We recently observed that the course of glomerular filtration rate (GFR) rapidly declines in a subgroup of Type 2 diabetic patients (D) with abnormalities of albumin excretion rate (AER) and typical diabetic nephropathy, despite tight blood pressure control. The aim of this study was to evaluate whether amelioration of blood glucose control, using insulin, improves the course of GFR. GFR decay was measured by spline modeling analysis of the plasma clearance rate of 51CR-EDTA, assessed every 6 months. We identified two groups of D using morphometric analysis of renal biopsy, who had values of glomerular basement membrane (GBM) and fractional mesangial volume (Vv mes/glom) respectively below (Group A: 38) or above (Group B: 50) the mean+2SD of values found in 27 kidney donors (GBM: 389 nm; Vv mes/glom: 0.25), as previously described in detail. Median AER was similar at base line in the 2 groups (109 microg/min, 29-1950, in Group A, 113 microg/min, 37-1845, in Group B; n.s.). Conventional metabolic therapy (sulphonylureas and/or biguanides) was used both in Group A and B during a 3 year follow-up period (Period 1). Group B was further divided in two subgroups with body mass index below (Group B, a) and above (Group B, b) the value of 30 kg/m2. Mean +/- SD HbA1c was 8.2 +/- 1.6% in Group A, 8.3 +/- 1.7% in Group B (a) (n.s.) and 9.1 +/- 1.7% in Group B (b) (n.s.). Tight blood pressure control was achieved and maintained using angiotensin converting enzyme inhibitors and/or beta blockers and/or calcium antagonists and/or thiazides. The mean arterial blood pressure (MAP) was 92 +/- 3 mmHg in Group A and 91 +/- 4 mmHg in Group B (n.s.). GFR decay was significantly greater in Group B than in Group A (Group A vs B: +1.21 +/- 0.71 vs -5.86 +/- 1.61 ml/min/1.73 m2/year). Median AER significantly rose in Group B (177 microg/min, p<0.05 vs base line) but not in Group A (134 microg/min, n.s.) during the third year of follow-up. Groups A and B were then followed over 4.1 years (range 3.1-4.4) (Period 2) maintaining the above described antihypertensive regimen, resulting in MAP values similar to those described during Period 1. Group A patients were treated with the same conventional glycemic control during Period 2. Group B (a) was conversely treated with intensive insulin therapy to achieve a HbA1c value below 7.5% (3 daily injections of regular and 1 or 2 daily injections of intermediate acting insulin associated with metformin 500 mg twice daily in 64% of the patients). Group B (b) patients were only treated by metformin (850 mg thrice daily) to achieve a HbA1c value below 7.5%. HbA1c decreased below the 7.5% target value in Group B (a) (7.0 +/- 1.6%, p<0.01 vs Period 1), but not in Group B (b) (8.0 +/- 1.6%, p<0.05 vs Period 1) and in Group A (8.3 +/- 1.7%, n.s. vs Period 1). The GFR decay of Group B, a during Period 2 was lower than that during Period 1 (Period 1 vs Period 2: -5.9 +/- 1.8 vs -1.8 +/- 0.7 ml/min/1.73 m2/year, p<0.01). GFR decay during Period 2 was similar to that observed during Period 1 in Group A (Period 1 vs Period 2: +1.21 +/- 0.71 vs +0.7 +/- 0.6 ml/min/1.73 ml/year, n.s.) and in Group B (b) (Period 1 vs Period 2: -4.4 +/- 0.71 vs -4.2 +/- 0.6 ml/min/1.73 m2/year, n.s.). Median AER did not significantly change in the fourth year of Period 2 , either in Group A or B (Group A vs B: 141 vs 152 microg/min, n.s.). In conclusion, our findings seem to suggest that amelioration of blood glucose control is attained both by insulin and metformin intensive treatment, but only insulin decreases and maintains HbA1c levels below 7.5%. These pattens of HbA1c appear to be a threshold value in order to significantly blunt GFR decay in a subgroup of Type 2 diabetic patients with typical diabetic glomerular lesions, who are less responsive to tight blood pressure control alone. Conversely, the cohort of patients with less severe diabetic glomerulopathy steadily show constant GFR patterns, despite similar abnormalities of albumin excretion rate, and HbA1c average values above 7.5%.