Can we measure frailty in trials for dementia and mild cognitive impairment?

Conference:

SAPC ASM 2022 - UCLan

Talk Code:

2D.5

Presenter:

Heather Wightman

Twitter:

@gppcglasgow

Co-authors:

Heather Wightman, Terry Quinn, Jim Lewsey, Frances Mair, David McAllister, Peter Hanlon

Author institutions:

University of Glasgow

Problem

Frailty, an age-related decline in physiological reserve, is increasing. There is growing interest in the relationship between physical frailty and cognitive impairment. Despite this, frailty is rarely measured or reported in randomised controlled trials (RCTs) for dementia. Identifying frailty in RCTs is important for assessment of trial representativeness and understanding clinical implications for treatment. This study aims to assess the prevalence of frailty in three RCTs for mild cognitive impairment (MCI) or dementia.

Approach

We analysed individual-level participant data from three RCTs available from the Yale Open Data Access repository. Two trials included participants with MCI, living in the community. One trial included participants with severe Alzheimer's dementia. For each trial, we constructed a frailty index, based on standard procedures based on Rockwood’s ‘cumulative deficit’ model of frailty. For each trial, we identified health related deficits (comorbidities, symptoms, functional limitations and laboratory deficits) which were combined as a non-weighted sum (ranging 0 to 1, higher values indicating a greater degree of frailty). Deficits were selected from baseline data in each of the three trials. For each trial, we calculated the mean frailty index. We also calculated the number of participants in each trial identified as living with frailty (based on a frailty index value of >0.24).

Findings

The trials for MCI recruited 1062 participants (553 female, mean age 70 years) and 994 participants (606 female, mean age 71 years), respectively. The trial for severe Alzheimer’s included 415 participants (332 female, mean age 82 years). Data were sufficiently complete to calculate the frailty index for 1054 (99%), 982 (99%) and 408 (98%) participants, respectively. In the MCI trials, the mean frailty index was 0.142 and 0.133, respectively, with a frailty prevalence of 7.4% and 5.3%. In the Alzheimer’s disease trial the mean frailty index was 0.269 and 60.5% of participants were classified as living with frailty. The 99th centile of the frailty index was 0.29 and 0.32 in the MCI trials and 0.47 in the Alzheimer’s disease trial.

Consequences

Our findings show that it is feasible to measure and report frailty in RCTs for cognitive impairment/dementia, using standard baseline measurements. We found frailty to be present in all three trials. As would be expected, frailty was more common in the population experiencing severe Alzheimer's dementia than in MCI. For all trials, however, the upper limit of frailty was lower than is found in many general-population studies, suggesting trial selection processes may be a barrier to inclusion of people with severe frailty. These findings offer further opportunities to explore the implications for frailty in people with cognitive impairment, such as the relationship between frailty and clinical outcomes within a trial setting, and assessment of the representativeness of trials for cognitive impairment.

Submitted by:

Peter Hanlon

Funding acknowledgement:

David McAllister is funded via an Intermediate Clinical Fellowship and Beit Fellowship from the Wellcome Trust, who also supported other costs related to this project such as data access costs and database licenses (Grant reference: 201492/Z/16/Z). Peter Hanlon is funded through a Clinical Research Training Fellowship from the Medical Research Council (Grant reference: MR/S021949/1).