Table 2.

Association between metformin and other anti-diabetic medication use and prostate cancer risk

Variable	None anti-diabetic medications	Non-metformin anti-diabetic medications			Metformin
	OR	OR	95% CI	P-value	OR	95%CI	P-value
Overall prostate cancer risk
No. with cancer/total	43/205		30/141			49/194
Univariable	Referent	1.02	0.60–1.72	0.946	1.27	0.80–2.03	0.311
Multivariable*	Referent	0.85	0.48–1.48	0.560	1.19	0.72–1.99	0.495
Disease grade, low grade#
No. with cancer/total**	33/195		20/131			31/176
Univariable	Referent	0.88	0.48–1.62	0.691	1.05	0.61–1.80	0.860
Multivariable*	Referent	0.79	0.42–1.49	0.463	1.01	0.57–1.81	0.961
Disease grade, high grade#
No. with cancer/total***	10/172		10/121			18/163
Univariable	Referent	1.46	0.59–3.62	0.415	2.01	0.90–4.50	0.089
Multivariable*	Referent	1.23	0.46–3.26	0.678	1.83	0.75–4.46	0.185

OR, odds ratio

CI, confidence interval

^*Adjusted for age, race, geographic region, prostate specific antigen levels, digital rectal examination findings, body mass index, prostate volume, family history of prostate cancer, coronary artery disease, smoking status, NSAIDs, statins, aspirin, and treatment group

^**Numbers reflect men included in the analysis: those with low-grade disease and those without cancer

^***Numbers reflect men included in the analysis: those with high-grade disease and those without cancer

^#Multinomial regression was used to model the outcomes of low vs. no prostate cancer and high grade disease vs. no prostate cancer.