. Author manuscript; available in PMC: 2011 Jan 1.

Published in final edited form as: Cancer Epidemiol Biomarkers Prev. 2010 Jan;19(1):9–17. doi: 10.1158/1055-9965.EPI-09-0777

Table 4.

Mean adjusted estimates and 95% CI of PSADT^* after recurrence among men with diabetes at surgery stratified by race and obesity^**

	No. Patients		Mean adjusted PSADT ^† (95% CI)		p	p – interaction by race^††	P – interaction by obesity^‡

	Total	Diabetic	Diabetic at surgery	Not diabetic at surgery
Overall	192	33	23.4 (15.7 - 34.8)	16.7 (13.0 - 21.4)	0.12	0.11
White	99	15	16.4 (8.8 - 30.3)	18.7 (12.8 - 27.3)	0.71		0.75
Non obese	69	5	20.6 (7.4 - 57.5)	19.8 (12.7 - 30.9)	0.94
Obese	30	10	11.0 (4.8 - 25.0)	22.7 (9.6 - 53.7)	0.24
Black	93	18	29.3 (17.0 - 50.5)	14.9 (10.6 - 21.0)	0.02		0.87
Non obese	61	11	42.5 (20.3 - 89.0)	21.6 (13.7 - 34.1)	0.06
Obese	32	7	27.5 (11.3 - 67.1)	9.6 (6.0 - 15.3)	0.02

Using log transformed PSA doubling time as a continuous variable in a linear regression model

^**

Obesity was defined as a BMI ≥ 30 kg/m²

^†

Adjusted for age, year of surgery, race, BMI, clinical stage, biopsy Gleason score, center, and pre-operative PSA

^††

p – interaction by race assessed by including the cross product term between diabetes and race in the model

^‡

p – interaction by obesity assessed by including the cross product term between obesity and race in the model