Digital Cognitive Assessment as a Digital Biomarker

Digital cognitive assessment means administering cognitive tests on a computer, tablet or phone rather than with pencil, paper and a trained administrator. It covers established computerised batteries with decades of use as well as newer app based tests designed for remote self administration.

The advantage is not that the tasks are better. It is that they can be given far more often. A paper battery requires a clinic visit and an administrator, which limits it to a few timepoints; a self administered digital battery can be repeated weekly at home, producing a trajectory instead of two distant points and averaging out the day to day variability that makes single assessments noisy.

The category also has a boundary worth defending. A validated digital cognitive assessment is developed to detect specific impairment and to withstand repetition; a consumer brain training product is not, and the visual similarity between them does not make them interchangeable.

Tasks are drawn from established neuropsychology and translated to a screen: paired associate learning for memory, set shifting and trail making for executive function, simple and choice reaction time for processing speed. Computerised batteries with a long translational research record, and publicly funded batteries developed for large scale use, provide the validated task pool most digital assessments draw from.

Two technical issues dominate. Practice effects mean the same task cannot simply be repeated, so batteries use alternate forms or adaptive item generation. Device dependence means timing sensitive tasks give different values on different hardware, so studies should fix the device or record it as a covariate.

Remote self administration adds a third: nobody is watching, so the testing environment, distractions and effort are uncontrolled in a way clinic testing is not.

The largest application is Alzheimer's disease and mild cognitive impairment research, where digital assessment is used for screening, enrichment of trial populations, and repeated measurement of decline. Reviews of digital technologies in this field have covered their use as biomarkers, as clinical outcome assessments and as recruitment tools, which is a wider role than most digital measures occupy.

Other uses include monitoring cognitive side effects of medication, tracking recovery after stroke or concussion, and assessing cognitive change in conditions with known cognitive involvement such as multiple sclerosis. Measures are reported alongside established instruments such as the Addenbrooke's Cognitive Examination, the Trail Making Test and informant questionnaires, which supply the clinical anchors and the perspective of someone who knows the person.

Technical validation is often good, and clinical validation is often the weak link. Demonstrating that a digital task correlates with a paper test is not the same as demonstrating that change on it corresponds to change that matters to the person, and that second step is missing for many available tests.

Self administration introduces uncontrolled conditions and raises questions about who actually took the test. Performance is affected by education, language, familiarity with technology and sensory impairment, and digital delivery can widen rather than narrow those gaps in older populations.

Practice effects are managed rather than eliminated, and residual improvement across repeated sessions can mask real decline in a study that assumes stability.

• Koo BM, Vizer LM. Mobile technology for cognitive assessment of older adults: a scoping review. Innov Aging. 2019. pubmed.ncbi.nlm.nih.gov
• Gold M, et al. Digital technologies as biomarkers, clinical outcomes assessment, and recruitment tools in Alzheimer's disease clinical trials. Alzheimers Dement (N Y). 2018. pubmed.ncbi.nlm.nih.gov
• Barnett JH, et al. The Paired Associates Learning (PAL) test: 30 years of CANTAB translational neuroscience from laboratory to bedside. Curr Top Behav Neurosci. 2016. pubmed.ncbi.nlm.nih.gov
• Ho EH, et al. NIH Toolbox cognition performance in older adults with normal cognition, mild cognitive impairment and dementia. Arch Clin Neuropsychol. 2025. pubmed.ncbi.nlm.nih.gov