Deaths linked to low HbA1c levels spark call for revised limits

EXPERTS are issuing new warnings about the dangers of excessive glucose lowering in patients with type 2 diabetes, after further evidence linking low HbA1c levels to increased mortality has emerged.

An analysis of almost 50,000 patients from a UK general practice database found a 52% increased risk of death in those with the lowest HbA1c levels, compared to median levels of 7.5%.

The relationship was U-shaped, with the greatest mortality risk at HbA1c levels above 10.5% and below 6.4 per cent. The lowest mortality appeared to be at an HbA1c of 7.5 per cent.

Following on from the findings of the high-profile ACCORD trial, the authors said the new findings could see guidelines changed to include lower, as well as upper, limits for HbA1c.

The ACCORD trial, which targeted HbA1c below 6%, was stopped early in 2008 after intensive glucose control was found to increase mortality by 22 per cent.

A new post-hoc analysis of the study reported that severe hypoglycaemia was three times as common in the intensive treatment group.

Australian experts recommended GPs continue to aim for an HbA1c level of 7% in most patients, but should be more mindful than ever about avoiding hypoglycaemia.

“We now know that going down to 6% is not necessarily a safe thing to do in everybody,” said Associate Professor Shane Hamblin, director of diabetes at Western Health. “This paper adds to the degree of caution about very tight control.”

“It’s fascinating that it didn’t seem to make much difference whether insulin or tablets were being used – the U-shaped curve seemed to be present in both types of treatment,” he added.

The authors’ recommendations to set minimum target limits were premature, said Associate Professor Stephen Twigg, president of the Australian Diabetes Society.

He said cause and effect remained unproven, since not all clinical trials had shown increased mortality with intensive control.

Associate Professor Jona-than Shaw, deputy director of the Baker IDI Heart and Diabetes Institute, agreed that the mechanism of the relationship was unclear.

He noted that patients who achieved the lowest HbA1c levels might have already faced a higher mortality risk.

“I think we should continue to aim for less than 7%, but we should be very cautious about hypoglycaemia, particularly in older people and particularly in people with prior cardiovascular disease,” he said.

“In such people, it’s usually better to have a target somewhat higher than 7% – maybe 7.5% or sometimes even 8 per cent.”

Lancet, online 27 January; BMJ 2010;340:b5444