Table 5.

Multiple regression models of serum inflammatory biomarker outcomes including steroidal and hormonal variables that had predicted sex.

Variables in the model of total subjects	ASAA outcome		sIL-2Rα outcome		sTNF-R1 outcome
	(N = 142)		(N = 159)		(N = 159)
	Beta∗	p values	Beta∗	p values	Beta∗	p values
17-OH P5	−0.216	0.023	−0.116	0.268	0.334	<0.001
DHEA	0.286	0.013	0.200	0.044	−0.013	0.872
Adione	0.022	0.853	−0.032	0.732	−0.387	<0.001
DHEAS	−0.147	0.129	−0.147	0.130	0.177	0.021
LH	0.084	0.365	0.030	0.772	−0.094	0.255
Entry age	0.226	0.016	−0.102	0.292	0.006	0.936
Variables in the model of control subjects	ASAA outcome		sIL-2Rα outcome		sTNF-R1 outcome
	(n = 113)		(n = 126)		(n = 126)
	Beta∗	p values	Beta∗	p values	Beta∗	p values
17-OH P5	−0.257	0.017	0.021	0.862	0.303	0.001
DHEA	0.198	0.146	0.204	0.073	0.046	0.604
Adione	0.170	0.209	−0.092	0.390	−0.450	<0.001
DHEAS	−0.185	0.092	−0.193	0.096	0.241	0.007
LH	0.054	0.613	0.088	0.470	−0.037	0.703
Entry age	0.199	0.054	−0.054	0.631	0.038	0.668
Variables in the model of pre-RA subjects	ASAA outcome		sIL-2Rα outcome		sTNF-R1 outcome
	(n = 29)		(n = 33)		(n = 33)
	Beta	p values	Beta∗	p values	Beta∗	p values
17-OH P5	0.127	0.533	−0.595	0.015	0.479	0.030
DHEA	0.496	0.015	0.367	0.094	−0.193	0.313
Adione	−0.618	0.015	0.123	0.552	−0.198	0.272
DHEAS	−0.166	0.417	0.031	0.873	0.022	0.896
LH	0.155	0.364	−0.237	0.254	−0.231	0.208
Entry age	0.210	0.307	−0.130	0.537	−0.077	0.671

^∗Standardized beta coefficients from linear regression of log-transformed, reported NEI biomarkers.