Neuropsychological assessment in cognitively healthy nonagenarians and centenarians: an updated systematic review and meta-analysis

Abstract

The assessment of cognitive impairment in individuals aged 90 years and older is based on normative data estimated in the younger elderly. A first systematic review had provided the reference values in the oldest population for a few neuropsychological tests. Robust estimations on additional cohorts are needed.

We designed a systematic review update (PROSPERO: CRD42022347327) encompassing the literature published from 2015 to June 2024, to include studies reporting raw neuropsychological test scores from at least ten individuals aged >90 without dementia. All types of cohort studies were eligible. The data set from the previously published systematic review and the studies retrieved in the update process were pooled. Random effect meta-analysis was applied to estimate the updated mean and cut-off values.

The systematic workflow provided 11 articles eligible for data abstraction. Based on the pooled data we estimated the updated reference scores for MMSE, BNT-SF, Semantic Fluency, TMT-A, TMT-B, Digit span forward and Digit span backward. Moreover, we estimated for the first time the reference values for the Word List Immediate and Word List Delayed tasks.

The systematic review provided an updated set of norms for the neuropsychological test that are frequently used to assess the cognitive profile in individuals aged >90 years. The heterogeneity of the assessments limited the quantitative synthesis. Further studies on large cohorts are needed to stratify the normative values by age, gender and education. The identification of country- and population-specific values would help clinicians and researchers to characterise the cognitive profile in the oldest old individuals.