SP - Diabetes

INTRODUCTION:

We now face an increasing challenge of managing type 1 diabetes (T1DM) in older people. 3% of patients with T1DM are over 80 years old - a number that is set to rise over the coming decades. Diabetes is a risk factor for frailty, and advanced age comes with a higher incidence and severity of comorbidities affecting patients’ ability to manage their treatments, such as arthritis or dementia. A recent systematic search in England highlighted the scarcity of data surrounding this, finding no articles specifically researching frailty in older adults with T1DM.

CASE STUDY:

●90-year-old lady with T1DM

●Living alone with mild frailty

●Mild cognitive impairment

Despite living alone and being independent with many of her activities of daily living, her mild cognitive impairment made her unable to manage her insulin safely. Consequently, she had recurrent hospital admissions due to labile blood sugars. Local services (even in central Manchester) offer a maximum of two visits each day, which is not a safe insulin regime for most patients living with T1DM. As family were not available to support, she had prolonged inpatient hospital stays and ultimately had to explore 24-hour nursing care, despite being otherwise independent.

The case provides a poignant illustration of an issue which is only going to become more frequent as the early users of insulin now reach older age.

Introduction

One in four Care Home (CH) residents have diabetes, making diet crucial for controlling glucose levels (GLs). Continuous blood glucose monitoring (CGM) now offers deeper insights into GL fluctuations. Diabetes in severe frailty is often overtreated, particularly with insulin, posing risks such as hypoglycemia, avoidable hospital admissions, and labour-intensive clinical supervision. While protein and vegetables can slow glucose absorption, dietary advice for CH residents typically emphasizes carbohydrates and may not be tailored to their frailty. This study investigates the impact of modifying protein intake in insulin-using diabetics to improve glycaemic control.

Method

A small pilot study assessed if protein-rich foods (e.g. eggs, peanut butter) given for breakfast stabilise GLs throughout the day. Eight diabetic CH residents using insulin were randomly selected over four months. A diabetic frailty pharmacist monitored GLs with the CGM device (Freestyle Libre) and analysed GLs after a protein-rich breakfast. Descriptive analysis and t-tests were conducted using R before and after the food intervention, and ANOVA was used to analyse significant differences in GLs.

Results

Six out of eight patients showed statistically significant reductions in GL spikes, sustained throughout the day. For the remaining two patients, the food intervention helped maintain target GLs. This led to the discontinuation of insulin in one patient, and in the second, problematic frequent hypoglycemia was mitigated by the food intervention. Clinical decisions on patient safety influenced outcomes for these two patients but were not excluded from analysis.

Conclusion

Six of the eight residents given additional protein at breakfast showed significant GL reductions, leading to decreased insulin dosing and simpler regimes. Carers reported improvements in mood, sleep, and energy levels anecdotally. A holistic dietary approach in managing diabetes in CH residents, emphasizing increased morning protein intake, should be considered to enhance GL control and allow deprescribing. A larger study is planned.

Background:

Type 2 Diabetes mellitus (T2DM) is the most common long-term metabolic condition in older people. In the UK, half of all diabetic patients are over 65 and prevalence reaches 10% in over 75s. Lifestyle interventions reduce diabetic complications and can achieve remission, however, there are concerns over the generalisability of these findings to the diabetic population, particularly elderly, complex patients, and those from ethnic minorities. This systematic review quantifies the disparity between diabetes clinical trial cohorts and the UK diabetic population.

Method:

This is a systematic review of UK-based randomised control trials (RCTs) of non-pharmaceutical interventions in adults with T2DM. Data was collected on characteristics of participants included in these studies, including age, sex, ethnicity, socioeconomic status and education of participants.

Results:

Our search strategy identified 5437 results, of which 161 met the criteria for full-text screening. After full-text screening and de-duplication, 80 RCTs were included in our analysis. Of 80 studies, 60% (48/80) reported a mean participant age under 60. Only 40 (50%) reported participant age range; of these the maximum participant age was under 65 in 20% and under 75 in 60%. Where the mean age of participants was over 60, 56% (18/32) restricted participation by comorbidities. Almost all of these precluded anyone with pre-existing CVD (17/18), one third precluded any comorbidities, and 5/18 precluded hypertensive patients. Only 26% of studies reported the ethnicity of participants. These cohorts were not representative of the UK diabetic population, with underrepresentation of Asian ethnic groups in 90% of trials.

