ACP: Best practice for diabetes is diet, then metformin, then any second drug

New diabetes guidelines from the American College of Physicians recommend prescribing a drug when lifestyle changes don't lower hyperglycemia. Specifically, start with metformin first and then add any second oral agent.
The three recommendations are:

• Add oral pharmacologic therapy in patients diagnosed with type 2 diabetes when diet, exercise and weight loss fail to improve hyperglycemia (Grade: strong recommendation; high-quality evidence). There are no best data on when to start drugs, so consider life expectancy, whether there are vascular complications, and the risk for adverse events.
• Prescribe monotherapy with metformin for initial pharmacologic therapy to treat most patients with type 2 diabetes (Grade: strong recommendation; high-quality evidence). It's cheaper than most other drugs, is more effective, has fewer side effects and does not result in weight gain.
• Add a second agent to metformin to treat patients with persistent hyperglycemia when lifestyle modifications and monotherapy with metformin fail to control hyperglycemia (Grade: strong recommendation; high-quality evidence). Adding a second drug can lower hyperglycemia by about 1 more percentage point. No evidence supports using one combination therapy over another.

The guideline is based on a 2011 review of the literature from 1966 through April 2010. It expands on a 2007 evidence report from the Agency for Healthcare Research and Quality that discussed drug therapy's effect on mortality, microvascular and macrovascular outcomes, intermediate outcomes and adverse effects. The 2011 review focuses on head-to-head comparisons and includes direct comparisons for monotherapy and dual therapy regimens. The College's recommendations appear in the Feb. 7 Annals of Internal Medicine.
In developing the recommendations, experts considered outcomes including all-cause mortality, cardiovascular morbidity and mortality, cerebrovascular morbidity, neuropathy, nephropathy and retinopathy.
Most diabetes medications reduced A1c levels to a similar degree, the authors noted. In most cases, metformin was more effective than other medications as a monotherapy and when used in combination therapy to reduce A1c levels, body weight, and plasma lipid levels. But it was difficult to draw conclusions about the comparative effectiveness of type 2 diabetes medications on all-cause and cardiovascular mortality, cardiovascular and cerebrovascular morbidity, and microvascular outcomes because of low-quality or insufficient evidence, the guideline writers said.
High-quality evidence shows that the risk for hypoglycemia with sulfonylureas exceeds the risk with metformin or thiazolidinediones and that the combination of metformin plus sulfonylureas is associated with six times more risk for hypoglycemia than the combination of metformin plus thiazolidinediones, the authors noted.
Moderate-quality evidence shows that the risk for hypoglycemia with metformin and thiazolidinediones is similar. But metformin is associated with an increased risk for gastrointestinal side effects, thiazolidinediones are associated with an increased risk for heart failure, and rosiglitazone and pioglitazone are contraindicated in patients with serious heart failure.