Blood Pressure Control And Diabetes: What Should Our Goal Be ACCORDing To The Evidence?

Are you treating to an evidence-based blood pressure goal in patients with diabetes? Consensus guidelines suggest that we should aim for a blood pressure of < 130/80 mmHg.^1,2 However, the support for this recommendation is derived from epidemiological data and subgroup analyses rather than well-designed, randomized controlled trials specifically designed to determine the optimal BP targets.^3-5  The recommended goal for diastolic blood pressure (DBP) is primarily based the results of the Hypertension Optimal Treatment (HOT) trial.  A planned sub-group analysis found that people with diabetes who were randomized to a target DBP of < 80mm Hg had significantly fewer major adverse cardiovascular events (MACE) compared to those randomized to a DBP of < 90mm Hg.⁴  However, subjects with diabetes were a mere 8% of the total study population in this trial, and the absolute number of events was small.  High-quality, grade A evidence for the goal systolic blood pressure (SBP) of less than 130 mmHg is similarly lacking.  To address this vacuum, the investigators of the Action to Control Cardiovascular Risk in Diabetes (ACCORD) blood pressure trial set out to test the hypothesis that intensive treatment to a SBP < 120 mmHg (compared to a SBP goal < 140 mmHg) would reduce the risk of MACE in patients with diabetes.⁶

The ACCORD study was a prospective, randomized, multi-center clinical trial that was designed to assess the benefits of intensively treating blood glucose, blood pressure, and blood lipids on cardiovascular outcomes in high-risk patients with diabetes mellitus.⁷  (SeeFigure 1 – ACCORD BP Study Design)  The primary outcome was the first occurrence of MACE, which was defined as the composite of nonfatal myocardial infarction, nonfatal stroke, or cardiovascular death.  With a planned enrollment of 4200 patients, the blood pressure arm of the trial was designed with 94% power to detect a 20% reduction in the primary outcome for the intensively treated group, assuming an event rate of 4% in the standard treatment group and an average follow-up of 5.6 years.

Baseline characteristics of the intensive and standard treatment groups were similar, with an average age of 62.2 years and a median duration of diabetes of 10 years.  Approximately 1 in 3 patients had a previous cardiovascular event.  The average blood pressure was 139.2/72.0 mmHg.  After randomization, patients were treated in a non-blinded fashion using currently approved antihypertensive medications.  Treatment selection was at the discretion of the individual investigator, but the use of medications that had previously been shown to lower cardiovascular event rates in patients with diabetes were recommended (e.g. thiazide diuretics, ACEIs, ARBs).  After the first year of therapy, patients in the intensive treatment group required 3.4 medications on average to achieve a mean SBP of 119.3 mm Hg, compared to an average of 2.1 medications to achieve a mean SBP of 133.5 mm Hg for the standard treatment group.  The mean duration of follow-up was 4.7 years.

Results:  Primary and secondary outcomes from the ACCORD blood pressure study

†Primary outcome

Overall, there was no significant benefit from intensive treatment.  There was a statistically significant reduction in the rates of total and nonfatal stroke, although the total number of events was small.  The NNT over 5 years to prevent one stroke was 89.  Patients in the intensive treatment group had more serious adverse events attributed to the use of antihypertensive medications compared to those in the standard treatment group (3.3% per year vs. 1.27% per year, p<0.001).  Small but significant increases in the number of patients with hypotension, bradycardia, and hyperkalemia were observed (all p<0.05).  There were also significant increases in the rates of hypokalemia (p=0.01) and elevated serum creatinine concentrations (p<0.001) in the intensive treatment group. 

Based on these results, it appears that there is little to gain from treating patients with diabetes and hypertension to a SBP goal of 120 mmHg or less.  Some have suggested that the standard treatment goal of 140/90 should be used instead.  However, there are some limitations to the ACCORD trial that must be considered.  First, the event rate for the primary outcome in the standard treatment group was approximately half of what was expected, which reduced the power of the study to detect a significant difference between the two groups.  Indeed, the confidence interval for the primary outcome does not rule out the possibility of a 27% reduction in the risk of MACE with intensive treatment – a difference, if true, that would be clinically important.  Secondly, there were trends towards a significant interaction between glycemic control and cardiovascular outcomes in the study.  Patients assigned to the standard glycemic control arm and those with HbA₁C < 8.0% at the time of study entry appear to have derived greater benefit from intensive blood pressure treatment (p=0.08 and p=0.11, respectively).  This suggests that certain patients may receive a significant benefit from intensive BP control based on their overall CV risk.  Third, the ACCORD study focused solely on SBP goals, and ignored the effect of diastolic blood pressure control (which averaged 64.4 mm Hg in the intensive treat group).  Very low diastolic pressures may increase the risk of MACE (the so-called J-hypothesis⁸), particularly in patients with pre-existing cardiovascular disease, and this may have blunted any potential benefit from intensive BP management.  Fourth, since the trial was designed to test a specific goal SBP rather than a particular drug, confounding effects from the use of various medication regimens is a possibility.  Some antihypertensive drug combinations are likely better than others (see the ACCOMPLISH study⁹).  Finally, ACCORD was primarily designed to examine the effects on macrovascular events, despite the fact that optimal blood pressure control may provide benefits on microvascular outcomes as well.¹⁰ Admittedly, the results of the ACCORD Eye Study¹¹ (a subset of the ACCORD trial) did not demonstrate any benefit from intensive blood pressure control on the progression of diabetic retinopathy, making this possibility less likely.

So, how should we interpret and apply the results of the ACCORD BP study to clinical practice?  Despite its limitations, the ACCORD blood pressure trial did provide some valuable information.  In this group of patients with long-standing type 2 diabetes, intensive treatment required more medications (with presumably higher costs) and increased the risk of adverse effects, while providing no significant reduction in cardiovascular events.  Therefore, we must conclude that there is no compelling reason to treat to a goal SBP < 120 mmHg.  Unfortunately, this still doesn’t answer our original question – what should the goal blood pressure be for patients with diabetes?  The fact that the ACCORD investigators chose a blood pressure of 120 mmHg instead of 130 mmHg for the intensive treatment arm makes this difficult to answer.  Moreover, BP control actually achieved in the standard treatment group (mean BP = 133.5/ 70.2 mmHg at the end of the first year) was better than what is often achieved in practice today and close to the currently recommend < 130/80 goal.

Taken together with the results of previous trials, I think that we should make several changes to our current practice.  First, we should abandon the use of open-ended goals that suggest that any blood pressure less than a particular number (i.e. <130/80) is appropriate.  I believe that this encourages the use of more medications without acknowledging the potential increased risk of adverse events.  Instead, we should aim for a specific target range.  Based on the results of ACCORD and previous trials, a target range of 130-139/70-79 mmHg would be appropriate for most patients with diabetes.  Second, we need to acknowledge that diabetes is not a disease but rather a heterogeneous group of diseases (with hyperglycemia as its common manifestation) that requires individualized goals and treatments.  A one-size-fits-all approach isn’t appropriate.  While intensive management of SBP may be inappropriate for those with long-standing diabetes or pre-existing cardiovascular disease, it may be an appropriate goal for younger patients or those newly-diagnosed, particularly if it can be accomplished safely with few medications.  Specific patient populations, such as those at higher risk of stroke or renal disease (e.g. African American), may benefit from more intensive blood pressure control.  Unfortunately, ACCORD BP didn’t give us the answers we were seeking … and may erroneously lead us to believe that achieving good BP control in people with diabetes is no longer important.  What do you think?