ADVANCE: Severe Hypoglycemia a Marker, Not Cause, of Poor Diabetes Outcomes?

July 08, 2010

July 8, 2010 (Orlando, Florida) — Development of severe hypoglycemia in patients from the ADVANCE trial correlated significantly with later cardiovascular and microvascular events as well as development of nonvascular disorders such as cancer and digestive and respiratory diseases, according to researchers presenting last week at the American Diabetes Association 2010 Scientific Sessions.

Moreover, reported Dr Sophia Zoungas (George Institute for International Health, Sydney, Australia) at the meeting, severe hypoglycemic events in the controversial trial were a significant predictor of such outcomes regardless of randomization group--that is, whether the patients with type 2 diabetes had been assigned to its "intensive" or standard glucose-lowering strategy.

In addition, she said, adverse events didn't appear to closely follow serious hypoglycemic episodes or to be related to the number of such episodes patients experienced.

We're looking at an alternative explanation, that severe hypoglycemia identifies sick patients who have other pathologies.

"There are two possible explanations," Zoungas told heartwire . One is that severe hypoglycemia directly causes the poor outcomes, and the other, "which we propose is more likely, is that hypoglycemia is a marker of increased vulnerability." Such vulnerability apparently raised the risk of both vascular and nonvascular events along with the risk of severe hypoglycemia, which in this analysis required symptoms and a degree of disability, she said.

Zoungas and her colleagues aren't necessarily rejecting the idea that the severe hypoglycemia actually causes poor outcomes, "it's just that we're looking at an alternative explanation, that severe hypoglycemia identifies sick patients who have other pathologies."

The ADVANCE analysis, she said, was somewhat in response to concerns that intensive glucose-lowering regimens in recent trials may have induced hypoglycemia severe enough to keep the aggressive therapy from showing a cardiovascular benefit, as occurred in ADVANCE and the VADT trial, or even to increase mortality, as suggested in ACCORD.

As previously reported by heartwire , ADVANCE enrolled 11 140 patients with a history of CV disease or microvascular disease with at least one other CV risk factor at 215 centers in 20 countries. There was no A1c or fasting-glucose requirement for study entry. They were randomized to receive the modified-release sulfonylurea gliclazide at 30 mg to 120 mg daily plus other glucose-lowering drugs as needed to achieve an A1c <6.5% or to standard guidelines-based management using glucose-lowering drugs at the physician's discretion.

If there were a true causal relationship, the more hypoglycemic episodes that occur, the greater the risk should be. And we didn't demonstrate that.

Severe hypoglycemia was strictly defined as "a blood glucose level <50 mg/dL (<2.8 mmol/L) and the presence of typical signs and symptoms of hypoglycemia with no other cause, associated with transient central-nervous-system dysfunction during which the participant was unable to treat himself/herself and required assistance from another person."

It occurred in 2.7% of the 5571 intensively treated patients and 1.5% of the 5569 patients on standard management over a median follow-up of five years, Zoungas reported, for rates of 0.7 and 0.4 events per person-years, respectively. The hazard ratio for severe hypoglycemia for intensive vs standard therapy was 1.86 (95% CI 1.40–2.40).

In multivariate analysis, significant predictors of severe hypoglycemia were advanced age, longer diabetes duration, higher creatinine levels, lower body-mass index (BMI), poorer cognitive function, use of >2 oral glucose-lowering drugs, history of smoking, history of microvascular disease (regardless of randomization group), and assignment to intensive glucose control (p<0.05 for all).

Severe hypoglycemia was an independent predictor of both of the trial's primary end points and its three clinical secondary end points, regardless of randomization group.

Adjusted Hazard Ratio for Primary and Secondary Clinical Outcomes, ADVANCE Patients Who Developed Severe Hypoglycemia vs Those Who Didn't

End point Severe hypoglycemia, n=231 (%) No severe hypoglycemia, n=10 909 (%) HR (95% CI)
Major macrovascular event* 15.9 10.2 3.53 (2.41–5.17)
Major microvascular event* 11.5 10.1 2.19 (1.40–3.45)
All-cause mortality 19.5 9.0 3.27 (2.29–4.65)
Cardiovascular mortality 9.5 4.8 3.79 (2.36–6.08)
Noncardiovascular mortality 10 4.3 2.80 (1.64–4.79)

*Primary end points. Major macrovascular event=CV death, nonfatal MI, or nonfatal stroke; major microvascular event=new or worsening nephropathy or retinopathy

Patients who experienced >2 episodes of severe hypoglycemia showed an independently increased risk of macrovascular events (CV death, nonfatal MI, nonfatal stroke), with an HR of 2.65 (95% CI 1.30–5.40; p=0.0072). However, there was otherwise no consistent significant relationship between the number of such episodes and development of microvascular events or other individual clinical end points.

Nor was a close temporal relationship observed between such episodes and outcomes. Major macrovascular events occurred a median of 1.56 years after initial severe hypoglycemia episodes; major microvascular events, a median of 0.99 years; death from any cause, 1.05 years; CV death, 1.31 years; and non-CV death, 0.74 years.

"If there were a true causal relationship, the more hypoglycemic episodes [that occurred], the greater the risk should be," Zoungas said. "And we didn't demonstrate that. And if it were directly causal, we should have been able to see that these outcomes occurred a short period after the actual experience of the hypoglycemia, and again we did not demonstrate that."

Also supporting her group's view that severe hypoglycemia is only a marker of increased risk, not a cause, she said, was that it correlated not only with vascular outcomes, it also predicted "a range of nonvascular outcomes where there is absolutely no plausible reason for there being a direct causal link."

Adjusted Hazard Ratio for Nonvascular Clinical Events and Complications, ADVANCE Patients Who Developed Severe Hypoglycemia vs Those Who Didn't

End point Severe hypoglycemia, n=231 (%) No severe hypoglycemia, n=10 909 (%) HR (95% CI)
Noncardiovascular mortality 10 4.3 2.80 (1.64-4.79)
Respiratory-system events 8.5 6.0 2.46 (1.43-4.23)
Digestive-system events 9.6 7.9 2.20 (1.31-3.72)
Skin diseases 2.7 1.3 4.73 (1.96-11.40)
Cancer events 2.2 1.4 2.11 (0.65-6.82)

As she noted in her presentation, if severe hypoglycemia were to be a direct cause of cardiovascular morbidity and mortality, "plausible mechanisms" might include resulting sympathoadrenal activation, abnormal cardiac repolarization, or increased thrombogenesis, inflammation, or vasoconstriction.

"Now, our data are limited in that severe hypoglycemia may also have occurred at the time of the [clinical] events but went unreported," Zoungas said to heartwire . That might have obscured any temporal link between them. "However, we believe that it's highly unlikely that for such a large number of events, there would have been that many unreported episodes of severe hypoglycemia."

ADVANCE was funded by the National Health & Medical Research Council of Australia and Servier International. Zoungas discloses support for serving on the speaker's bureau and travel from Servier and receiving fellowship support from the National Health and Medical Research Council of Australia and Diabetes Australia.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: