When You Lose Something You Can't Replace: Exploring Loneliness in Bereavement

Abstract

Human beings are inherently social, driven by a fundamental need for connection that promotes safety and well-being (Bowlby, 1982). When a close relationship is abruptly severed by the death of a significant other, this loss can trigger a grieving process that is, for most, a natural part of life. However, for a minority, grief becomes chronic and impairing - a condition recently recognized as Prolonged Grief Disorder (PGD) in both the DSM-5-TR (DSM-5-TR, American Psychiatric Association, 2022) and ICD-11 (ICD-11, World Health Organization, 2019). This dissertation examines the role of loneliness - defined as the negative subjective experience that arises when social relationships are deficient in quality or quantity (Perlman & Peplau, 1981) - in the context of grief. Loneliness is conceptualized as emotional loneliness (EL), the absence of a close emotional bond, and social loneliness (SL), the lack of a broader social network (De Jong Gierveld & Van Tilburg, 2010; DiTommaso & Spinner, 1997; Weiss, 1973). Early theories of grief already suggested that attachment, loneliness, and grief are interconnected (e.g., Bowlby, Weiss), with particular emphasis on the role of EL. Nevertheless, systematic research into these relationships has long been limited. This dissertation therefore addresses the central question of to what extent, and in what ways, loneliness plays a role in grief (Part 1), and how EL and SL relate to attachment and prolonged grief symptoms (PGS; Part 2). Part 1 presents two systematic reviews (89 studies in total) integrating psychological and (geronto)sociological perspectives on loneliness in bereavement. Across studies, loneliness proved highly prevalent after loss. Bereaved individuals reported more loneliness than non-bereaved peers, and widowed individuals more than those divorced or single. Mental and physical challenges and low perceived social support emerged as key correlates. Most studies employed unidimensional measures such as the UCLA Loneliness Scale (UCLA-LS; Russell, 1996), while only a minority differentiated between EL and SL - raising concerns about conceptual precision. Given that EL, rather than SL, is more consistently linked to psychopathology (Walsh et al., 2025), future research should adopt multidimensional measures and longitudinal designs. Part 2 investigates how EL and SL relate to PGS. Longitudinal analyses over four years after partner loss showed that EL - but not SL - was associated with changes in PGS. Increases in PGS were associated with later increases in EL. Cross-sectional analyses in younger and more diverse bereaved samples confirmed that only EL, not SL, was related to PGS, and that the association between insecure attachment styles and PGS was fully mediated by EL. These findings highlight the crucial distinction between EL and SL in understanding grief processes. Together, these findings suggest that EL, rather than SL, is central to grief. Clinically, these findings suggest that greater attention to EL in grief interventions could be beneficial, complementing approaches that focus primarily on enhancing social contact. At the same time, these insights may contribute to a broader social acknowledgment of grief and its accompanying EL - helping to make grief more openly discussed, and more livable, within society.