• Cellularity of subcutaneous adipose variation tissue at 6 month [ Time Frame: 6 months ]
• HbA1c, lipid level, adiponectin, CRP variation at 6 month [ Time Frame: 6 months ]
• inflammation gene expression in sub-cutaneous fat [ Time Frame: 6 months ]

• Cellularity of subcutaneous adipose variation tissue at 6 month
• Hepatic adipose tissue variation at 6 month
• HbA1c, lipid level, adiponectin, CRP variation at 6 month

In type 2 diabetic patients with poor glycemic control despite maximum "classic" oral treatment, bed time insulin therapy may lead to a parallel increase in abdominal visceral and subcutaneous fat, whereas pioglitazone treatment should lead to a stability (or even a decrease ) in visceral and an increase in subcutaneous abdominal fat. As visceral fat mass is correlated with insulin-resistance and cardio-vascular risk, the evolution of visceral abdominal fat in type 2 diabetic patients is of great importance.

Main objective:

To compare visceral and subcutaneous abdominal fat compartment after a six-month bed time insulin or pioglitazone treatment in type 2 diabetic patients with poor glycemic control despite a maximal oral treatment with metformin and sulfonylureas.

The study hypothesis is that quantity of visceral and subcutaneous abdominal adipose tissue should differently evolute comparing a 6 month treatment with pioglitazone® (30 or 45mg/j) or NPH " bed-time " insulin (0.2u/kg/

In type 2 diabetic patients with poor glycemic control despite maximum "classic" oral treatment, bed time insulin therapy may lead to a parallel increase in abdominal visceral and subcutaneous fat, whereas pioglitazone treatment should lead to a stability (or even a decrease ) in visceral and an increase in subcutaneous abdominal fat. As visceral fat mass is correlated with insulin-resistance and cardio-vascular risk, the evolution of visceral abdominal fat in type 2 diabetic patients is of great importance.

The study hypothesis is that quantity of visceral and subcutaneous abdominal adipose tissue should differently evolute comparing a 6 month treatment with pioglitazone® (30 or 45mg/j) or NPH " bed-time " insulin (0.2u/kg/

• Drug: UMULINE NPH
• Drug: pioglitazone

• Active Comparator: UMULINE NPH at bed time
• Experimental: pioglitazone 30 mg

Inclusion Criteria:

• Type 2 diabetes
• BMI= 26kg/m2
• Maximal treatment with metformin and sulfonylurea
• HbA1c between 7.5 and 9.5%

Exclusion Criteria:

• Anterior treatment with glitazones
• Anterior treatment with insulin
• Known heart failure
• Hepatopathy
• Renal filtration less than 60ml/min, Hb<10g/dl
• Corticoids treatment