Longitudinal change in neuropsychological performance using latent growth models: a study of mild cognitive impairment

Abstract

The goal of the current study was to examine cognitive change in both healthy controls (n = 229) and individuals with mild cognitive impairment (MCI) (n = 397) from the Alzheimer's Disease Neuroimaging Initiative (ADNI). We applied latent growth modeling to examine baseline and longitudinal change over 36 months in five cognitive factors derived from the ADNI neuropsychological test battery (memory, executive function/processing speed, language, attention and visuospatial). At baseline, MCI patients demonstrated lower performance on all of the five cognitive factors when compared to controls. Both controls and MCI patients declined on memory over 36 months; however, the MCI patients declined at a significantly faster rate than controls. The MCI patients also declined over 36 months on the remaining four cognitive factors. In contrast, the controls did not exhibit significant change over 36 months on the non-memory cognitive factors. Within the MCI group, executive function declined faster than memory, while the other factor scores changed slower than memory over time. These findings suggest different patterns of cognitive change in healthy older adults and MCI patients. The findings also suggest that, when compared with memory, executive function declines faster than other cognitive factors in patients with MCI. Thus, decline in non-memory domains may be an important feature for distinguishing healthy older adults and persons with MCI.

Fig. 1

Model Diagram of a Multiple Indicator Parallel Process Latent Growth Model for the ADNI Neuropsychological Test Battery. This figure shows the parameterization of the model of growth processes for memory and other cognitive domain (executive function, visuospatial ability, language, and attention) across five study visits in ADNI (baseline, 6, 12, 24 and 36 months). Latent variable outcomes (in circles) representing cognitive domains are measured by observed cognitive indicators (in squares) at each ADNI study wave. Latent variables capturing baseline or initial status (intercept) and annual trajectories (slope) were estimated using the latent variable outcomes at each visit. Numbers on arrows from growth parameters to latent variable outcomes are fixed factor loadings representing time steps from the baseline visit. The domain-specific indicators used at each study visit for executive function, visuospatial ability, language, and attention are shown in Table 1 and described in the Methods section

Fig. 2

Model-estimated trajectories of performance in cognitive factors by diagnostic status: Results from healthy control and MCI groups (n=626)