Abstract
The aim of this study was to examine the longitudinal relationship between different levels of cognitively stimulating leisure activity (CSLA) participation and three cognitive functions (i.e., memory, working memory, attention and processing speed) among older adults with mild cognitive impairment (MCI). This study used Health and Retirement Study (HRS) data from 2012 to 2020 (n = 5932) and conducted repeated-measured multivariate analysis of covariance (RM-MANCOVA). The results indicated the following: (a) the high level CSLA participation group were found to have higher levels of memory, working memory, and attention and processing speed than the mid and low-level participation groups. The mid-CSLA participation group was found to have higher levels of working memory, and attention and processing speed than the low level CSLA participation group; (b) significant differences in all three cognitive functions were found between years with a declining slope, but the differences between 2014 and the other years of the data set that were examined were not significant; (c) the high-level CSLA participation group consistently exhibited higher cognitive function levels during the study period and maintained a similar level of cognitive functions compared to the other groups. By incorporating the findings of this study into clinical guidelines, healthcare providers will be able to offer an optimal level of CSLA programming at least three to four times a week. Furthermore, efforts should be made to create strategies to overcome the barriers to CSLA participation encountered by MCI older adults including inadequate caregiver support and financial constraints by establishing public care services and community support networks.