(PS10-24) Loneliness and Grief: Can Ecological Momentary Assessment Help Better Predict the Persistence of Prolonged Grief?

Abstract Text: Background 

Loneliness following bereavement has been found to prospectively predict prolonged grief. However, most research on loneliness has utilized retrospective self-reports and, thus, is limited by the risk of recall bias. Ecological momentary assessment (EMA), a method used to assess experiences as they are experienced moment to moment in daily life, is less vulnerable to recall bias and can provide richer information about the dynamics of loneliness. Recently, a study of loneliness in bereaved adults using EMA found that individuals with prolonged grief not only experience more frequent loneliness, but also more variable and more intense loneliness. In this study, we examined whether this additional information about the intra-individual dynamics of loneliness could be leveraged to better predict the persistence of prolonged grief. We hypothesized that EMA-assessed loneliness would predict the persistence of prolonged grief symptom severity at 6-month follow-up and that specific loneliness dynamics (i.e., frequency, intensity, and variability) would each uniquely predict the persistence of prolonged grief.  

Methods 

Bereaved adults (n=136) completed the Inventory of Complicated Grief, a measure of prolonged grief, at baseline and at 6-month follow up. After baseline, participants completed six 12-item EMA surveys per day for 17 days. Each survey contained a single item assessing loneliness on a 0-100 scale. Mean survey adherence was 94%. For each participant, we calculated the within-person mean (mean of all responses), frequency (proportion of non-zero responses), intensity (mean of all non-zero responses) and time-dependent variability (root mean square of successive differences) of loneliness.  

Results 

Mean EMA-assessed loneliness predicted prolonged grief symptom severity at 6-month follow-up after controlling for baseline prolonged grief, b=.07, p=.005. Adding EMA-assessed loneliness to the model improved variance accounted for by 2% (R2=.71) relative to what was accounted for by baseline prolonged grief alone (R2=.69). In a separate model, neither frequency (b=2.23, p=.20), intensity (b=.06, p=.06), nor time-independent variability (b=-.15, p=.43) uniquely predicted prolonged grief.  

Discussion 

Consistent with prior research using retrospective reports, EMA-assessed loneliness prospectively predicted prolonged grief symptom severity at 6-month follow-up. However, EMA-assessed loneliness provided relatively modest improvement in prediction and individual dynamic characteristics did not uniquely predict prolonged grief, though loneliness intensity was only narrowly non-significant. Together, these findings leave open the possibility that EMA may be leveraged to better predict the persistence of prolonged grief, but further research is needed to determine whether it can meaningfully advance our ability to identify those whose grief is likely to persist. In future studies, it will be valuable to evaluate the prediction afforded by EMA data for more symptoms and for more dynamic characteristics, ideally in a sample more homogenous in their time since loss, to determine whether EMA-assessed experiences can provide meaningfully greater predictive value.