Hidden figures: Unveiling differences in diabetes care in people living with mild and moderate-to-severe dementia

In this CRIS blog, Dr Usha Gungabissoon investigates what happens when people with diabetes develop dementia in terms of care received and adverse outcomes.  In her PhD at King's College London Dr Gungabissoon used CRIS, the unique linkage between mental healthcare, and local primary care, Lambeth DataNet. These NIHR-supported findings were published this week in Journal of Post Acute and Long Term Care Medicine (JAMDA).

Up to one fifth of people with dementia also have type 2 diabetes, and diabetes care can be especially challenging when dementia occurs. Diabetes is a mostly self-managed condition; patients are asked to take regular medications, monitor their blood sugar and follow dietary guidelines. All of these tasks become harder for patients who are also experiencing memory difficulties and other symptoms of dementia. 

Surprisingly, little is known about diabetes control or outcomes in people with dementia. This is important because poor diabetes control results in a higher risk of diabetes-related complications such as stroke, heart failure and kidney disease, and a higher risk of emergency hospitalisations and death. As well as the impact on patients, their families and carers, dementia in diabetes also places demands on an already stretched NHS. Diabetes costs the NHS around £15.1bn per year and a substantial proportion arises from diabetes-related complications.

What did we do?

To understand if there were differences in healthcare and diabetes-related measures, we looked at general practice information in Lambeth, South London. We investigated patients with type 2 diabetes with newly diagnosed dementia and compared them to patients of the same age and sex without a dementia diagnosis.

We found important differences. Although patients with dementia had more healthcare use in general around the dementia diagnosis, they were less likely to have had routine diabetes-specific management. Compared to those without dementia, these patients were more likely to experience cardiovascular complications and foot problems that occur in diabetes. They also had a higher mortality risk. 

We investigated further to see if there were any differences based on the severity of cognitive impairment in dementia.  We found that that people with more severe dementia received less routine diabetes management and had more diabetes-related complications than those with mild dementia.  We also found differences in glucose control over time (as measured by HbA1c levels which is used to diagnose and monitor diabetes) according to both the presence of and severity of dementia. For example, our study found that those with mild dementia had a steeper increase in HbA1c levels over time than those without dementia. This could be due to challenges in self-management of diabetes experienced by people with dementia. 

Meeting unmet need in older people with diabetes and dementia

Our findings highlight possible unmet needs in older people with diabetes who develop dementia and suggest there could be ways in which the NHS could improve clinical outcomes, and quality of care in this vulnerable population.

For example, dementia assessment and care should be much broader than the current focus on memory difficulties and symptoms around behaviour since many people with dementia already have other longstanding health conditions which are going to be harder to self-manage over time. We need primary care and general medical services to recognise the importance of dementia as a challenge for other health conditions and more proactive support for those affected.

In the case of diabetes this could be done by tailoring management and adapting diabetes care related indicators in the primary care setting, to account for the effects of frailty and the extent of memory problems  amongst those with dementia. This could help improve clinical outcomes, quality, access, and efficiency of care in this vulnerable population.

Continuing research

Alongside this there is also an ongoing need for more research to improve our understanding of the physical health impacts of dementia. Approaches such as the one we used in this study that extract information from clinical e-records and link to other datasets could be instrumental in this.

If the extent of memory problems in dementia had not been accounted for in our analyses through the power of data, then important differences in HbA1c patterns and other measures of diabetes-related care would have remained hidden. This was only possible because of the unique linkage between Lambeth DataNet which includes detailed information on diabetes care in general practice and CRIS, which includes detailed information on dementia.