The influence of depression on the relationship between loneliness and grief trajectories in bereaved older adults

Abstract

Loneliness is associated with adverse mental health outcomes in older adults. Bereavement triggers intense feelings of loneliness. This pilot study explored the association between baseline loneliness and grief symptom trajectories in bereaved elders and explored if this association is moderated by depressive symptom changes. 56 individuals aged 50 years and older, within 13-months following bereavement, completed assessments. Loneliness was measured at baseline using the UCLA loneliness scale-version 3. Grief and depressive symptoms were measured over 26 weeks using the inventory of complicated grief (ICG) and the 17-item Hamilton Depressive Rating (HAM-D) scales, respectively. Linear regression explored the cross-sectional association between loneliness and grief symptoms, after adjusting for covariates including depressive symptoms. A mixed-effects linear model tested whether baseline loneliness was related to grief symptom trajectory over 26 weeks, after accounting for depressive symptom changes. Loneliness was associated with grief symptom severity at baseline; however, this cross-sectional association was not significant after adjusting for depressive symptoms. Longitudinally, baseline loneliness was positively associated with grief symptom trajectories; however, depressive symptom changes moderated this association. Depressive symptom alterations appear to weaken the loneliness-grief symptom change association. These exploratory findings point to opportunities for interventions targeting loneliness and depression that may reduce grief intensity over time in bereaved elders.

Keywords: Bereavement; depression; grief; loneliness; older adults; prolonged grief disorder; trajectories.

Figure 1.. Longitudinal Inventory of Complicated Grief trajectories during the six-month follow-up.

Figure 1a shows the longitudinal raw ICG score for each participant plotted over the six-month (i.e., 26 week) period. Since participants show different grief symptom severity at the first follow-up timepoint (higher in some, lower in others) compared to baseline, latent class-based model was used over the standard assumptions commonly used, for the longitudinal mixed model analysis. Participants had a higher probability to belong in one of the three latent clusters (or classes) based on their respective ICG trajectories. Figures 1b, 1c and 1d show the latent clusters, with standardized ICG scores, plotted against time, with the center of each cluster being plotted with a bold black line. ICG: Inventory of Complicated Grief