Conclusions:

Representation of elderly patients with comorbidities and those belonging to ethnic minority groups is severely limited in UK based T2DM RCTs of lifestyle interventions. Failure to include a representative population in clinical trial cohorts risks guidance that is not generalisable to the UK diabetic population, potentially exacerbating existing health inequities.

Older people with diabetes are often admitted with falls, dizziness or confusion that may stem from undiagnosed episodes of hypoglycaemia. We examined the use of a 10-day period of round the clock glucose monitoring (CGM), to detect hypoglycaemia in older people with diabetes with symptoms potentially related to hypoglycaemia. 

Methods 

Population: Age 75 years and older, on sulfonylureas and/or insulin, presenting to hospital with a fall and/or symptoms suggestive of unrecognised hypoglycaemia. 

Design: Single-centre, observational study (no change to standard diabetes care). Intervention: 10 days of CGM with Dexcom G6 sensor and Android app on smartphone to continuously transmit data. 

Primary outcomes: Proportion of participants with captured hypoglycaemia; within that group, time spent in the hypoglycaemic range (Battelino T, Danne T, Biester T, et al. Diabetes Care. 2019;42(8):1593-603.). 

Secondary outcomes: Overall time in range; emergency department re-attendances and/or hospital re-admissions for falls, fractures, heart attacks, ischaemic strokes and death within 30 days. REC IRAS project ID: 301286. 

Results 

26 eligible participants of which 13 consented to participate. At the time of writing, nine participants (mean age 81 years) completed the study.

There were no reports of pain or skin reactions from the participants.

Hypoglycaemic events were captured in 3 of 9 participants, with two participants suffering >1 hour below 3.9mmol/L. Only 3 participants achieved >50% time in range target (3.9-10.0mmol/L). 

Discussion 

We have detected significant hypoglycaemic episodes in our participants. CGM should be used more widely in older patients with diabetes who present with falls, dizziness or confusion. 

Limitations include issues around data capture due to participants struggling to navigate the mobile phone app. Despite this, all participants felt that CGM was better than finger-prick glucose testing. Future work is needed to explore how CGM can be deployed after acute admissions in this patient group.

INTRODUCTION

Type 2 diabetes mellitus (T2DM) is associated with poor health outcomes (1) and few people aged >70 years likely benefit from HbA1c <53mmol/mol(2) with ≤64mmol/mol generally accepted in people with moderate-severe frailty. We analysed fallers with T2DM to evaluate their outcomes and frailty status.

METHODS

Older Persons Assessment Service (OPAS) is an Emergency Department service which accepts patients on frailty criteria (aged >70 years, falls, confusion, care dependence, polypharmacy and poor mobility). OPAS databank was retrospectively analysed for people with T2DM admitted with a fall June 2020 to April 2022. Interactions between clinical outcomes with therapeutic agents used, age, Charlson Co-morbidity index (CCMI) and Clinical Frailty Score (CFS) were evaluated.

RESULTS

Six-hundred and seventy-nine patients were assessed; 191 (28.1%) had diabetes with a mean HbA1c 56.7 (IQR: 43.0 – 61.5) mmol/mol, 245 (36.1%) were male. Patients with diabetes had a similar mean CFS (5.3 vs 5.3, p=0.52) and age (83.8 vs 83.2 years, p=0.28) as those without diabetes, but had a higher mean CCMI (5.0 vs 7.0, p<0.001). People with diabetes were more likely to die within 12 months (31.4% vs 25%, p<0.05), and there was a trend to greater mortality in patients with diabetes who used insulin and/or gliclazide compared to those who used other agents (49.6% vs 30.9%, p=0.12).

CONCLUSIONS

Falls are a significant burden, and hypoglycaemic agents may contribute to the greater mortality observed in people with diabetes. Clinician awareness of the poorer prognosis associated with diabetes to support de-prescribing diabetes therapies (2) for patients with significant frailty and HbA1c <64mmol/mol. A frailty assessment should be part of any interaction(3) in the older T2DM patient. We are currently writing a local guideline on Diabetes management in Older Adults in Swansea Bay UHB.