The Masking of Mourning: Social Disconnection After Bereavement and Its Role in Psychological Distress

Abstract

Social support has been shown to facilitate adaptation after bereavement in some studies but not others. A felt sense of social disconnection may act as a barrier to the utilization of social support, perhaps explaining these discrepancies. Factorial and psychometric validity of the Oxford Grief-Social Disconnection Scale (OG-SD) was tested in a bereaved sample (N = 676). A three-factor solution (negative interpretation of others' reactions to grief expression, altered social self, and safety in solitude) fit the data best and demonstrated excellent psychometric validity. A second three-wave longitudinal sample (N = 275) recruited 0 to 6 months following loss and followed up 6 and 12 months later completed measures of prolonged grief disorder, posttraumatic stress disorder, depression, and the OG-SD at each time point. High levels of baseline social disconnection were associated with concurrently high psychological distress. The extent to which social disconnection declined over time predicted resolution of psychological distress.

Keywords: affective disorders; grief; posttraumatic stress disorder; psychometrics; social cognition.

Fig. 1.

Structural equation model of the curve of factors model of psychological distress and the latent growth curve of social disconnection. Unstandardized coefficients are shown. PGD = prolonged grief disorder score (Prolonged Grief Disorder Inventory) with social symptom composite item removed; PTSD = posttraumatic stress disorder score (Posttraumatic Stress Disorder Checklist for DSM–5) with social symptom composite item removed; DEP = depression score (Patient Health Questionnaire); SD = social disconnection score (Oxford Grief-Social Disconnection Scale); i = intercept; s = slope. Scalar invariance was specified using the marker variable approach. The residual variances among the same indicators over time were correlated but are not shown in the figure for simplicity. Solid lines indicate statistically significant paths. Broken lines indicate nonsignificant paths. χ²(51) = 160.28, χ²:df = 3.14, comparative fit index = .96, Tucker-Lewis index = .95, root mean square error of approximation = 0.088, standardized root mean square residual = .05. Asterisks indicate significance of parameters (*p < .05, ***p < .001).