Psychosomatic Medicine 66:973-974 (2004)
© 2004 American Psychosomatic Society
LETTERS TO THE EDITOR
POSTPONING DEATH: ANOTHER FAILURE TO REPLICATE
Jonathan D. Lane and
Richard D. Lane, MD, PhD
Washington University in St. Louis, St. Louis, MO
Department of Psychiatry, University of Arizona, Tucson, AZ
We would like to share additional data that address the question of whether death can be postponed. As noted in your recent miniseries (1), Phillips et al. (2) analyzed the months of birth and death of 348 people listed in Four Hundred Notable Americans and concluded that death could be postponed. We sought to replicate and extend this finding in two ways: by analyzing a larger data set of 540 people (the deceased among 1851 individuals listed on the Web site "Who’s Alive and Who’s Dead" [www.whosaliveandwhosdead.com] as of February 03, 2001) and by increasing the temporal resolution, comparing the week of death rather than the month relative to the birth date. If death could be postponed, we predicted a dip before and an increase after the person’s birthday, particularly during the 4 weeks immediately before and after the birthday. We also predicted a more random pattern during the rest of the year.
Ignoring the year of birth or death, we assigned a negative value (minus weeks) to deaths that occurred in the 6 months before the birth date and a positive value (plus weeks) to deaths that occurred in the 6 months after the birth date. The number of deaths that occurred each week before or after one’s birthday is displayed in Figure 1.
There were 267 deaths during minus weeks and 271 deaths during plus weeks (two subjects were excluded because they died 183 days from their birth date). There were 51 deaths during the first 4 minus weeks and 47 deaths during the first 4 plus weeks. There were 15 deaths during the first and 16 deaths during the fourth plus week. The expected number of deaths per week (538/52) was 10.35 or 1.92%. Using the binomial distribution, the probabilities of at least 15, at least 16, and at most 7 deaths (during minus week 2) per week were .0997, .0593, and .189, respectively. None of these values met the minimum probability threshold of .05 for statistical significance.
To examine the trend across the entire 12-month period, we subtracted the number of deaths during minus weeks from the number of deaths during plus weeks because we expected the number of deaths to be greater after the birthday than before. The difference between the number of deaths during the plus weeks and minus weeks is shown in Figure 2.
It is apparent from inspection of Figure 2 that the pattern observed during the 4 weeks before and after the birthday does not differ substantially from that observed during the rest of the year. In fact, there were nine more deaths during minus week 2 than during plus week 2, a finding in the direction opposite of that predicted.
Based on methodology similar to that previously published (2), we conclude that there is no evidence that death can be postponed. This is in agreement with the conclusion from the recently published miniseries (1). However, we should add that in our view, the hypothesis in question has not been adequately tested. Just as Schmale (3) postulated that giving up may be a final common pathway preceding adverse changes in physical health, it is reasonable, in light of evidence of the health-damaging effects of negative emotions (4) and the health-promoting effects of positive emotions (5), to hypothesize that in certain circumstances, a greater will to live may contribute to the postponement of death. A more refined test, in our view, would be to test this hypothesis in relation to an occasion that is meaningful to each specific person (eg, a child’s wedding). Another is to consider cause of death. There are some clinical contexts, such as chronic cardiac or pulmonary disorders, in which such psychological factors could conceivably play a role, whereas such effects should not be evident in situations in which the person has no control over the timing of death, such as a natural disaster or homicide.
REFERENCES
- Skala JA, Freedland KE. Death takes a raincheck. Psychosom Med 2004; 66: 382–6.[Abstract/Free Full Text]
- Phillips DP, Feldman KA. A dip in deaths before ceremonial occasions: some new relationships between social integration and mortality. Am Social Rev 1973; 38: 678–96.
- Schmale AH. Giving up as a final common pathway to changes in health. Adv Psychosom Med 1972; 8: 20–40.[Medline]
- Creed F. Stress and psychosomatic disorders. In: Goldberger L, Breznitz S, editors. Handbook of stress: theoretical and clinical aspects. New York: Free Press; 1993: 496–510.
- Danner DD, Snowdon DA, Friesen WV. Positive emotions in early life and longevity findings from the nun study. J Pers Soc Psychol 2001; 80: 804–13.[CrossRef][Medline]
