Trajectories of resilience among widows: A latent transition model

TitreTrajectories of resilience among widows: A latent transition model
Type de publicationJournal Article
Year of Publication2019
AuteursBennett, KMary, Morselli, D, Spahni, S, Perrig-Chiello, P
JournalAging & Mental Health
Mots-cléslatent transition analysis, later life, resilience, widowhood, www2

Objectives: In 2015 we identified three profiles of adaptation following spousal bereavement: Vulnerables; Copers and Resilients (Spahni, Morselli, Perrig-Chiello, & Bennett, 2015). However, adaptation to spousal bereavement is a dynamic process. Thus, we examine the trajectories of the same participants longitudinally over two years. We identify the stability and change in profiles of adaptation to widowhood; probability of stability and change; factors that influence trajectories in profile membership.Methods: Data stem from a longitudinal questionnaire study of 309 older widowed people. The questionnaire included five measures of well-being, serving as the dependent variables of this analysis, and measures of personal resources and contextual factors, including social support, marital happiness, psychological resilience, and demography. Data was analysed using latent transition analysis of the variables loneliness, hopelessness, depressive symptoms, life satisfaction, and subjective health.Results: The analysis replicated the three Wave 1 profiles as the best theoretical fit: Vulnerables; Copers; and Resilients. Stability was most common, but some participants moved to more or less adaptive profiles, the former being more frequent. Younger age, longer time since widowhood, new life perspectives facilitated adaptation. Those transitioning to less adaptive profiles were more likely to be women and older.Discussion: The path to adaptation was not linear. Many of the explanatory variables contributed both to positive and negative adaptation. These include previous caring experience, education, psychological resilience and personal strength. This suggests these explanatory variables do not act in isolation but are likely to interact with each other, and with other, yet not measured, factors.