New England Journal of Medicine JUNE 12, 2008

Effects of Intensive Glucose Lowering in Type 2 Diabetes

The Action to Control Cardiovascular Risk in Diabetes (ACCORD) Study Group

Bottom Line

In patients with type 2 diabetes at high risk for cardiovascular disease, an intensive glycemic control strategy targeting HbA1c below 6.0% increased all-cause and cardiovascular mortality compared to a standard strategy targeting 7.0–7.9%.

Key Findings

1. The intensive glucose-lowering strategy resulted in a 22% increase in the risk of all-cause mortality compared to standard therapy (Hazard Ratio [HR] 1.22; 95% Confidence Interval [CI] 1.01–1.46; P=0.04).

2. Cardiovascular mortality was significantly higher in the intensive treatment group compared to the standard group (HR 1.35; 95% CI 1.04–1.76; P=0.02).

3. While the composite primary outcome (first occurrence of nonfatal myocardial infarction, nonfatal stroke, or cardiovascular death) did not significantly differ between groups (HR 0.90; 95% CI 0.78–1.04), the intensive strategy was associated with a significant reduction in nonfatal myocardial infarction (HR 0.76; 95% CI 0.62–0.92).

4. Severe hypoglycemia requiring medical assistance occurred more frequently in the intensive treatment group (16.2% vs. 5.1%; P<0.001).

Study Design

Design

RCT

Open-Label

Sample

10,251

Patients

Duration

3.4 yr

Median

Setting

Multicenter, US and Canada

Population Patients aged 40–79 with type 2 diabetes, HbA1c ≥7.5%, and either established cardiovascular disease or high risk of cardiovascular disease based on age and comorbid factors.

Intervention Intensive glycemic control targeting HbA1c <6.0% using various antihyperglycemic agents.

Comparator Standard glycemic control targeting HbA1c 7.0–7.9%.

Outcome A composite of first occurrence of nonfatal myocardial infarction, nonfatal stroke, or cardiovascular death.

Study Limitations

• The trial was terminated early at a median follow-up of 3.4 years due to the observed excess mortality, which may have limited the power to detect long-term cardiovascular benefits.

• The study population was limited to high-risk patients with established cardiovascular disease or significant risk factors, potentially limiting the generalizability to lower-risk patients.

• The intensive intervention was complex and involved rapid escalation of multiple pharmacological agents, making it difficult to isolate whether the risk was due to the A1c target itself, the rate of decline, or specific drug combinations.

• The study was conducted as part of a 2x2 factorial design, and interactions between the glycemic intervention and other concurrent lipid or blood pressure treatments may have influenced outcomes.

Clinical Significance

The results of the ACCORD trial fundamentally shifted clinical guidelines by demonstrating that targeting near-normal HbA1c levels (<6.0%) in high-risk patients with long-standing type 2 diabetes is associated with significant harm, particularly increased mortality. This necessitates a more personalized approach to glycemic targets, prioritizing safety and avoiding severe hypoglycemia in older patients or those with extensive cardiovascular comorbidities.

Historical Context

Prior to ACCORD, the UK Prospective Diabetes Study (UKPDS) had established that intensive glycemic control reduced microvascular complications in patients with newly diagnosed type 2 diabetes. Given the high burden of cardiovascular disease in this population, ACCORD was launched to determine if even more intensive control could reduce macrovascular (cardiovascular) events. The trial's unexpected finding of increased mortality challenged the prevailing dogma that 'lower is always better' for blood glucose management.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

Based on the metabolic response to insulin and counter-regulatory hormones, why might a rapid drop in blood glucose to near-normal levels (HbA1c <6.0%) trigger fatal cardiac arrhythmias in patients with long-standing type 2 diabetes?

Key Response

Intensive insulin therapy increases the risk of severe hypoglycemia. Hypoglycemia triggers a massive sympathoadrenal response, releasing epinephrine which can cause hypokalemia (via intracellular potassium shifts) and prolong the QT interval, both of which are potent triggers for fatal arrhythmias, particularly in a heart already compromised by autonomic neuropathy or CAD.

Resident

Contrast the macrovascular findings of the ACCORD trial with those of the earlier UKPDS trial; how should these differing results influence your selection of a glycemic target for a newly diagnosed 40-year-old versus a 70-year-old with established coronary artery disease?

Key Response

UKPDS showed that early, intensive control in newly diagnosed patients provides a 'legacy effect' of cardiovascular protection. In contrast, ACCORD showed that intensive control in older patients with established CVD or long-standing diabetes increases mortality. Current guidelines (ADA/EASD) reflect this by recommending stringent targets (<6.5-7.0%) for the young/healthy and relaxed targets (7.5-8.0%+) for those with high comorbidity burdens.

