. Author manuscript; available in PMC: 2020 Jun 1.

Published in final edited form as: Minerva Endocrinol. 2018 Apr 12;44(2):176–184. doi: 10.23736/S0391-1977.18.02824-9

Table III.

Crude and adjusted prevalence odds ratios^a (PORs) and 95% confidence intervals (CI) evaluating associations between vitamin D categories and PCOS phenotypic subgroups^b.

	PCOS phenotype #1 (n=55) ovulatory dysfunction polycystic ovaries		PCOS phenotype #2 (n=15) ovulatory dysfunction androgen excess		PCOS phenotype #3 (n=67) ovulatory dysfunction polycystic ovariesandrogen excess
Vitamin D Exposure Categories (ng/mL)	Crude POR (95% CI)	Adjusted POR^c (95% CI)	Crude POR (95% CI)	Adjusted POR^c (95% CI)	Crude POR (95% CI)	Adjusted POR^c (95% CI)
Vitamin D categories
Sufficient (≥30)	Referent	Referent	Referent	Referent	Referent	Referent
Insufficient (21–29)	0.77 (0.36, 1.64)	0.64 (0.29, 1.41)	3.86 (1.12, 13.26)	2.98 (0.80, 11.06)	2.06 (1.01, 4.18)	1.47 (0.68, 3.17)
Deficient (≤20)	1.29 (0.51, 3.25)	0.97 (0.35, 2.64)	0.96 (0.10, 9.32)	0.94 (0.09, 9.61)	3.31 (1.40, 7.82)	2.19 (0.83, 5.80)

^a Prevalence odds ratios were calculated using logistic regression.

^bThere were no women in our study who exhibited only polycystic ovaries and hyperandrogenism. All PCOS women were oligo or anovulatory.

^cAdjusted for BMI category and race/ethnicity