Transition in the cognitive continuum from normal cognition to dementia: the interplay of lifelong cognitive reserve with brain aging and genetic predisposition

Abstract

AbstractBackgroundHigher cognitive reserve (CR) has been associated with lower risks of cognitive decline and dementia. In this study, we sought to investigate associations of lifelong CR capacity with transitions of cognitive phenotypes in older adults, while accounting for age, sex, APOE genotype, and pathological load of brain aging.MethodThis population‐based Swedish National study on Aging and Care‐Kungsholmen included 2631 participants (age≥60 years, 61.4% female) free of dementia at baseline (2001‐2004); among them 517 had brain MRI. At baseline, lifelong CR capacity was assessed by integrating early‐life education, midlife occupational complexity, later‐life social network and leisure activity. The load of pathological brain aging was assessed using volumes of grey matter (GM), white matter, hippocampus, and white matter hyperintensity, and count of perivascular space and lacunes. Cognitive impairment no dementia (CIND) and dementia were defined at baseline and across up to 15 years until 2016‐2019. Data were analyzed using the Markov multi‐state model.ResultDuring the mean 10.7 years of follow‐up, 9.0% had incident dementia (95%CI: 7.8%–10.3%). Among people free of CIND at baseline (n = 2023), 27.4% had incident CIND during follow‐up (25.3%–29.6%). Higher lifelong CR capacity (mean value of composite CR score = 0.2; range: ‐4.2‐3.6) was associated with a lower risk of direct transitions from non‐dementia (normal or CIND) to dementia (multivariable‐adjusted HR = 0.84; 95% CI 0.73–0.95) and from normal cognition to CIND (0.79; 0.73–0.85), but not with the direct transitions from normal cognition to dementia (0.86; 0.72–1.03), from CIND to dementia (0.93; 0.76–1.14), or from CIND to normal cognition (1.15; 0.98–1.35). Higher lifelong CR capacity was associated with a lower risk of direct transition from normal cognition to dementia in people with APOE‐ε4 and those with high GM volume (2nd‐4th quartiles), but not in those without APOE‐ε4 or with low GM volume (1st quartile) (P‐for‐interaction<0.05). There was no statistical interaction of CR with age, sex, or other brain MRI markers on cognitive transitions.ConclusionHigher lifelong CR capacity may benefit late‐life cognitive status by reducing the risk of cognitive deterioration to dementia, especially in people with genetic predisposition or with low neurodegenerative load.