Subject characteristics at enrolment (?1 month) for 100 patients with type-1 diabetes.

Data represent number of patients [%], mean (standard deviation) or median (range).aInformation about daily alcohol intake was missing in 23 and 24 patients allocated metformin and placebo, respectively. The percentage of patients refers to those for whom data were available.bNot compared statistically due to equal proportions of patients or too few patients in treatment groups.cPrevious ischaemic heart disease, stroke, transient ischaemic attack or peripheral arterial disease.dNormo-, micro- and macro-albuminuria: 24-hour urinary albumin-excretion��29 mg, 30�C299 mg and ��300 mg, respectively, in two out of three consecutive samples. One patient, in the metformin group, had a kidney transplant five years prior to enrolment due to polycystic kidney disease. The patient presented with normal levels of creatinine and microalbuminura. Due to having prior kidney disease, the patient is included in the macroalbuminuria group.eSymptomatic peripheral or autonomic neuropathy or clinical signs thereof.fOne patient, in the metformin group, received continuous subcutaneous insulin-infusions. The patient is included in the group with four daily injections.gDuring the run-in period, in the metformin group, one patient changed from thiazid to loop diuretics, one patient initiated statin treatment and two patients initiated dietary (n = 1) or potassium (n = 1) supplementations, respectively, whereas, in the placebo group, one patient was temporarily treated with non-steroid anti-inflammatory drugs. Thus, at baseline (0 months), in the metformin and placebo groups, respectively, a total of 24 (49%) and 14 (27%) patients received ongoing statin treatment (p = 0.039 between treatments); five (10%) and one (2%) patients received dietary supplementations (p = 0.108 between treatments) and 15 (31%) and 12 (24%) patients received potassium supplementations (p = 0.502 between treatments). In contrast, for the whole group of thiazid or loop diuretics as well as for non-steroid anti-inflammatory drugs, despite changes herein during the run-in period, a similar number of patients received such treatments at baseline compared to enrolment. Otherwise, patients did not start and/or stop other non-study medications as listed during the run-in period.hOver-the-counter vitamin pills with unspecified cobalamin and folic acid content.Abbreviation: ACE: Angiotensin converting enzyme.