The primary objective of this study is:

• To investigate whether insulin aspart with meals is better than a standard treatment with insulin NPH at bedtime, evaluated by HbA1c.

The secondary objectives of this study are:

• To study if a combination treatment with metformin and/or rosiglitazone and insulin aspart with meals is better than a standard treatment with insulin NPH combined with one or more of the above oral antidiabetic drugs. According to the hypothesis, special focus will be given to the treatment group with insulin aspart combined with metformin and rosiglitazone. The treatment effect will be evaluated by HbA1c.

• To examine the effects of the treatments on glucose metabolism and beta cell function, evaluated by diurnal blood glucose, fasting plasma glucose, insulin, C-peptide, and lactate.

• To examine the effects of the treatments on cardiovascular risk factors evaluated by serum lipid profiles, serum free fatty acids, urine albumin/creatinine ratio, and electrocardiogram (ECG).

• To quantify and describe the patients' subjective experiences of the two different insulin treatments (quality of life assessment)

• To examine patients with type 2 diabetes for the presence of variability in a series of genes, which are known to or are assumed to:

  • affect the long term outcome;
  • determine the responsiveness to treatment with diet, exercise and drugs targeting the known risk markers for late diabetic complications; and
  • after intervention, to analyse the complex interrelationships between genotypes and clinical endpoints and the responsiveness to actual treatment modalities.

Inclusion Criteria:

• Ages between 30 and 70 years
• Fasting C-peptide >300 pmol/l
• Body mass index (BMI) > 25 kg/m2
• Diabetes for more than 2 years
• Pharmacological antidiabetic treatment for more than 3 months
• 7.0%<HbA1c<12.0% at randomisation
• Patient willing to sign informed consent
• Fertile women: negative pregnancy test and use of oral or intra-uterine contraception or depot gestagen.

Exclusion Criteria:

• S-creatinine > 120 μmol/l
• History of intolerance to metformin or glitazones
• S-ALAT/S-ASAT > 2.5 x upper normal limit
• Total cholesterol > 10 mmol/l
• Total triglyceride > 8 mmol/l
• Hemoglobin (Hb) < normal range
• Treatment with glitazone preceding 30 days New York Heart Association (NYHA) functional class III or IV
• Night work
• Present or planned pregnancy
• Poor vision impeding insulin administration
• Unawareness of hypoglycaemia (complete or partly)
• Mental illness or alcohol abuse
• Clinically relevant major organ or systemic illness
• Uncontrolled hypertension >180/110 mmHg, systolic or diastolic
• Steroid treatment
• Severe lung disease
• A history of malign disease
• An expectation that the patient will not be collaborative or will not be able to understand the character of this trial