Pooling data from studies on prolonged grief: the Multi-region Archive of Research data on Bereavement and Loss from Empirical Studies (MARBLES) project.

Introduction: Prolonged Grief Disorder (PGD) was newly included in the 11th edition of the International Classification of Diseases (ICD-11) and the text revision of the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5-TR). There is a need to increase our understanding of the prevalence, nature, and risk factors of this condition. Research in this area can be advanced by pooling existing datasets.Objective: The MARBLES (Multi-region Archive of Research data on Bereavement and Loss from Empirical Studies) project was initiated to create a consortium of researchers conducting studies on PGD. Its goal is to harmonize and combine data from separate studies into a unified, individual participant data-archive, boosting bereavement research.Methods: To date, data from 16,666 bereaved individuals have been compiled from 32 datasets across nine countries. All datasets include data on PGD symptoms and sociodemographic and loss characteristics; most also include data on emotion regulation and concurrent symptoms. For preliminary analyses presented in this paper, data on sociodemographic and loss-related characteristics and on PGD, posttraumatic stress, and depression symptoms were harmonized.Results: Regarding the preliminary analyses, participants' mean age was in their mid-40s, about 70% of the participants were female, one in four participants had lost a partner, and one in four participants experienced an unnatural loss. Analyses (with multiple imputation) indicated that 18.3% met criteria for probable PGD (ICD-11).Conclusion: The MARBLES archive demonstrates that it is feasible to build a FAIR archive of PGD symptom data. Although the archive can already be used to study the prevalence, pathogenesis, and predictors of PGD symptoms and associated emotional problems, it is intended as an evolving resource. Its potential will expand as additional datasets are added and use of the archive grows. Future extensions should prioritize the inclusion of underrepresented groups (e.g. migrants) and currently missing social, cognitive, and neurobiological variables.