RESUMO
OBJECTIVE: To estimate the prevalence of frailty in older adults in the IMIAS population, to examine associations between lifelong domestic violence and frailty and possible pathways to explain these associations. METHODS: A cross-sectional study with 2002 men and women in the International Mobility in Aging Study, aged between 65 and 74 years old living in five cities of Tirana (Albania), Natal (Brazil), Kingston and Saint-Hyacinthe (Canada), and Manizales (Colombia). Domestic physical and psychological violence by family and intimate partner was assessed by the Hurt, Insult, Threaten and Scream (HITS) scale. Fried's phenotype was adopted to define frailty. Logistic regressions were fitted to estimate between frailty and lifelong violence. Mediation analyses using the Preacher and Hayes method was used to examine potential health pathways. RESULTS: Frailty prevalence varies across cities, being lowest in Saint-Hyacinthe and Kingston, and highest in women in Natal. Women had a higher prevalence in Tirana and Natal. Adjusting for age, sex, education and research city, those reporting childhood physical abuse (CPA) had higher odds of frailty (ORâ¯=â¯1.68; 95% CI: 1.01; 2.78); those who had been exposed to psychological violence by their intimate partner had also higher odds of frailty (ORâ¯=â¯2.07; 95% CI: 1.37; 3.12). CPA effect on frailty was totally mediated by chronic conditions and depression symptoms. Effects of psychological violence by intimate partner were partially mediated by chronic conditions and depression symptoms. CONCLUSIONS: Childhood physical abuse and psychological violence during adulthood leave marks on life trajectory, being conducive to adverse health outcomes and frailty in old age.