Skip to main content
. 2018 Oct 27;8(10):e021860. doi: 10.1136/bmjopen-2018-021860

Table 2.

Summary of binary regression analysis for variables predicting PCOS with the OR of each risk factor adjusted for other variables in the model

OR (95% CI)
BMI <25 kg/m2
(normal)
BMI 25–30 kg/m2(overweight) BMI ≥30 kg/m2
(obese)
CRP/albumin ratio*
 <0.097 1 1 1
 ≥0.097 11.21 (3.28 to 39.75) 19.32 (5.07 to 72.17) 34.5 (7.75 to 153.52)
Insulin resistance†
 No 1 1 1
 Borderline 3.34 (0.645 to 17.33) 5.58 (0.907 to 34.41) 3.13 (0.53 to 18.48)
 Yes 9.21 (1.63 to 51.93) 8.81 (1.75 to 44.31) 17.94 (1.81 to 177.61)
Free Androgen Index‡
 <3.95 1 1 1
 ≥3.95 2.28 (0.536 to 9.75) 0.86 (0.22 to 3.35) 3.79 (0.59 to 24.42)

*Cut-off values were derived from the sensitivity and specificity analysis of the receiver operating characteristic curve.

†Categories of insulin resistance based on normal values used for HOMA-insulin resistance.

‡Cut-off values from normal laboratory reference ranges for Free Androgen Index.

CRP, C reactive protein; HOMA, homeostasis model assessment; PCOS, polycystic ovary syndrome.