Fellow

The ACCORD trial noted a reduction in non-fatal myocardial infarction but an increase in total mortality in the intensive arm. What are the leading theories regarding this divergence, and how does the concept of 'glucose variability' versus 'absolute HbA1c' potentially explain this paradox?

Key Response

While intensive control may reduce the atherosclerotic progression (lowering non-fatal MI), the 'lethal' cost of achieving that target—specifically frequent iatrogenic hypoglycemia and high glycemic variability—may exceed the benefit. Higher glucose flux and frequent 'lows' are associated with increased oxidative stress and inflammation, which may lead to sudden cardiac death, offsetting the reduction in MI.

Attending

ACCORD challenged the 'lower is better' dogma that dominated endocrinology for decades. How do you integrate these findings when teaching trainees about the limitations of surrogate endpoints (HbA1c) versus patient-centered outcomes (mortality)?

Key Response

ACCORD is a landmark lesson in 'medical reversal.' It demonstrates that improving a surrogate marker (HbA1c) does not always translate to improved survival and can sometimes cause harm. Clinicians must prioritize the safety of the intervention over the aesthetics of the lab result, shifting from 'glucocentrism' to a holistic risk-reduction strategy including BP and lipid management.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

The intensive arm of ACCORD was terminated 17 months early. From a statistical standpoint, how does this early termination for harm affect the 'overestimation' of the effect size, and what challenges does it pose for analyzing the secondary outcomes and microvascular benefits?

Key Response

Early termination often results in 'random highs' in effect size estimation. Because the trial stopped early due to the mortality signal, it may have lacked sufficient power to fully evaluate microvascular endpoints or specific drug-drug interactions. It also creates a 'truncation bias,' where the long-term potential benefits of intensive control (which take years to manifest) are systematically undervalued compared to acute harms.

Journal Editor

If reviewing the ACCORD manuscript today, how would you address the potential confounding factor of 'weight gain and polypharmacy' in the intensive group, where patients received significantly more medications and gained more weight than the standard group?

Key Response

A critical reviewer would ask if the mortality was caused by the target (low glucose) or the means (excessive insulin, TZDs, and multiple drug combinations). The rapid escalation of therapy and the resulting metabolic side effects (fluid retention, weight gain) represent a 'treatment-pattern' confounder that makes it difficult to isolate the biological effect of a 6.0% HbA1c level itself.

Guideline Committee

In light of ACCORD, ADVANCE, and VADT, how should the strength of recommendation for 'intensive glycemic control' be phrased for patients with high cardiovascular risk, and which specific ADA/EASD criteria should trigger a relaxation of the HbA1c target to 8.0%?

Key Response

Evidence from these trials moved guidelines away from a universal <7.0% target. Current guidelines (e.g., ADA Standards of Care) now assign a 'Level A' recommendation to individualizing targets. Relaxation to <8.0% is specifically recommended for patients with a history of severe hypoglycemia, limited life expectancy, advanced microvascular/macrovascular complications, or long-standing diabetes where the target is difficult to attain.

Clinical Landscape

Noteworthy Related Trials

1998

UKPDS 33 Trial

n = 3,867 · Lancet

Tested

Intensive glucose control with sulfonylureas or insulin

Population

Newly diagnosed patients with type 2 diabetes

Comparator

Conventional dietary therapy

Endpoint

Any diabetes-related endpoint

Key result: Improved glycemic control substantially decreased the risk of microvascular complications, although it did not show a statistically significant reduction in myocardial infarction.

2008

ADVANCE Trial

n = 11,140 · NEJM

Tested

Intensive glucose control (target HbA1c < 6.5%)

Population

Patients with type 2 diabetes at high cardiovascular risk

Comparator

Standard glucose control

Endpoint

Major macrovascular and microvascular events

Key result: Intensive glucose control reduced the incidence of combined major macrovascular and microvascular events, primarily driven by a reduction in nephropathy, but did not significantly reduce major cardiovascular events or mortality.

2009

VADT Trial

n = 1,791 · NEJM

Tested

Intensive glucose control targeting HbA1c < 6.0%

Population

Veterans with poorly controlled type 2 diabetes and high cardiovascular risk

Comparator

Standard glucose control

Endpoint

Major cardiovascular events (MACE)

Key result: Intensive glucose-lowering therapy did not significantly reduce the rate of major cardiovascular events or cardiovascular death in this high-risk population.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis