Table 4. Associations of serum K (scaled/continuous), rs5215, and rs5219 with incident diabetes in 11,812 ARIC-JHS participants, overall and by race.

Minimally Adjusted, Model 1	Overall (n = 11,812)	P	Caucasians (n = 8653)	P	African Americans (n = 3159)	P
	OR (95% CI)		OR (95% CI)		OR (95% CI)
Serum K (mmol/L) (scaled)	0.86 (0.82, 0.89)	0.00	0.88 (0.83, 0.92)	0.00	0.82 (0.76, 0.88)	0.00
Fully Adjusted, Model 2 rs5215
Serum K (mmol/L) (scaled)	0.88 (0.83, 0.94)	0.00	0.87 (0.8, 0.95)	0.00	0.9 (0.82, 0.99)	0.03
rs5215	0.67 (0.35, 1.3)	0.24	0.55 (0.25, 1.22)	0.14	0.49 (0.07, 3.2)	0.45
Interaction (serum K X rs5215)	1.04 (0.97, 1.11)	0.23	1.06 (0.98, 1.15)	0.13	1.07 (0.88, 1.31	0.50
Fully-adjusted, Model 2 rs5219
Serum K (mmol/L) (scaled)	0.89 (0.84, 0.95)	0.00	0.88 (0.81, 0.95)	0.00	0.9 (0.82, 0.99)	0.02
rs5219	0.79 (0.4, 1.54)	0.49	0.6 (0.27, 1.32)	0.20	0.38 (0.04, 3.54)	0.40
Interaction (serum K X rs5219)	1.03 (0.96, 1.1)	0.45	1.05 (0.97, 1.14)	0.19	1.12 (0.88, 1.41)	0.35

Model 1- adjusted for age, sex, race (Overall only)

Model 2- adjusted for age, sex, race (Overall only), BMI, systolic blood pressure, presence of hypertension, use of ABD medications, serum creatinine, serum sodium, fasting glucose, fasting insulin, parental history of diabetes mellitus, physical activity, income, serum K (quartiles), SNP (rs5215 or rs5219), and interaction terms (SNP X K)