Dementia Increases Death Risk From Hypoglycaemia in Diabetes

Liam Davenport

October 04, 2018

BERLIN — Older diabetes patients with dementia have a markedly increased mortality risk following an episode of hypoglycaemia than those without dementia, reveals a large UK data analysis.

Dr Katharina Mattishent, Alzheimer's Society clinical research fellow, at Norwich Medical School, University of East Anglia, examined the records of almost 20,000 diabetes patients aged at least 65 years.

Presenting her findings here as a poster at the EASD 2018 annual meeting, she showed that those who had dementia were 67% more likely to die following a hypoglycaemic event than those without dementia.

Minimising Risk

Mattishent commented in a release that hypoglycaemia is "under-recognised" as a risk factor for death in older diabetes patients with dementia, and emphasised that strategies to minimise hypoglycaemia should be prioritised.

James Pickett, Head of Research at the Alzheimer's Society, which funded the study, added: "With no new dementia drugs in 15 years, minimising risk and improving care is key.

"We know that diabetes can raise the risk of developing dementia, and with both of these illnesses on the rise we urgently need to understand this relationship better."

He added: "Very low blood sugar levels are clearly dangerous to anyone with diabetes, and this suggests the effects might be even more extreme in people with dementia.

"The study didn't show cause and effect but, given the dangers of low blood sugar levels, clearly it should be managed carefully."

Mattishent nevertheless told Medscape News UK that it is "biologically plausible" that hypoglycaemia would increase the mortality risk in older diabetes patients with dementia.

This could be due to such patients having difficulties "with monitoring and maintaining their blood sugars steady in the recommended range".

She said: "Some researchers have found that hypoglycaemia can disrupt the heart rhythm, whilst others have suggested that the brain may experience harm when starved of glucose," adding: "Dementia itself can be seen as a marker of increased frailty."

A Fresh Look at Targets

On the basis of the current findings, Mattishent urged clinicians and patients to "move away from relentless pursuit of strict glucose-lowering targets", saying in the news release that the "focus must be directed at rigorous detection of hypoglycaemia using continuous glucose monitoring devices [CGM].

"This close monitoring will guide treatment choices with regards to drugs that have a low risk of hypoglycaemia in this patient group."

In the UK, a number of CGM devices are available, but Mattishent told Medscape News UK that "these are not typically funded on the NHS for older people, and self-funding of these devices may prove to be a barrier.

"We believe that wider use of continuous glucose monitoring in older people with diabetes and dementia will require a significant shift in the current, narrow-minded approach to CGM," she said.

"So far, the benefits of CGM have been mainly targeted at the younger insulin-users, with the benefits of improved lifestyle and HbA1C outcomes."

Mattishent added: "Whilst the NHS is prepared to invest hundreds of pounds per year for expensive insulins that have no major improvement in clinical benefit or safety in randomised controlled trials, there seems to be less of an appreciation of the seriousness of undetected hypoglycaemia and its major consequences in frail older patients."

Hypoglycaemia Risks

Recent studies have suggested that hypoglycaemia is associated with an increased risk of death, cardiovascular events, falls, and fractures.

This may be a particular issue in older patients, as self-management is challenging and they have higher rates of drug-induced hypoglycaemia.

The researchers therefore set out to examine the association between hypoglycaemia and all-cause mortality in older diabetes patients with dementia.

They looked at data from the UK Clinical Practice Research Datalink database on individuals aged 65 years and over with a new or existing diabetes diagnosis between 1997 and 2016.

Linking to the Hospital Episode Statistics database the team divided the participants into those with dementia and diabetes but no hypoglycaemia (group 1); dementia, diabetes and hypoglycaemia (group 2); and diabetes and hypoglycaemia but no dementia (group 3).

In all, the team gathered information on 19,995 patients with a first prescription of glucose-lowering drug therapy between 1997 and 2016, of whom 6134 were in group 1, 1679 were in group 2, and 12,182 were in group 3.

While the baseline characteristics were comparable across the three groups, patients with dementia were older than those without, at an average age of 82 years and 83 years, respectively, in groups 1 and 2 versus 77 years in group 3.

Insulin use was higher in group 2 than in group 1.

Using mortality data from the Office of National Statistics, the team found that, over a follow-up period of up to 5 years, there were 3853 (63%) deaths in group 1, 1369 (82%) in group 2, and 6494 (53%) in group 3.

Comparing groups 2 and 3, they calculated that dementia had a substantial effect on the mortality risk in patients with hypoglycaemia, at a hazard ratio adjusted for multiple patient and clinical characteristics of 1.67.

They also determined, by comparing groups 2 and 1, that hypoglycaemia had a marked effect on mortality in patients with dementia, at an adjusted hazard ratio of 1.66.

Next Steps

The researchers acknowledge that their study may be limited by under-reporting of hypoglycaemic episodes or missing data on factors such as body mass index (BMI), HbA1c levels, and smoking and alcohol status.

They nevertheless conclude: "The results of this retrospective cohort study underscore the serious consequences of hypoglycaemia in patients with comorbid diabetes and dementia, and the need for a personalised management strategy tailored towards avoidance of hypoglycaemia."

To further understand the association between hypoglycaemia and serious complications, Mattishent and colleagues are looking at the associations with myocardial infarction and ischaemic stroke, and with falls and fractures.

"In addition, we are also conducting a feasibility study on the acceptability of flash glucose monitoring in older people with diabetes and memory problems," she explained.

The study was supported by the Alzheimer's Society.

Mattishent is funded through a clinical training fellowship from Alzheimer's Society with support from Celesio UK. 

EASD 2018: Poster 913. Presented October 2nd.


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: