TABLE 3.

Relative risk of renal cell cancer in relation to BMI^* at four distinct age periods, United States, 1995–1996 through 2003^†

	BMI (kg/m2)						p-trend‡
	<18.5	18.5–<22.5	22.5–<25	25–<27.5	27.5–<30	≥30
Baseline BMI
Men§
HR*		1.0	1.12	1.51	1.74	1.87	<0.0005
95% CI*		Referent	0.73, 1.72	1.01, 2.26	1.15, 2.63	1.24, 2.82
No. of cases	0	28	88	169	127	152
Women
HR	3.65	1.0	1.66	2.44	2.27	2.67	0.002
95% CI	1.23, 10.9	Referent	0.92, 2.98	1.39, 4.26	1.23, 4.20	1.53, 4.66
No. of cases	4	17	33	46	27	64
BMI at age 50 years
Men
HR	1.24	1.0	1.12	1.49	1.37	1.90	<0.0005
95% CI	0.30, 5.13	Referent	0.78, 1.60	1.06, 2.11	0.94, 2.01	1.31, 2.78
No. of cases	2	40	111	195	83	107
Women
HR	0.49	1.0	1.57	2.02	2.00	2.12	0.001
95% CI	0.07, 3.54	Referent	1.03, 2.40	1.30, 3.14	1.13, 3.53	1.27, 3.53
No. of cases	1	37	53	46	19	33
BMI at age 35 years
Men
HR	0.40	1.0	1.08	1.36	1.61	1.65	<0.0005
95% CI	0.10, 1.63	Referent	0.85, 1.39	1.07, 1.74	1.19, 2.20	1.16, 2.33
No. of cases	2	108	150	166	68	49
Women
HR	0.61	1.0	1.27	1.46	1.26	1.44	0.06
95% CI	0.22, 1.67	Referent	0.90, 1.80	0.92, 2.31	0.58, 2.76	0.71, 2.93
No. of cases	4	87	52	24	7	9
BMI at age 18 years
Men
HR	0.83	1.0	0.87	1.28	1.55	1.29	0.02
95% CI	0.63, 1.10	Referent	0.69, 1.09	0.98, 1.67	0.99, 2.42	0.72, 2.31
No. of cases	59	288	98	67	21	12
Women
HR	1.20	1.0	0.89	0.87	2.31	0.78	0.81
95% CI	0.81, 1.78	Referent	0.56, 1.43	0.40, 1.87	1.01, 5.28	0.19, 3.20
No. of cases	33	105	21	7	6	2

^*BMI, body mass index; HR, hazard ratio; CI, confidence interval.

^†Subcohort analysis. Pearson's correlations for BMI at different ages: 18 and 35 years, 0.65; 18 and 50 years, 0.47; 18 years and baseline, 0.34; 35 and 50 years, 0.75; 35 years and baseline, 0.56; 50 years and baseline, 0.78. Participants were 51–72 years of age at baseline. Statistical interactions were observed between BMI at age 18 years and baseline (modeled by continuous variable cross-products; likelihood ratio test p values: 0.06 for men, 0.65 for women, <0.05 for men and women combined.

^‡Linear trend of category medians.

^§All analyses were adjusted for age, smoking status and dose, physical activity, protein intake, history of diabetes, and history of hypertension.