Table 3.

ORs for the association of pioglitazone exposure and prostate cancer

Exposure definitions	Cases (n=756)	Controls (n=2942)	Crude OR (95% CI)	Adjusted* OR (95% CI)
Pioglitazone exposure
Never	714	2740	reference	reference
Ever	42	202	0.800 (0.560 to 1.143)	0.759 (0.502 to 1.148)
Time since initiating pioglitazone (years)
Never			reference	reference
<4.5	31	160	0.747 (0.499 to 1.118)	0.640 (0.401 to 1.023)
4.5–8	10	38	1.023 (0.498 to 2.099)	1.388 (0.612 to 3.147)
>8	1	4	0.990 (0.098 to 10.010)	1.569 (0.126 to 19.500)
Test for trend, p			0.371	0.535
Duration of pioglitazone (months)
Never			reference	reference
<12	15	87	0.659 (0.375 to 1.159)	0.504 (0.267 to 0.949)
12–23	6	47	0.503 (0.213 to 1.188)	0.559 (0.211 to 1.479)
24–35	12	22	2.147 (1.039 to 4.435)	2.035 (0.820 to 5.050)
36–59	6	34	0.699 (0.287 to 1.705)	0.822 (0.302 to 2.235)
≥60	3	12	0.958 (0.263 to 3.483)	1.984 (0.454 to 8.679)
Test for trend, p			0.627	0.970
Cumulative pioglitazone dose (mg)
Never			reference	reference
1–9000	14	91	0.587 (0.330 to 1.045)	0.443 (0.234 to 0.837)
9001–25 000	15	60	0.984 (0.555 to 1.741)	1.265 (0.640 to 2.501)
25 001–50 000	10	32	1.201 (0.580 to 2.488)	1.151 (0.492 to 2.693)
≥50 001	3	19	0.607 (0.175 to 2.104)	1.004 (0.252 to 3.996)
Test for trend, p			0.514	0.808

*Adjusted for: most recent PSA screening test result prior to index date; history of BPH drug treatment (α-blockers) at index date.

BPH, benign prostatic hyperplasia; PSA, prostate-specific antigen.