Table 4.

Associations between metformin medication and incident breast cancer in a prospective population-based study (N = 4,323)

Exposure categoriesCases (n = 393)Controls (n = 3,930)Conditioned on countyaAdjusted for diabetes complications and clinical obesitybMultivariable adjustedcMultivariable adjusted imputed parityd
OR95% CIOR95% CIOR95% CIOR95% CI
No metformin2972,7761.001.001.001.00
Metforminf961,1540.770.61–0.990.820.64–1.080.820.64–1.080.810.63–0.96
 Formerg151960.710.41–1.230.730.42–1.270.720.41–1.260.720.42–1.26
 Recenth819580.780.60–1.010.810.61–1.060.840.64–1.100.830.63–1.08
 5-y metformin duration354180.750.51–1.090.780.54–1.150.820.56–1.210.830.56–1.22
Other medications2191,9781.001.001.001.00
Metforminf961,1540.740.57–0.960.760.58–0.990.780.60–1.030.780.59–1.01
 Formerg151960.640.37–1.120.660.38–1.150.640.37–1.120.810.61–1.08
 Recenth819580.760.58–1.000.780.59–1.030.820.62–1.090.640.36–1.12
 5-y metformin duration354180.710.48–1.060.740.50–1.110.800.53–1.200.800.53–1.31
Diet and exercise onlye787981.001.001.001.00
Metforminf961,1540.890.64–1.240.880.62–1.250.870.60–1.240.870.61–1.25
 Formerg151960.960.52–1.790.920.48–1.740.820.43–1.590.850.44–1.64
 Recenth819580.880.62–1.240.880.61–1.250.870.60–1.260.870.60–1.26
 5-y metformin duration354180.800.48–1.320.780.46–1.330.720.41–1.270.740.43,1.29
  • aConditional logistic regression used to adjust for confounding and selection bias introduced by matching on county.

  • bConditional logistic regression used to adjust for confounding and selection bias introduced by matching on county and adjusting for complications due to diabetes and clinical obesity.

  • cConditional logistic regression used to adjust for confounding and selection bias introduced by matching on county and adjusting for complications due to diabetes, clinical obesity, age at index date, postmenopausal hormone use, and parity at index date.

  • dConditional logistic regression used to adjust for confounding and selection bias introduced by matching on county and adjusting for complications due to diabetes, clinical obesity, age at index date, postmenopausal hormone use, and multiple imputation to impute missing parity.

  • eWomen who were treated with a diet and exercise regimen only.

  • fMetformin users were women who used metformin for a minimum of 1-year duration.

  • gFormer users were women whose last metformin prescription was greater than 1 year from the index date.

  • hRecent users were women whose last metformin prescription was within 1 year before the index date.