Treatment to Low Glucose Targets Cuts CV Risk in Less-Sick Diabetics: Cohort Study

December 16, 2009

December 16, 2009 (Irvine, California and Santa Maria Imbaro, Italy) — Addressing issues raised recently, to much controversy and debate, a prospective observational study in type 2 diabetics suggests that aggressive control of glycated hemoglobin (HbA1c) levels--that is, to no higher than 6.5% or 7%--significantly improves cardiovascular risk over five years, but only in patients who aren't too old or sick at the outset [1]. In the analysis of several thousand patients from diabetes clinics and community practices in Italy, those who started out with a lot of comorbidity didn't show a decline in CV risk from attaining such HbA1c levels; nor, on the other hand, was their CV risk any worse.

The study is published in the December 15, 2009 issue of the Annals of Internal Medicine.

Several randomized trials over the past decade, notably ACCORD, ADVANCE, and VADT, have shaken traditional diabetes management, which had been supported by the earlier UKPDS trial, by suggesting no CV benefit or perhaps even heightened CV risk from treating to such low HbA1c targets, as chronicled by heartwire . But post hoc analyses of the those studies and various meta-analyses have pointed to a more complex picture, in which such treatment may cut CV risk if it's started before diabetes is well established and comorbidities are few, but perhaps not in longtime diabetics with manifest heart disease.

These poor people, they're on six or seven medications, many of them taking 25 or 30 pills a day. They're trying to eat right and get some exercise, with their arthritis or other conditions. . . . What we're saying with this study is, Ease up a little bit, they'll get just as much out of having an HbA1c between 7 and 8.

"The trouble with post hoc analyses is that they're considered exploratory," observed Dr Sheldon Greenfield (University of California, Irvine), lead author of the current cohort study, in which patients were prospectively stratified by comorbidity scores that accounted for any coexisting vascular, lung, genitourinary, and gastrointestinal diseases; and arthritis, vision loss, and other conditions; each weighted according to their clinical and functional impact. "Our study is much stronger and more quantitative, but it's perfectly convergent," he told heartwire .

"If you add UKPDS and the other studies together," Greenfield said, "the message seems to be pretty darned clear: get [HbA1c] as low as you can for people who are relatively early in their diabetes, with respect to both age and the disease, and maybe lighten up on the people with a lot of comorbidity. [In that group,] go after the other stuff--cholesterol, blood pressure--more aggressively."

Greenfield and his colleagues followed 2613 randomly selected community-based diabetics for a median of five years; all had been classified at baseline as carrying either a low-to-moderate or high comorbidity load according to what they call the Total Illness Burden Index (TIBI), which they have tested in various clinical settings.

Overall, a TIBI risk score of >12 pointed to a multivariate-adjusted hazard ratio (HR) of 1.52 (95% CI 1.21–1.89) for total cardiovascular events (p<0.001) and a trend toward increased all-cause mortality (p=0.074), compared with lower scores.

The 1115 patients with TIBI scores at that threshold or above, the high-comorbidity group, were older (mean 64.3 years vs 61.7 years, p<0.001) and less likely to be female (50.2% vs 58.3%, p<0.001) than the 1498 with lower TIBI scores, the low-comorbidity group. They were also significantly less likely to show a reduction in cardiovascular events (angina, myocardial infarction, transient ischemic attack, stroke, percutaneous coronary intervention or coronary artery bypass graft surgery, lower-limb vascular complications, or cardiovascular death) by attaining HbA1c levels of 6.5% or 7.0% with medical therapy.

Adjusted Hazard Ratio (HR) for Cardiovascular Events Over Median Five Years in 2613 Diabetic Outpatients, by Achieved HbA1c and TIBI Comorbidity

Achieved HbA1c Low-comorbidity group (TIBI<12), HR (95% CI) High-comorbidity group (TIBI>12), HR (95% CI)
<6.5% 0.60 (0.42–0.85)* 0.92 (0.68–1.25)
<7.0% 0.61 (0.44–0.83) * 0.86 (0.64–1.14)
TIBI=Total Illness Burden Index


"If we take Greenfield and colleagues' conclusions at face value, future studies should include patients with few comorbid conditions and, consequently, lower CVD risk than in previous studies," Dr David M Nathan (Massachusetts General Hospital, Boston) writes in an accompanying editorial [2]. Based on current data, such patients "might have baseline CV disease rates that approach 1% per year"; therefore, a trial testing the current study's findings could potentially require "30 000 volunteers followed for six years," according to Nathan.

"We must carefully consider whether the putative and modest benefit of achieving an HbA1c level of 6.0% or 6.5%, compared with that of a level 1% higher, is worth exploring in a study that would consume far more resources than previous efforts." Such a trial, he said, is unlikely to be cost-effective. "Meanwhile, clinicians can be confident that aiming for an HbA1c level less than 7% is justified by the strong evidence that it prevents or delays progression of microvascular and neurologic complications and seems to reduce CV disease."

Although the current study suggests that aggressive HbA1c management is more likely to prevent cardiovascular benefits if it's carried out in younger patients with few or no comorbidities, Greenfield said, "actually, our study is stronger in considering the high-comorbidity patients. There the message is cleaner. These poor people, they're on six or seven medications, many of them taking 25 or 30 pills a day. They're trying to eat right and get some exercise, with their arthritis or other conditions. Their lives are already complicated enough. What we're saying with this study is, Ease up a little bit, they'll get just as much out of having an HbA1c between 7 and 8."

The study was funded by Pfizer of Italy. The authors and Nathan disclosed no potential conflicts of interest.