Abstract
This study aimed to delineate changes in the patterns of prolonged grief severity in widowed adults following group bereavement interventions and to identify the predictive factors impacting these patterns. We performed a secondary analysis of data from a randomized trial of 125 Chinese widowed older adults who participated in either LOGBI or DPGBI. Self‐ratings of prolonged grief symptoms via Inventory of Complicated Grief (ICG) were subjected to latent class growth analysis to identify subgroups according to symptomatic changes among baseline, post‐treatment, and 8‐week follow‐up. Multinomial regression analysis examined whether participant classification could be predicted by demographics, loss‐related characteristics, and the relationship with the deceased and baseline symptom severity. In the results, intervention sample comprised 96 completers with an attendance greater than 60%. A three‐class categorization yielded the best model fit for changes in prolonged grief severity for both randomized and intervention samples: improved class (41.6% in the randomized sample, 42.8% in the intervention sample), partial responded class (48.1%/45.5%), and relapse class (10.3%/11.8%). Older widowed adults experiencing a high level of grief at baseline and bereavement following chronic illness had an elevated risk of symptomatic relapse after the completion of intervention. These findings indicated that Chinese widowed adults experienced different changing patterns of prolonged grief severity following group bereavement interventions. Practically, widowed adults with a higher risk of symptomatic relapse might need more intensive or longer term intervention or follow‐up support.