Iterative and data-driven development of a digital measure of motivation, impulsivity, and risk-taking (Preprint)

BACKGROUND Basal ganglia (BG) dysfunction is the hallmark of several neurological and neuropsychiatric disorders. BG dysfunction impairs cost-benefit decision-making, a brain computational process of trading off between reward and costs, and may underlie clinical observations of neuropsychiatric symptoms such as amotivation, impulsivity, and risk-taking. There is an unmet need for a remote digital monitoring tool for cost-benefit decision-making that can be administered repeatedly by a patient at home to provide a longitudinal proxy of basal ganglia functioning to inform clinical practice and drug development efforts. OBJECTIVE This report describes a data-driven and iterative prototyping effort of a smartphone-based, cost-benefit, decision-making test battery developed to measure basal ganglia functioning in three cost domains: effort, delay, and risk. Via an iterative process consisting of multiple rounds of data collection with online community-based samples, we aimed to establish feasibility, proof-of-concept, test-retest reliability over up to 4.5 months, and convergent validity with relevant clinical symptoms, namely apathy and impulsivity. METHODS We converted a lab-based, decision-making test battery to a URL-based test administered on private smartphones. An online pool of n = 879 volunteers aged 40-80 years participated. Using a data-driven approach, we iteratively (6 rounds) refined and shortened the test to a prototype with a target participant burden of less than 5 minutes while preserving its psychometric properties. Proof of concept, test-retest reliability (intraclass correlation coefficients (ICCs)), and convergent validity with the Starkstein Apathy Scale (SAS) and Barrat Impulsiveness Scale (BIS-11) were established for the final version. RESULTS Consistent monotonic effects of cost levels in all cost domains were observed, demonstrating proof of concept. Measures of motivation, impulsivity, and risk-taking showed good to excellent test-retest reliability (ICCs 0.69 – 0.97). Metrics of impulsivity showed convergent validity with its clinical anchor, BIS-11 (t = -2.83, p = 0.005)), and the metric of motivation showed convergent validity with SAS (t = -2.29, p = 0.022). CONCLUSIONS This accessible, short, scalable, and replicable digital smartphone-based decision-making test battery may represent an efficient and robust means to estimate cost-benefit decision-making associated with basal ganglia functioning and its changes over time. CLINICALTRIAL This is not a RCT