High tissue iron is a risk factor for diabetes even within the broad normal range of normal human values. In order to demonstrate the benefits of reducing iron on glucose homeostasis and to better define the parameters for larger clinical trials, the investigators will subject individuals with prediabetes (impaired glucose tolerance, IGT) or early type 2 diabetes to phlebotomy in order to reduce serum ferritin concentrations and determine the effect on glucose homeostasis as revealed by oral and frequently sampled intravenous glucose tolerance testing.

Male and female subjects aged 35-65 will be recruited from the clinic populations at the Medical Centers of the Universities of Colorado, New Mexico, or Utah who have prediabetes, or diabetes relatively well-controlled (Hemoglobin A1c [HbA1c] ≤ 8.0%) with a single oral agent. Diabetes/prediabetes status is verified by oral glucose tolerance testing (OGTT) using criteria of the American Diabetes Association. Serum ferritin will be determined, and subjects chosen with values in the upper 50% of the normal range, 110-400 ng/mL for males, or 80-240 for females. Exclusion criteria include: hereditary hemochromatosis; cancer (except cases currently with no evidence of disease); serum creatinine >1.5; anemia (Hgb < lower limit of normal); chronic inflammatory conditions (rheumatologic conditions such as rheumatoid arthritis or giant cell arteritis, or chronic infections such as hepatitis B or C) that could affect ferritin; erythrocyte sedimentation rate or C-reactive protein>1.5 times the upper limit of normal (UNL); serum transaminases 2 x UNL; hemophilia, warfarin therapy, or history of GI bleeding; and current glucocorticoid therapy.

After determination of eligibility, subjects are randomized in a 2:1 ratio for treatment and controls arms. In the fasted state, blood is drawn for determination of adiponectin and lipids (HDL, direct LDL, triglycerides). Subjects have the option of participating in a frequently sampled intravenous glucose tolerance test (FSIVGTT). All subjects receive counseling to follow a healthy diet and regular exercise. Phlebotomy occurs at the rate of 500 ml (one Unit) of blood per month over the period of 3-6 months. At two-month intervals, serum ferritin and complete blood counts are determined. When serum ferritin reaches the lowest quartile of normal (<50 ng/mL for females and <70 ng/mL for males, but no sooner than 3 months or later than 6 months after beginning phlebotomy, metabolic status will be reassessed by fasting lipids, adiponectin, OGTT, and optionally, FSIVGTT. Controls are recalled for repeat testing 6-8 months after enrollment.

120 min OGTT and FSIVGTT are performed on separate days after an overnight fast. FSIVGTT data are analyzed using MINMOD Millennium software. Areas under the glucose curve (AUCglucose) are calculated using the trapezoidal method.

Inclusion Criteria:

• Age 35-65
• Prediabetes, or diabetes relatively well-controlled (Hemoglobin A1c [HbA1c] ≤ 8.0%) with a single oral agent as defined by oral glucose tolerance testing (OGTT) using criteria of the American Diabetes Association.
• Serum ferritin values in the upper 50% of the normal range, 110-400 ng/mL for males, or 80-240 for females.

Exclusion Criteria:

• hereditary hemochromatosis;
• cancer (except cases currently with no evidence of disease);
• serum creatinine >1.5;
• anemia (Hgb < lower limit of normal);
• chronic inflammatory conditions (rheumatologic conditions such as rheumatoid arthritis or giant cell arteritis, or chronic infections such as hepatitis B or C) that could affect ferritin;
• erythrocyte sedimentation rate or C-reactive protein>1.5 times the upper limit of normal (UNL);
• serum transaminases 2 x UNL;
• hemophilia,
• warfarin therapy,
• history of GI bleeding;
• current glucocorticoid therapy.