WAIS-IV clinical validity samples of individuals with Alzheimer’s disease (AD; n ≤ 42) and mild cognitive impairment (MCI, n ≤ 59) reveal subtest group differences compared to demographically matched healthy controls. The Information subtest, often considered a “hold” test, had the second highest effect size for subtest differences between AD and controls. The MCI cohort did not significantly differ from controls on IN but differed on Coding and PRI subtests (ES range .50 to .66), possibly related to the inclusion of both amnestic and non-amnestic subtypes. The Information subtest examines general factual knowledge, and is arguably a measure of semantic memory, a common deficit in AD. Here we report the WAIS-IV profiles to evaluate amnestic and non-amnestic subtype differences with the hypothesis that Information will be the lowest VCI subtest score in amnestic subtypes at both the MCI and dementia stages.

Participant data were exported from the National Neuropsychology Network data repository. Out of 11033 participants with available ICD-10 diagnoses, 947 (54% female, 83% white, age range 18 to 90+) were clinically diagnosed with the following conditions of interest: MCI (n = 403), dementia (n = 142), and unspecified neurocognitive disorder (n = 402). Participants in all groups were divided into “amnestic” and “non-amnestic” subtypes using the psychometric component of Petersen’s criteria (≥ 1.5SD below the normative mean on at least one memory measure), using delayed recall from one or more of the following tests: the WMS-IV Logical Memory and Verbal Paired Associates subtests, CVLT-3, and HVLT-R. We conducted ANOVA analyses to compare amnestic and non-amnestic groups across WAIS-IV subtests. The Games-Howell correction for multiple comparisons was applied.

Significantly lower scores were seen on Matrix Reasoning, Digit Span, Information, and Similarities in aMCI compared to naMCI.  In the dementia group, the Information mean difference scores approached significance in the amnestic v. non-amnestic subgroup (p = .085). In the unspecified diagnostic group, the amnestic subgroup had significantly lower scores on all WAIS-IV subtests. The IN subtest was the lowest VCI subtest mean score in all amnestic groups. For a subset of participants, neurodegenerative etiology was specified (i.e., AD n = 66, cerebrovascular disease n = 45, and PD/Lewy Body n = 16). Trend-level results emerged with lower mean scores on the Information subtest in the AD versus PD/Lewy Body group (p = .06). While not significant, the AD group had a lower mean score than the cerebrovascular group.

Differences emerged in WAIS-IV subtest profiles between amnestic and non-amnestic groups. Findings indicate that the Information subtest is a useful test as part of a workup for amnestic syndromes, prior to the dementia stage, and that its status as a “hold” test should be reassessed.