Results Of the 59,545 participants, 6,431 (10.8%) were frail onset at follow-up. After multiple imputations (minimum 64,212, maximum 64,287), the risk of frailty onset at follow-up was lower for eight types of social participation, excluding senior citizens' clubs, (nursing care [risk ratio; 0.91], paid work [0.90], volunteer groups [0.87], neighborhood associations [0.87], learning or cultural groups [0.87], activities intended to teach skills or pass experiences to others [0.85], hobby groups [0.81], and sports groups or clubs [0.80]; P<0.05), than no social participation. Additionally, individuals who participated in more types of social participation were at a lower risk of frailty than those with no social participation (P for trend <0.001).