Is the five stages of grief model supported by evidence?

Largely no. People do often feel emotions such as disbelief, anger, sadness and acceptance after a loss, but there is little support for the idea that these arrive in a set sequence of stages. A review by Stroebe, Schut and Boerner concluded that there is evidence people experience some of these reactions some of the time, but little to support their sequential development in stages. One prominent study did report a fit with the sequence, and that study is itself disputed. [1]

The five stages come from Elisabeth Kübler-Ross's 1969 book On Death and Dying, which described reactions among dying patients, not bereaved people. According to the review, her knowledge came from contact with over two hundred dying patients, not from systematic empirical investigation. The psychiatrist Colin Murray Parkes, quoted there, described the book as a collection of case studies in the form of conversations with dying patients. Other summaries note that critics object to the stages being applied too rigidly and linearly, and that Kübler-Ross herself aimed to describe experiences a patient might have, not a fixed path. [1] [2]

The main study in favor is Maciejewski and colleagues' 2007 analysis in JAMA of 233 bereaved people in Connecticut, followed from 1 to 24 months after a loss. Counter to stage theory, disbelief was not the initial dominant indicator, and acceptance was the most frequently endorsed item throughout. However, when the researchers rescaled each indicator, the peaks occurred in the predicted order: disbelief, yearning, anger, depression, then acceptance. They calculated that this exact order would occur by chance with a probability of .008. They acknowledged limitations, including that nobody was interviewed in the first month after the death. [3]

Critics reject this reading. Stroebe and colleagues report that letters to JAMA raised questions about the study's design and methods, and that Prigerson and Maciejewski later suggested relabeling the stages as multidimensional grief states that evolve and diminish in intensity over time. A 2010 replication attempt by Holland and Neimeyer found results neither corroborative nor entirely inconsistent with stage theory. [1]

Other research that did not test stages directly also counts against them. In studies by George Bonanno and colleagues tracking people before and after a loss, nearly half of one sample (46%) were resilient, with stable low distress. Only 11% showed the pattern traditionally assumed to be common, in which distress is high after the death and then declines. In other caregiver studies, about 17% showed that pattern. [1]

The prevailing view among grief researchers is that the stage model is not well supported. Stroebe, Schut and Boerner argue it should be discarded, citing absent empirical evidence, weak conceptual clarity, little practical use for identifying people at risk, and possible harm to those who feel they are not grieving the expected way. A separate review in the research literature calls the persistence of the theory stubborn, given how little evidence supports it, and points to alternatives such as continuing bonds and tasks-of-grieving models. [1] [4]