analysis.
References
1. Ehrmann DA: Polycystic ovary syndrome. N Engl J Med 2005; 352:1223
2. Roe AH, Dokras A: The diagnosis of polycystic ovary syndrome in adolescents.
Rev Obstet Gynecol 2011; 4:45
3. Ehrmann DA, Barnes RB, Roseneld RL, et al: Prevalence of impaired glucose
tolerance and diabetes in women with polycystic ovary syndrome. Diabetes
Care 1999; 22:141
4. Wild RA, Rizzo M, Clifton S, et al: Lipid levels in polycystic ovary syndrome:
systematic review and meta-analysis. Fertil Steril 2011; 95:1073
5. Thomson RL, Spedding S, Buckley JD: Vitamin D in the etiology and
management of polycystic ovary syndrome. Clin Endocrinol (Oxf) 2012; 77:343
6. Parikh G, Varadinova M, Suwandhi P, et al: Vitamin D regulates steroidogenesis
and insulin-like growth factor binding protein-1 (IGFBP-1) production in
human ovarian cells. Horm Metab Res 2010; 42:754
7. Ranjzad F, Mahban A, Shemirani AI, et al: Inuence of gene variants related to
calcium homeostasis on biochemical parameters of women with polycystic
ovary syndrome. J Assist Reprod Genet 2011; 28:225
8. Wehr E, Trummer O, Giuliani A, et al: Vitamin D-associated polymorphisms are
related to insulin resistance and vitamin D deciency in polycystic ovary
syndrome. Eur J Endocrinol 2011; 164:741
9. Thys-Jacobs S, Donovan D, Papadopoulos A, et al: Vitamin D and calcium
dysregulation in the polycystic ovarian syndrome. Steroids 1999; 64:430
 M. Sadhir et al. / J Pediatr Adolesc Gynecol 28 (2015) 378e381
10. Li HW, Brereton RE, Anderson RA, et al: Vitamin D deciency is common and
associated with metabolic risk factors in patients with polycystic ovary
syndrome. Metabolism 2011; 60:1475
11. Krul-Poel YH, Snackey C, Louwers Y, et al: The role of vitamin D in metabolic
disturbances in polycystic ovary syndrome: a systematic review. Eur J
Endocrinol 2013; 169:853
12. Rashidi B, Haghollahi F, Shariat M, et al: The effects of calcium-vitamin D and
metformin on polycystic ovary syndrome: a pilot study. Taiwan J Obstet
Gynecol 2009; 48:142
13. Raja-Khan N, Shah J, Stetter CM, et al: High-dose vitamin D supplementation
and measures of insulin sensitivity in polycystic ovary syndrome: a
randomized, controlled pilot trial. Fertil Steril 2014; 101:1740
14. Pal L, Berry A, Coraluzzi L, et al: Therapeutic implications of vitamin D and
calcium in overweight women with polycystic ovary syndrome. Gynecol
Endocrinol 2012; 28:965
15. Lerchbaum E, Obermayer-Pietsch B: Vitamin D and fertility: a systematic
review. Eur J Endocrinol 2012; 166:765
16. Tsakova AD, Gateva AT, Kamenov ZA, et al: 25(OH) vitamin D levels in
premenopausal women with polycystic ovary syndrome and/or obesity. Int J
Vitam Nutr Res 2012; 82:399
381
17. Yildizhan R, Kurdoglu M, Adali E, et al: Serum 25-hydroxyvitamin D concen-
trations in obese and non-obese women with polycystic ovary syndrome. Arch
Gynecol Obstet 2009; 280:559
18. Mahmoudi T, Gourabi H, Ashra M, et al: Calciotropic hormones, insulin
resistance, and the polycystic ovary syndrome. Fertil Steril 2010; 93:1208
19. Gordon CM, De Peter KC, Feldman HA, et al: Prevalence of vitamin D deciency
among healthy adolescents. Arch Pediatr Adolesc Med 2004; 158:531
20. Dong Y, Pollock N, Stallmann-Jorgensen IS, et al: Low 25-hydroxyvitamin D
levels in adolescents: race, season, adiposity, physical activity, and tness.
Pediatrics 2010; 125:1104
21. Holick MF, Binkley NC, Bischoff-Ferrari HA, et al: Endocrine Society: Evaluation,
treatment, and prevention of vitamin D deciency: an Endocrine Society clinical
practice guideline. J Clin Endocrinol Metab 2011; 96:1911
22. Wortsman J, Matsuoka LY, Chen TC, et al: Decreased bioavailability of vitamin D
in obesity. Am J Clin Nutr 2000; 72:690
23. Auble B, Elder D, Gross A, et al: Differences in the management of adolescents
with polycystic ovary syndrome across pediatric specialties. J Pediatr Adolesc
Gynecol 2013; 26:234
24. Nicandri KF, Hoeger K: Diagnosis and treatment of polycystic ovarian syndrome
in adolescents. Curr Opin Endocrinol Diabetes Obes 2012; 19:497