Is the Stage Theory of Grief Empirically Valid?

Home>
Specialty Care>
Psychiatry>
Summary and Comment

Is the Stage Theory of Grief Empirically Valid?

This attempt at confirming the popular theory falls short.

Some theoreticians have conceived of the human response to bereavement and loss as a sequence of characteristic stages beginning with denial, progressing to various kinds of emotional turmoil, and ending with acceptance. The theory is appealing to clinicians. These researchers attempted a test of Jacobs’s four elements — disbelief, separation distress (i.e., yearning, anger), depression, and acceptance — by prospectively examining their frequency at three successive time points over 2 years. Each item was assessed with a single question.

Recruiting primarily elderly bereaved spouses, the investigators had a 55% response rate. Of the 317 who responded, 26% were then excluded because of bereavement by traumatic death, missing data, and "complicated grief" reactions, resulting in a 41% participation rate. Many of the 233 remaining participants failed to provide data at all time points.

The authors’ graphic representation of the response trajectories shows that the participants’ grief did not occur in stages. In fact, all of Jacobs’s elements except acceptance peaked within 6 months and occurred simultaneously for much of that period, and acceptance was highest at each time point. The authors assert, nevertheless, that the responses achieved their maximal values in a sequence consistent with the theory.

Comment: While attempting to study stages of grief, the authors of this study overstated their findings. In addition to reviewing a large body of literature selectively, they used single-item outcomes derived post hoc from longer scales, conducted cross-sectional (rather than longitudinal) analyses that left out a large portion of the study population, and drew oversimplified conclusions that reinforce formulaic, unhelpful ways of thinking about bereavement. For bereaved people not progressing through expected stages, this could lead to unwarranted intervention, unnecessary increased distress, or both.

— Peter Roy-Byrne, MD, and M. Katherine Shear, MD

Published in Journal Watch Psychiatry March 26, 2007