Frailty is a geriatric syndrome which leads to poor health outcomes in older adults, such as falls, disability, hospitalization, institutionalization, and death. Due to the dramatic growth in the U.S. aging population and the health care costs associated with frailty (estimated at more than $18 billion per year), frailty is a major health care problem. There has been little research into potential pharmacologic interventions that would delay or reduce the incidence of frailty. Thus, the major goal of this study is to test metformin as a novel intervention for the prevention of frailty. The investigators propose that diabetes/insulin resistance and inflammation are major contributors to frailty, and that the use of metformin to modulate diabetes/insulin resistance and inflammation will prevent and/or ameliorate the progression of frailty.

Physical frailty is a geriatric syndrome that leads to poor health outcomes such as falls, disability, institutionalization, and death. The prevalence of frailty is estimated to be 7-15% among community-dwelling older U.S. adults. The associated costs of frailty were estimated to be more than $18 billion in 2000 and these will continue to increase over the next two decades. Thus, an increasingly frail older population will have major implications for the demand for health care services, including hospital usage, home care, and long-term care.

Data from several studies have suggested strong roles for diabetes and insulin resistance, which are associated with increased inflammation, in the physiological basis of frailty. The investigators' recent epidemiological research with a community-based population of older adults showed that diabetes was the most significant predictor of frailty onset and worsening over time. While the importance of frailty and its impact on an aging U.S. society have been widely recognized, to date there are no effective interventions to prevent or treat frailty. Therefore, the major goal of this study is to test a drug with insulin-sensitizing and anti-inflammatory properties, such as metformin, as a novel intervention for frailty prevention.

The investigators hypothesize that metformin will lead to reduced inflammation and insulin resistance present in older glucose-intolerant subjects and that these changes will consequently prevent the onset and/or progression of frailty in this sub-population of older adults. Subjects with impaired glucose intolerance will be enrolled in this study because this group encompasses approximately 1/3rd of the older population, this group is at increased risk for developing diabetes and frailty, and is the most likely to benefit from a potential anti-inflammatory and insulin-sensitizing intervention.

Inclusion Criteria:

• Men and women
• All ethnic groups
• Age 65 and older
• Community-dwelling
• Pre-diabetic based on oral glucose tolerance test with 2 hour values of 140 - 199 mg/dL after an oral glucose load, and no diagnosis of diabetes in the past 12 months
• Subjects must have the following laboratory values: Hematocrit ≥ 33%, aspartate aminotransferase < 2 X upper limit of normal, alanine aminotransferase < 2 X upper limit of normal, alkaline phosphatase < 2 X upper limit of normal, normal urinalysis (no clinically significant white blood cells, red blood cells or bacteria), platelets ≥ 100,000, prothrombin time < 15 seconds and partial thromboplastin time < 40 seconds, glomerular filtration rate (GFR) ≥ 45ml/min and urine protein ≤ 100 mg/dL by lab urinalysis.

Exclusion Criteria:

• Characterized as frail, defined as the presence of 3 or more of: 1) weak hand grip strength, 2) slow walking speed, 3) low physical activity, 4) unintentional weight loss of ≥ 10 pounds over the past year, 5) self-reported exhaustion
• Resident of nursing home or long-term care facility
• Subjects with diabetes with range fasting glucose in diabetes range (≥ 126 mg/dL), or 2-hour glucose within diabetes range on OGTT (≥ 200 mg/dL).
• Subjects taking drugs known to affect glucose homeostasis
• Untreated depression or Geriatric Depression Scale (GDS) score on 15-item scale >7
• Diagnosis of any disabling neurologic disease Parkinson's Disease, Amyotrophic Lateral Sclerosis, multiple sclerosis, cerebrovascular accident with residual deficits (muscle weakness or gait disorder), severe neuropathy, diagnosis of dementia or Mini-mental State Exam (MMSE) score <24, cognitive impairment due to any reason such that the patient is unable to provide informed consent.
• History of moderate-severe heart disease (New York Heart Classification greater than grade II) or pulmonary disease (dyspnea on exertion upon climbing one flight of stairs or less; abnormal breath sounds on auscultation)
• Poorly controlled hypertension (systolic >160 mmHg, diastolic >100 mmHg)
• Peripheral arterial disease (history of claudication)
• Moderate to severe valvular heart disease
• Subjects who have been treated with long term (>30 days) systemic steroids, anabolic steroids, growth hormone or immunosuppresants within the last 6 months. Males with a medical history of testosterone deficiency who are on a stable dose of testosterone replacement (for ≥ 3 months) are allowed.
• Subjects who have been treated with short term (<30 days) systemic steroids, anabolic steroids, growth hormone or immunosuppressants within the last 1 month.
• Chronic inflammatory condition, autoimmune disease, or infectious processes (e.g., active tuberculosis, Human Immunodeficiency Virus, rheumatoid arthritis, systemic lupus erythematosus, acute or chronic hepatitis B or C)
• Active tobacco use (within 6 months)
• Illicit drug use
• Active malignancy, non-skin
• Disease or condition likely to cause death within 5 years
• Hypersensitivity to metformin or pioglitazone
• Any disease or condition considered to be exclusionary based on the clinical opinion and discretion of the PI

Adverse events were collected from baseline to 2 years.

All cause mortality and serious adverse events, related and unrelated to study intervention, are reported.