Table 2.

Summary of evidence relating RAR and neurobiological measures.

Stability (IS)	Lower IS was associated with:  • Cerebral microbleeds (p = 0.06) (19)  • Occiputal periventricular and frontal white matter lesions in Oosterman et al. (20); not global white matter lesions in Zuurbier et al. (19)  • Lower medial temporal lobe volume, but this association was attenuated when adjusting for age (21).
Fragmentation (IV)	Higher IV was associated with:  • Greater levels of amyloid deposition markers in the brain and cerebrospinal fluid (22).  • Lower medial temporal lobe volume (21)  • White matter lesion volumes in Zuurbier et al. (19) but not Oosterman et al. (20)
Goodness of fit measures (pseudo-F)	Better fit was associated with better memory performance, and this association was statistically mediated by hyperactivation in the hippocampus (23)
Amplitude (cosine-based model or nonparametric)	Lower amplitude was associated with lower medial temporal lobe volume, but this association was attenuated when adjusting for age (21).