Blood pressure reduced to new guideline goals in patients with high‐normal glucose further reduces cardiovascular events

The SPRINT (Systolic Blood Pressure Intervention) trial heralded a new era of lower blood pressure (BP) for those with high cardiovascular risk1 It also shaped recent guidelines mandating lower levels of BP as defining hypertension.2 SPRINT evaluated 9361 patients without diabetes mellitus but with elevated cardiovascular risk. They were randomized to intensive (target systolic BP < 120 mm Hg) vs standard (target systolic BP < 140 mm Hg) therapy. Intensive BP lowering was associated with a 25% lower relative risk of major cardiovascular events compared with a standard BP‐lowering strategy. Intensive therapy also reduced the relative risk of all‐cause mortality by 27%1.

Given that the SPRINT cohort did not have diabetes mellitus, the question remains whether the SPRINT results apply to patients with diabetes mellitus. Some meta‐analyses support the concept that lowering BP to levels <130/80 mm Hg does reduce all types of cardiovascular outcomes in high‐risk patients with diabetes3 Interestingly, the benefits derived from more intensive BP control were more prominently seen among those without diabetes mellitus when systolic BPs were <130 mm Hg.4 Moreover, the SPRINT data did not support an increase in cardiovascular events when systolic BP levels were <125 mm Hg, events presumably caused by small‐vessel ischemia and underperfusion.

The ACCORD (Action to Control Cardiovascular Risk in Diabetes) trial randomized patients to different levels of BP control similar to SPRINT but in patients with type 2 diabetes mellitus. However, in the main trial results of ACCORD, the intensive BP‐lowering group did not have fewer nonfatal myocardial infarctions and cardiovascular deaths compared with the group with standard BP lowering (hazard ratio, 0.88; 95% confidence interval, 0.73–1.06)5 Moreover, the results were confounded by the fact that more people died in the intensive glucose‐lowering group, hence affecting the power of the BP outcome in that trial. Secondary analyses of ACCORD and the recent ACCORDION trial, an 8‐year follow‐up of ACCORD participants, demonstrate a clear reduction in cardiovascular events in the intensive BP‐lowering group.6, 7

There have been several studies that evaluated the effect of impaired fasting glucose on the development of hypertension and cardiovascular outcomes. Almost all such studies defined impaired fasting glucose as a morning glucose level between 110 and 125 mg/dL. Henry and colleagues8 showed that impaired fasting glucose was associated with a substantial increase in 8‐year overall and cardiovascular mortality in patients with baseline systolic BP ≥ 140 mm Hg in a study that included 63 443 men with relatively low cardiovascular risk. Impaired fasting glucose was the prime determinant of mortality associated with stage 2 systolic hypertension (systolic BP: 140–159 mm Hg) based on the new nomenclature.

The study by Gong and colleagues9 in the current issue of the Journal of Clinical Hypertension performed a similar analysis to Henry and colleagues but in SPRINT participants with a glucose level in the range generally described as normal, ie <100 mg/dL. Gong and associates dichotomized 5425 SPRINT participants into two groups: those with fasting blood glucose <85 mg/dL (low‐normal group) and those with fasting blood glucose 85 to 99 mg/dL (high‐normal group). Their analysis found that participants randomized to intensive BP control with glucose in the range of 85 to 99 mg/dL had a 27% risk reduction in major cardiovascular events compared with patients randomized to standard BP control. The authors concluded that the benefits of intensive BP lowering were more likely to be seen in patients who were closer to having impaired fasting glucose levels.

This is a novel, post hoc analysis that is based on very little a priori data upon which to base a hypothesis. Its validity, therefore, may be affected by a several factors. The authors used automated procedures for variable selection in this subset analysis. Automated procedures are useful but must be used cautiously. They are best used for exploratory purposes, rather than for final modeling. While data extracted in this way can be used for prediction or explanation of outcome, the confounding with nonsignificant variables should be included in final models to obtain the most accurate inferences in observational studies. This aspect was missing in this analysis.

In addition, the authors based their analysis on baseline serum glucose levels and dichotomized patients into two groups depending based on a single measurement (low‐normal, <85 mg/dL, vs high‐normal, 85–99 mg/dL). This by itself may be a source of bias, as fasting plasma glucose levels are not static and are influenced by a host of factors. Intraindividual variability as well as interindividual variabilities in fasting blood glucose are reported10 Acute stress can also affect β‐cell function and alter fasting blood sugar in patients without diabetes mellitus. Poor sleep quality and short sleep duration are well‐known contributors to elevated fasting plasma glucose levels11

Among the participants examined, the mean fasting glucose was 80 mg/dL in the low‐normal group. A 15% change would yield a 12‐mg/dL difference, thus shifting a person into the high‐normal group. Glycated hemoglobin is a more reliable predictor for such an analysis and would be less likely affected by this variability but was unavailable.

The results of this post hoc SPRINT analysis presents a hypothesis that those with “high‐normal” glucose (85–99 mg/dL) treated to BP levels <130 mm Hg garner greater cardiovascular risk reduction. The study extends our fund of knowledge by emphasizing the potential benefits of aggressively lowering BP among patients who are in the normal glucose range of 85 to 99 mg/dL. While the new American College of Cardiology/American Heart Association BP guidelines do not specifically address this subgroup, they do address a benefit of lower cardiovascular risk by achieving lower BP among patients with impaired fasting glucose2 These data extend this range but need prospective validation.

CONFLICT OF INTEREST

Neither Drs Al Dhabyi or Bakris have any relevant conflicts of interest regarding this topic.

Al Dhaybi O, Bakris GL. Blood pressure reduced to new guideline goals in patients with high‐normal glucose further reduces cardiovascular events. J Clin Hypertens. 2018;20:625–626. 10.1111/jch.13242