Evaluation of the optimal cutoff number of types of social participation activities in older people.

INTRODUCTION: Social participation activities have a close association with health aging. However, the clinical significance of numbers of social participation activities and its cutoff value has not been defined.

METHODS: We recruited 516 people aged ≥55 years. Twelve social participation behaviors modified according to Taiwanese culture were investigated, and the adequacy of cutoff number was determined by the area under the receiver operating characteristic curve (AUC) according to the results of cluster analysis of individual activities and scores of the Brief Symptom Rating Scale-5 (BSRS-5) and the Chinese Happiness Inventory (CHI). Demographic, BSRS-5 and CHI data were then compared according to the candidate cutoff numbers.

RESULTS: The distribution of the numbers of social activities suggested that the highest partition of numbers of social activities was 3 in women and 4 in men. The AUC regarding the cluster of activity types was 0.917, with the highest Youden's J value located between 3 and 4. The AUC regarding the cluster of activity types and scores of the BSRS-5 and the CHI was 0.929, with similar cutoffs. If 3 and 4 were used as cutoffs, the between-group differences of both the CHI and the BSRS-5 were significant. More types of social activities had a different engaging frequency with the 3 and 4 cutoffs.

CONCLUSION: Our findings found an adequate cutoff with better differential power in the psychopathology and happiness of older people that provided a basis for application in intervention and policy formation.