New England Journal of Medicine April 29, 2010

Effects of Intensive Blood-Pressure Control in Type 2 Diabetes Mellitus

The ACCORD Study Group (William C. Cushman et al.)

Bottom Line

In high-risk patients with type 2 diabetes, targeting a systolic blood pressure of less than 120 mm Hg, compared to less than 140 mm Hg, did not significantly reduce major cardiovascular events but did decrease stroke incidence at the cost of increased adverse events.

Key Findings

1. At 1 year, mean systolic blood pressure (SBP) was successfully lowered to 119.3 mm Hg in the intensive-therapy group versus 133.5 mm Hg in the standard-therapy group.
2. The primary composite outcome (nonfatal MI, nonfatal stroke, or cardiovascular death) occurred at an annual rate of 1.87% in the intensive group versus 2.09% in the standard group, showing no significant difference (HR 0.88; 95% CI, 0.73 to 1.06; P=0.20).
3. There was no significant difference in all-cause mortality, with annual rates of 1.28% in the intensive group and 1.19% in the standard group (HR 1.07; 95% CI, 0.85 to 1.35; P=0.55).
4. Intensive blood pressure control significantly reduced the annual rate of stroke, a prespecified secondary outcome, from 0.53% to 0.32% (HR 0.59; 95% CI, 0.39 to 0.89; P=0.01).
5. Serious adverse events attributed to antihypertensive therapy (such as hypotension, hypokalemia, and elevated serum creatinine) were significantly more frequent in the intensive-therapy group (3.3%) compared to the standard-therapy group (1.3%) (P<0.001).

Study Design

Design
RCT
Open-Label
Sample
4,733
Patients
Duration
4.7 yr
Median
Setting
Multicenter, North America
Population Patients with type 2 diabetes mellitus at high risk for cardiovascular events, defined as having clinical or subclinical cardiovascular disease, or at least two additional major cardiovascular risk factors.
Intervention Intensive antihypertensive therapy targeting a systolic blood pressure (SBP) of less than 120 mm Hg.
Comparator Standard antihypertensive therapy targeting a systolic blood pressure (SBP) of less than 140 mm Hg.
Outcome A composite of nonfatal myocardial infarction, nonfatal stroke, or death from cardiovascular causes.

Study Limitations

The overall cardiovascular event rate in the trial was lower than originally anticipated, meaning the study might have been underpowered to detect a small but potentially clinically meaningful cardiovascular benefit.
The mean follow-up of 4.7 years may have been too brief to fully realize the long-term benefits of intensive blood pressure control.
The complex 2x2 factorial design with the ACCORD glycemia trial could have introduced confounding, especially since the intensive glycemia arm was prematurely halted due to increased mortality.
The open-label nature of the blood pressure intervention may have influenced physician behavior, patient adherence, and the reporting of adverse side effects.

Clinical Significance

The ACCORD-BP trial profoundly influenced clinical guidelines by demonstrating that aggressive blood pressure targets (SBP <120 mm Hg) do not confer a broad cardiovascular or survival benefit over standard targets (<140 mm Hg) for most patients with type 2 diabetes. It highlighted that pushing for lower blood pressure thresholds in diabetics yields diminishing returns and increases the risk of medication-related adverse events, though it may still be considered in highly selected patients specifically at high risk for stroke.

Historical Context

Before ACCORD-BP, widespread epidemiological data drove a 'lower is better' paradigm for blood pressure targets in diabetic patients, yet rigorous randomized controlled trials validating targets below 130 or 120 mm Hg were lacking. ACCORD was a monumental, NIH-funded factorial trial examining intensive glycemic, lipid, and blood pressure control. When ACCORD-BP returned neutral primary results for BP targets, it reshaped diabetes care. Its findings later stood in sharp contrast to the 2015 SPRINT trial, which showed a definitive mortality benefit for targeting SBP <120 mm Hg in non-diabetic patients, creating a notable bifurcation in modern hypertension guidelines based on diabetic status.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

The ACCORD-BP trial found an increase in adverse events like elevated serum creatinine in the intensive blood pressure control group. Physiologically, why might aggressively lowering systolic blood pressure below 120 mm Hg in patients with long-standing diabetes lead to an acute decline in glomerular filtration rate (GFR)?

Key Response

Long-standing diabetes causes hyaline arteriolosclerosis, which can stiffen the renal afferent arterioles. To maintain GFR, the kidney relies on adequate systemic perfusion pressure. Aggressive systemic BP lowering reduces this perfusion pressure. In the setting of stiff, non-compliant arterioles and concurrent RAAS blockade (which dilates the efferent arteriole), hydrostatic pressure in the glomerulus drops, leading to an acute decline in GFR and a corresponding rise in serum creatinine.

Resident
Resident

ACCORD-BP showed no significant benefit in the primary composite cardiovascular outcome but did demonstrate a significant reduction in stroke risk for the intensive control group. How should this isolated secondary outcome benefit influence your shared decision-making with a diabetic patient when titrating antihypertensive therapy?

Key Response

While the stroke risk reduction is notable, residents must weigh it against the primary outcome failure and the significant increase in adverse events (hypotension, syncope, hyperkalemia, and AKI) seen in the intensive group. Clinicians should reserve more aggressive targets for patients at exceptionally high risk for stroke who concurrently have a low risk for falls, polypharmacy complications, or acute kidney injury, demonstrating a nuanced, patient-centered approach rather than rigid algorithmic care.

