Cross-National Disability Comparison Among Older Adults Living in the United States,Mexico,and South Korea Using Rasch Analysis

BACKGROUND: As the availability of national microdata have emerged, cross-national comparisons evaluating patterns of population aging have become available in national data sets such as the Korean Longitudinal Study of Aging (KLoSA), the U.S. Health and Retirement Study (HRS), and the Mexican Health and Aging Study (MHAS). Cross-national disability comparison studies allow researchers to examine differences in disparities across countries globally. Information obtained from these studies can be used to inform policymakers and health care providers to promote healthy aging among older adult populations. However, differences in national survey designs cause challenges for accurate comparison.

The first objective of this study was to create a common disability scale that met Rasch model expectations for disability comparison across three national data sets (KLoSA, HRS, and MHAS) using survey items related to basic activities of daily living (ADLs), instrumental activities of daily living (IADLs), and mobility. On successful creation of a common disability scale, the secondary objective of this study was to compare levels of disability differences among older adults living in Korea, the United States, and Mexico.

METHOD: A cross-sectional design was used for retrieval and analysis of data from 4,134 older adults who completed 17 ADL and IADL items from the 2012 KLoSA database and from 6,367 older adults from the 2012 HRS and 10,017 from the 2012 MHAS database who completed 9 ADL and IADL items and 13 mobility items. Participants were age >65 and lived in community settings. The Rasch common-item equating method was applied to create a common disability scale using the nine common items across the three surveys. We conducted multivariate regression analysis to compare disability levels using the Rasch-calibrated person measures across the three countries.

RESULTS: The mean age of the sample was 74.6 (SD = 6.9), and the majority were female (56.9%) and unemployed (76.9%). The common measurement scale met the expectations of the Rasch model, including unidimensional construct and good internal consistency (person reliability = .90). The multivariate regression model revealed that Korean older adults had significantly higher disability status (M = 1.92 logits, SD = 2.33) compared with older adults living in Mexico (M = 2.81 logits, SD = 2.8) and in the United States (M = 3.65 logits, SD = 3.2). Increased age (ß = -0.08), female gender (ß = -0.45), single marital status (ß = -0.14), higher body mass index (ß = -0.07), history of stroke (ß = -1.17), history of arthritis (ß = -0.78), and history of diabetes (ß = -0.28) were significantly related to a higher levels of disability status.

CONCLUSION: The Rasch model indicates that older adults living in Korea have higher disability levels than older adults living in the United States and Mexico. We speculate that the majority of the differentials across the three countries may be attributable to differential rates of aging among the populations or different health polices affecting aging populations. Future research is recommended to investigate the variation in disability-driving factors for older adults living in Korea and Mexico. Furthermore, adoption of U.S. health policies should be considered to improve healthy aging among older adults living in the two countries.

IMPACT STATEMENT: The proposed research methods successfully compared disability levels across three countries. The study methods can be applied to 30 similar population-based longitudinal aging surveys, and the estimated disability levels will contribute to future research directions in global aging studies targeting healthy aging among community-dwelling older adults.

References