Transfer of metformin into human milk.

Objectives. Our objectives were to determine the milk-to-plasma ratio of metformin in lactating mothers and to estimate infant exposure. Methods. Two studies were performed. In study 1, 3 nursing mothers taking metformin were studied through-out a dosing interval at steady state. Blood samples were obtained from 2 suckling infants. In study 2, 5 healthy lactating women who volunteered to express milk after weaning were given metformin, 500 mg, at weaning and were studied for up to 72 hours. In both studies, areas under the plasma and milk concentration-time curves were estimated, and the milk-to-plasma concentration ratio based on area under the concentration-time curve analysis was derived. The infant dose was calculated by standard methods. Results. In study 1 the milk-to-plasma concentration ratios based on area under the concentration-time curve analysis were 0.37, 0.50, and 0.71. The estimated "doses" of metformin that would be ingested by the breast-fed infants were 0.18%, 0.20%, and 0.21% of the maternal doses, adjusted for weight. In the breast-fed infants, no metformin was detected (n = 2) or adverse effects noted (n = 3). In study 2, the milk-to-plasma concentration ratio based on area under the concentration-time curve analysis was unable to be calculated for 3 subjects because of the unexpected persistence of metformin in milk beyond the study period. For the 2 subjects studied for 72 hours, the milk-to-plasma concentration ratios based on area under the concentration-time curve analysis were 0.27 and 0.47 and the infant doses were 0.11% and 0.25%. The concentration-time profile for metformin in milk in all subjects was unexpectedly flat. Conclusions. Metformin appears to be "safe" during lactation because of low infant exposure. The unusual concentration-time profile for metformin in milk suggests that the transfer of metformin into milk is not solely dependent on passive diffusion.