Concurrent validity, test-retest reliability, and normative properties of the Ignite app: a cognitive assessment for frontotemporal dementia

Digital biomarkers can provide frequent, real-time monitoring of health-related behaviour and could play an important role in the assessment of cognition in frontotemporal dementia (FTD). However, the validity and reliability of digital biomarkers as measures of cognitive function must first be determined. The Ignite computerised cognitive app contains several iPad-based measures of executive function, social cognition, and other domains known to be affected in FTD. Here we describe the normative properties of the Ignite tests, evaluate the associations with gold-standard neuropsychology tests, and investigate test-retest reliability through two healthy controls studies. Over 2,000 cognitively normal adults aged 20-80 years (mean=55.2, standard deviation=15.8) were recruited to complete the Ignite app through a remote data collection study. Significant associations were found between age and performance on several Ignite measures of processing speed ( r =0.42 to 0.56, p <0.001) and executive function ( r =0.43 to 0.62, p <0.001), suggesting the tests are sensitive to cognitive decline observed in normal ageing. A separate cohort of 98 healthy controls were recruited to an observational study (mean age=51.2 years, standard deviation=17.3), completing Ignite at two timepoints (7 days apart), a gold-standard pen and paper neuropsychology battery of corresponding tests, and a user experience questionnaire (10-items). The Ignite tests demonstrated moderate to excellent test-retest reliability (ICCs=0.54 to 0.92) and significantly correlated with their pen and paper counterparts ( r =0.25 to 0.72, p <0.05). The majority of participants (>90%) also rated the app favourably, stating it was enjoyable and easy to complete unsupervised. These findings suggest the Ignite tests are valid measures of cognitive processes, capture a stable picture of performance over time and are well accepted in healthy controls, speaking to the feasibility of administering the app remotely. Therefore, the results have important implications for the utility of Ignite as a cognitive endpoint in upcoming FTD clinical trials.