Fellow
Fellow

The ACCORD-BP trial failed to show a primary outcome benefit for a systolic target <120 mm Hg in diabetics, whereas the later SPRINT trial showed a clear mortality benefit for the same target in non-diabetics. How do you reconcile these conflicting results, and what role did trial design and baseline event rates play in this discrepancy?

Key Response

The discrepancy is largely attributed to statistical power and population differences. ACCORD-BP had a 2x2 factorial design and an overall event rate that was much lower than anticipated, rendering the blood pressure arm underpowered to detect a difference in the primary composite endpoint. Additionally, diabetics often have distinct vascular pathology (autonomic dysfunction, extreme vascular stiffness) that alters the risk-benefit ratio of aggressive BP lowering. SPRINT explicitly excluded diabetics partially based on ACCORD's findings, highlighting the necessity of contextualizing BP targets within specific disease states and trial statistical power.

Attending
Attending

Given the ACCORD-BP results demonstrating harm without primary composite benefit at a target <120 mm Hg, how do you teach your clinical team to recognize the 'J-curve' phenomenon in older diabetic patients, and at what point do you actively mandate deprescribing antihypertensives?

Key Response

The J-curve hypothesis suggests that excessive lowering of BP (especially diastolic) can compromise coronary perfusion, paradoxically increasing cardiovascular events. In older diabetics with stiff vessels (creating a wide pulse pressure), pushing SBP <120 often drives DBP dangerously low. Attendings must teach trainees to monitor for signs of over-treatment, such as orthostatic hypotension, rising BUN/Cr, and fatigue, and to champion deprescribing to optimize safety, shifting the focus from 'treating numbers' to holistic patient vitality.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

ACCORD utilized a complex 2x2 factorial design to simultaneously assess intensive glycemic, lipid, and blood pressure control. From a methodological standpoint, how did the unexpected early termination of the glycemic arm impact the interpretability and statistical power of the ongoing blood pressure arm?

Key Response

Factorial designs assume no interaction between the interventions. The early termination of the intensive glycemic arm (due to increased mortality) meant that patients in the BP arm had their glycemic management protocols altered mid-study. This introduces a major time-varying confounder. It also affects the trial's statistical power, as the overall event rate and loss-to-follow-up dynamics shifted, complicating the ability to isolate the true causal effect of the BP intervention on the primary outcome.

Journal Editor
Journal Editor

As a peer reviewer assessing the initial ACCORD-BP manuscript, what concerns would you raise regarding the prominence of the secondary outcome (stroke reduction) in the abstract and discussion, given the negative primary composite outcome?

Key Response

Highlighting a positive secondary outcome when the primary composite outcome is negative is a classic form of 'spin.' A rigorous editor would flag this as a risk for the multiple testing fallacy (inflated Type I error) and demand that the stroke findings be explicitly framed as hypothesis-generating. The editor would ensure the abstract strongly emphasizes the lack of primary benefit and the increase in adverse events to prevent misinterpretation by the broader clinical community.

Guideline Committee
Guideline Committee

The ACCORD-BP trial was a pivotal reason earlier guidelines maintained a <140/90 mm Hg target for diabetics. However, the 2017 ACC/AHA guidelines recommended lowering the target to <130/80 mm Hg for high-risk patients, including those with diabetes. How does the guideline committee justify this more stringent target in light of ACCORD-BP's null primary outcome?

Key Response

The 2017 ACC/AHA guideline shift (Class I, Level of Evidence B-R) was heavily influenced by a meta-analysis approach combining ACCORD-BP, SPRINT, and other trials. The committee noted that while ACCORD-BP was negative for its primary endpoint (target <120), the achieved SBP in the standard group was ~133 mmHg. Consequently, a target of <130 mmHg was deemed the 'sweet spot' that captures the stroke benefit seen in ACCORD-BP and the broader CV benefits seen in similar populations, while minimizing the severe adverse events observed when pushing SBP below 120 mmHg.

Clinical Landscape

Noteworthy Related Trials

1998

UKPDS 38

n = 1,148 · BMJ

Tested

Tight blood pressure control (target <150/85 mm Hg)

Population

Patients with type 2 diabetes and hypertension

Comparator

Less tight BP control (target <180/105 mm Hg)

Endpoint

Any diabetes-related endpoint, diabetes-related death, and all-cause mortality

Key result: Tight BP control significantly reduced the risk of diabetes-related deaths, stroke, and microvascular complications.
2008

ADVANCE Trial

n = 11,140 · Lancet

Tested

Routine administration of perindopril-indapamide

Population

Patients with type 2 diabetes and high CV risk

Comparator

Placebo

Endpoint

Composite of major macrovascular or microvascular events

Key result: Routine BP lowering reduced the relative risk of major vascular events, driven mainly by a reduction in nephropathy.
2015

SPRINT Trial

n = 9,361 · NEJM

Tested

Intensive BP control (target systolic <120 mm Hg)

Population

Adults at high CV risk without diabetes

Comparator

Standard BP control (target systolic <140 mm Hg)

Endpoint

Composite of MI, ACS, stroke, heart failure, or CV death

Key result: Intensive BP control significantly reduced cardiovascular events and all-cause mortality.

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