Risk of Progression to Alzheimer's Disease for Different Neuropsychological Mild Cognitive Impairment Subtypes: A Hierarchical Meta-Analysis of Longitudinal Studies

Mild Cognitive Impairment (MCI) is a heterogeneous condition between normal aging and dementia. Upon neuropsychological testing, MCI can be divided into 4 groups: single-domain amnestic MCI (sd-aMCI), multiple-domain amnestic MCI (md-aMCI), single-and multiple-domain nonamnestic MCI (sd-naMCI, md-naMCI). Some controversy exists about whether the risk of progression to Alzheimer's disease (risk-AD) is increased in all MCI subtypes. We meta-analyzed the risk-AD for 4 MCI groups using random-effects metaregression with the Hierarchical Robust Variance Estimator and sample size, criterion for objective cognitive impairment, length of follow-up and source of recruitment as covariates. From a pool of 134 available studies, 81 groups from 33 studies (N = 4,907) were meta-analyzed. All the studies were rated as having a high risk of bias. aMCI is overrepresented in studies from memory clinics. Multivariate analyses showed that md-aMCI had a similar risk-AD relative to sd-aMCI, whereas both sd-naMCI and md-naMCI showed a lower risk-AD compared with sd-aMCI. The risk-AD was significantly associated with differences in sample sizes across studies and between groups within studies. md-aMCI had a similar risk-AD relative to sd-aMCI in studies from memory clinics and in studies in the community. Several potential sources of bias such as blindness of AD diagnosis, the MCI diagnosis approach and the reporting of demographics were associated with the risk-AD. This work provides important data for use in both clinical and research scenarios.