Development of a leisure activity scale for contemporary older adults: Examination of its association with cognitive function

Abstract

Objectives　Leisure activities are important for older adults, not only to maintain their subjective well-being but also to prevent bedridden states. This study aimed to develop a leisure activity scale for contemporary older adults and examine its psychometric properties, based on a previous study from Iwasa et al. (2018). As people who actively engage in leisure activities are reportedly less likely to experience cognitive decline, the relationship between the scale score and cognitive function should be assessed while developing the scale. Specifically, the study was conducted to examine the reliability of the scale and its factor structure, confirm basic statistical characteristics, examine the scale's gender- and age-based differences, and the relationship between the scale score and cognitive function.

Methods　We surveyed Japanese older adults living in a community (aged 70-84 years; N=594) and used data from 306 participants (151 men and 155 women). We developed and administered a scale comprising 11 items that were measured using a 4-point Likert-type scale. Additionally, we used cognitive function scales including the Mini-Mental State Examination (MMSE), the Memory Impairment Screen (MIS), and the Word Fluency Test. Covariates in tests for independent associations between the leisure activity scale score and cognitive function were socioeconomic status, chronic disease, functional capacity, and living alone. We conducted two web surveys with two-week intervals for test-retest reliability purposes and used data from 192 of those participants (aged 70-79 years; 101 men and 91 women).

Results　A confirmatory factor analysis upheld the fact that the scale was comprised of one factor. The scale obtained high indicators of reliability: Cronbach's alpha coefficient (0.81) and test-retest reliability (0.81). The mean, standard deviation, median, skewness, and kurtosis of the scale score were 14.44, 7.13, 15, −0.12, and −0.73, respectively. The analysis of variance for the scale score indicated significant age-based differences (i.e., the score for those who were 70-74 years old was higher than for those who were 80-84 years old) and no significant gender differences. Multiple regression analyses demonstrated that the scale score was significantly and independently correlated with MMSE (β=0.31), MIS (β=0.24), and word fluency (β=0.25).

Conclusion　This study confirmed the psychometric properties of the leisure activity scale, including factor structure, reliability, basic statistical characteristics, no gender differences, significant age-based differences, and relationship to cognitive function. Future studies should examine the longitudinal relationship between the leisure activity scale score and cognition among older adults in community settings.