Does the intact nephron hypothesis provide a reasonable model for metformin dosing in chronic kidney disease?