Tight Glycemic Control May Not Be Best in Type 2 Diabetes

[Raptor]

Tight Glycemic Control May Not Be Best in Type 2 Diabetes

April 20, 2009 — Tight glycemic control may not be best in patients with type 2 diabetes, according to the results of a review and critique of recent large randomized trials reported online April 20 in the Annals of Internal Medicine. In addition to summarizing findings from trials that evaluated tight glycemic control in patients with type 2 diabetes, the article offers practical, evidence-based suggestions for management.

"Some diabetes guidelines set low glycemic control goals for patients with type 2 diabetes mellitus (such as a hemoglobin A level as low as 6.5% to 7.0%) to avoid or delay complications," write Victor M. Montori, MD, MSc, from the Mayo Clinic in Rochester, Minnesota, and Mercè Fernández-Balsells, MD, from Hospital Universitari de Girona Doctor J. Trueta in Girona, Spain. "Our review and critique of recent large randomized trials in patients with type 2 diabetes suggest that tight glycemic control burdens patients with complex treatment programs, hypoglycemia, weight gain, and costs and offers uncertain benefits in return. We believe clinicians should prioritize supporting well being and healthy lifestyles, preventive care, and cardiovascular risk reduction in these patients."

The reviewers examined findings from large randomized trials comparing clinical outcomes for patients with type 2 diabetes managed using tight vs less tight glycemic targets. Trials of interventions designed to reduce multiple risk factors, to test the efficacy of a single antihyperglycemic agent vs placebo, or to evaluate glycemic durability were not included.

Except for the United Kingdom Prospective Diabetes Study (UKPDS) metformin trial, findings from the other trials indicate that all-cause or cardiovascular death, stroke, amputations, and microvascular complications are not decreased by tight glycemic control. However, effect estimates were imprecise, because few patients developed complications, and effects were heterogeneous across trials, making the findings inconclusive.

The risk for nonfatal myocardial infarction might be lowered by about 16% by intensive glycemic control, based on reported findings from all trials, at the expense of doubled or tripled risk for severe hypoglycemia. The incidence of hypoglycemia was also highest in trials with the lowest glycosylated hemoglobin (HbA1c) target levels. In all but the metformin vs conventional comparison in the UKPDS trial, intensive glycemic control was also associated with a 2% weight gain.

Low hemoglobin A (HbA) has been not only a clinical goal but also a quality measure for healthcare, with conventional wisdom advocating that to reduce diabetes complications, routine management of type 2 diabetes should aim for tight glycemic control. Instead, the reviewers suggest that optimal glycemic control requires individualizing hemoglobin A1c targets, so that those targets and the interventions needed to achieve them reflect patient-specific factors including their individual preferences.

Promoting patient well-being and healthy lifestyles, preventive care, and cardiovascular risk reduction should be key priorities for patients with type 2 diabetes, the reviewers recommend. Evidence reviewed from randomized trials does not strongly suggest that tight glycemic control is more beneficial with regard to lowering the risk for diabetes complications than it is harmful.

In certain circumstances, such as patients with newly diagnosed diabetes, early tight glycemic control may be appropriate. However, tight control has been shown to cause increases in patient burden, cost, and the harm of serious hypoglycemia. Therefore, further studies are needed to confirm or refute the potential net benefit of tight control before patients are subjected to the demonstrated harms.

"Given that patients with diabetes often have comorbid conditions, clinicians should avoid glycemic control interventions that overwhelm the patients' capacity to cope clinically, psychologically, and financially," the review authors write. "Tight disease-centered goals that require highly complex and burdensome treatment programs may promote frustration, nonadherence, and financial stress in some patients. For instance, many patients will not benefit and could reduce or eliminate glucose self-monitoring."

Glucose levels greater than 10 µmol/L (>180 mg/dL) may produce symptoms, whereas medication is associated with risk for hypoglycemia. For many patients, a reasonable and feasible

HbA1c target is between 7% and 7.5% (estimated average glucose level, 8.5 – 9.5 µmol/L [150 – 160 mg/dL]). To achieve this goal for other patients, such as those with severe insulin deficiency, management may need to be intensive, requiring interventions such as physiologic insulin dosing (basal-bolus regimens) and intense monitoring. Once clinicians and patients have chosen an individualized, optimal HbA1c target, they should design a management regimen to achieve that target.

Because glycemic targets should be individualized, HbA1c targets for clinical use should not be the same as those used to measure quality of care. If HbA1c is used as a performance measure, an upper limit — for example, HbA1c level greater than 9% — may be more appropriate to indicate possibly inadequate care than one that would fail to consider individual patient needs and goals (such as HbA1c level < 7%).

"Because we cannot confidently distinguish the relative effectiveness of different diabetes medications in reducing complications, we recommend basing their selection on such factors as burden of administration and side effects," the reviewer authors conclude. "We have developed and are studying tools to promote patient involvement in choosing diabetes medications. We hope that tools and tactics that encourage patient involvement in treatment decisions prove to be effective and lead to treatment programs that are both evidence-based and consistent with patients' context, values, and preferences."

Dr. Montori received a Clinician Investigator grant from the American Diabetes Association in 2004. Novo Nordisk, a maker of insulin, subvented the American Diabetes Association granting program. Dr. Fernández-Balsells received grants from Instituto de Salud Carlos III, Ministerio de Sanidad y Consumo, Government of Spain.

Ann Intern Med. Published online April 20, 2009.