Table 2:

Pearson’s correlation coefficients and multivariable linear regression for the association between log-transformed serum hormones and log-transformed urine hormones for participants with matching serum and urine collection dates within 16-24 days of the menstrual cycle

	Pearson’s Correlation			Multivariable Linear Regression^a
Hormone	N	Pearson’s r	p-value	N^b	β (SE)	95%CI	R²	p-value^c
LH	159	0.573	<0.0001	156	0.44 (0.05)	(0.34, 0.53)	0.42	<0.0001

FSH	168	0.845	<0.0001	164	0.76 (0.03)	(0.66, 0.79)	0.76	<0.0001

E2/E1c	170	0.574	<0.0001	167	0.84 (0.09)	(0.65, 1.02)	0.37	<0.0001

P/uPdg	170	0.837	<0.0001	167	1.10 (0.05)	(0.99, 1.20)	0.74	<0.0001

Model adjusted for ethnicity, age, log-body mass index at urine collection and smoking status

LH: 2 observations with missing body mass index and 1 outlier excluded; follicle stimulating hormone (FSH): 2 observations with missing body mass index and 2 outliers excluded; estradiol/estrone conjugates (E2/E1c): 2 observations with missing body mass index and 1 outlier excluded; progesterone/urine pregnanediol glucuronide (P/uPdg): 2 observations with missing body mass index and 1 outlier excluded

Significance level adjusted for multiple testing of serum LH, FSH, E2 vs urine LH, FSH, E1C=0.0125; and of P as related to uPdg= 0.01