Table 3.

Association of Metformin Treatment (1 Defined-Daily-Dose Increment) With Incidence of All-Sites Cancer and Specific Cancers Among Israelis With Diabetes, Controlling for Use of All Other Glucose-Lowering Medications^a and Adjusting for Confounding Variables^b, 2004–2012

Cancer Site	No. at Riskc	No. of Cancer Events	Period of Metformin Treatment Previous to the Current Quarterd
			Years 2–4		Years 2–7
			HR	95% CI	HR	95% CI
All sitese	294,770	11,898	0.96	0.82, 1.12	0.98	0.82, 1.18
Colon	310,698	2,131	1.13	0.79, 1.63	1.05	0.67, 1.63
Bladder	313,133	764	0.91	0.50, 1.68	0.98	0.49, 1.97
Lung	313,460	1,265	0.85	0.53, 1.38	1.02	0.59, 1.78
Breast (women only)	163,461	1,835	0.95	0.64, 1.40	0.88	0.56, 1.39

Abbreviations: CI, confidence interval; HR, hazard ratio.

^a Adjusted for use of insulin, α-glucosidase inhibitor, rosiglitazone, sulfonylureas, dipeptidyl peptidase-4 inhibitors, glucagon-like peptide-1 receptor agonists, and meglitinides.

^b Confounding variables included age, sex, socioeconomic status, ethnic origin, smoking (for bladder and lung cancers), and parity (for breast cancer).

^c Numbers reflect numbers of patients at risk for the particular cancer at the time they were diagnosed with diabetes (and were without any previous cancer diagnosis), excluding those who completed follow-up within 2 years of their diabetes diagnosis.

^d The first year prior to the current period was excluded, because an undiagnosed cancer could cause perturbations in glucose levels, particularly in the year prior to diagnosis.

^e Without prostate and pancreatic cancers.