Table 3. Separate multivariate linear regression models of the relation of serum CML, sRAGE, and esRAGE at baseline with estimated GFR in women, aged ≥65 years, in the Women's Health and Aging Study I in Baltimore, Maryland¹.

Serum CML2 (μg/mL)	Model adjusted for age			Model adjusted for age, race			Model adjusted for age, race, hemoglobin A1c, chronic diseases3
	Beta	SE	P	Beta	SE	P	Beta	SE	P
	-4.03	0.65	<0.001	-4.11	0.64	<0.001	-4.07	0.69	<0.001
Serum sRAGE2 (ng/mL)	Model adjusted for age			Model adjusted for age, race			Model adjusted for age, race, hemoglobin A1c, chronic diseases3
	Beta	SE	P	Beta	SE	P	Beta	SE	P
	-4.11	0.65	<0.001	-3.99	0.68	<0.001	-3.32	0.74	<0.001
Serum esRAGE2 (ng/mL)	Model adjusted for age			Model adjusted for age, race			Model adjusted for age, race, hemoglobin A1c, chronic diseases3
	Beta	SE	P	Beta	SE	P	Beta	SE	P
	-3.92	0.66	<0.001	-3.72	0.69	<0.001	-3.90	0.73	<0.001

¹GFR (glomerular filtration rate), CML (carboxymethyl-lysine), sRAGE (soluble receptor for advanced glycation end products), esRAGE (endogenous secretory receptor for advanced glycation end products). Separate multivariate linear regression models shown for serum CML, sRAGE, and esRAGE in which estimated GFR is the dependent variable.

²Betas are expressed per 1 SD change in serum CML, sRAGE, and esRAGE (0.28 μgmL, 0.21 ng/mL, and 0.71 ng/mL, respectively).

³Chronic diseases were coronary heart disease, congestive heart failure, and peripheral artery disease.