Table 4.

Multivariate binary logistic regression analysis showing polycystic ovary syndrome as a dependent variable on independent variables age, body mass index, serum fasting glucose, serum fasting insulin and serum homocysteine levels

Variables	B	SE	Wald statistic	P	OR	95.0% CI for OR
						Lower	Upper
Age	0.044	0.068	0.407	0.524	1.045	0.914	1.194
Fasting glucose	0.093	0.060	2.411	0.120	1.098	0.976	1.234
Fasting insulin	−0.107	0.353	0.091	0.763	0.899	0.450	1.796
HOMA-IR	0.522	1.594	0.107	0.743	1.685	0.074	38.340
Serum homocysteine	0.158	0.065	5.977	0.014	1.172	1.032	1.330
BMI	−0.032	0.079	0.159	0.690	0.969	0.829	1.132
Constant	−10.626	5.612	3.585	0.058	<0.001
Classification table based on binary logistic regression
Projected group	Group		Total
	PCOS	Controls
PCOS	29	7	36
Controls	6	28	34
Classifying efficacy of predictive model
Sensitivity	Specificity	PPV	NPV	Accuracy
82.9	80.0	80.6	77.8	81.4

BMI=Body mass index, SE=Standard error, OR=Odds ratio, CI=Confidence interval, HOMA-IR=Homoeostatic assessment of insulin resistance, PCOS=Polycystic ovary syndrome, PPV=Positive predictive value, NPV=Negative predictive value