If an end-stage Alzheimer disease patient with difficulty swallowing ends up with a fatal lung infection, does she die from respiratory failure or AD? The answer can be both, since most death certificates allow for multiple entries. Despite this option, Alzheimer disease is infrequently noted as a cause of death, according to a research letter published in yesterday’s Journal of the American Medical Association.

Why AD is underreported is unclear, but “it is partly an issue of Alzheimer disease not being recognized as a terminal illness,” suggested Susan Mitchell, the principal author of the study, in an interview with ARF. The short shrift may have public health ramifications. The National Center for Health Statistics (NCHS) currently lists AD as the fifth leading cause of death in the U.S. in people older than 65, and seventh overall. The new data suggest those stats could be misleading. “People pay attention to those numbers because they help in planning health services,” said Mitchell. She is also concerned that NCHS data could potentially influence research funding or funding for end-of-life care programs.

Mitchell, who is at the Hebrew SeniorLife Institute for Aging Research, Boston, and colleagues examined death certificate data from the Choices, Attitudes, and Strategies for Care of Advanced Dementia at End-of-Life study. This study prospectively followed 323 advanced dementia patients in 22 Boston-area nursing homes. In Massachusetts, conditions contributing to death can be listed in either of two parts on the death certificate, and there is no limit on the number of causal entries that can be made. Death certificates are also fairly uniform from state to state. In spite of this, first author Melissa Wachterman, from Brigham and Women’s Hospital, Boston, found that of the 165 patients who died between 2003 and 2007, dementia was not listed in 37 percent of cases. In the 114 patients who had received a specific diagnosis of AD during their lives, the situation was even worse, with only 27 percent of death certificates listing AD. This is a crucial finding since the NCHS only counts cases listed specifically as AD. “If a patient had Alzheimer’s but dementia was listed on the death certificate, then that person doesn’t get counted in the statistics,” said Mitchell.

This is not the first time underreporting of AD deaths has been documented (see, e.g., Olichney et al., 1995). What makes this data particularly troubling is that it comes solely from patients with very late-stage dementia for whom there is no diagnostic ambiguity. “If AD is not mentioned in the death certificate for these patients, then when will it be used?” asked Mitchell. She suggested a multi-pronged approach based on research and education of clinicians, and policy changes will be needed to address the issue.—Tom Fagan


  1. Under-reporting of AD in death certificates is an old story, dating back to the time Alzheimer disease was considered a presenile dementia. In our study of early onset familial AD in Italy (1), affected members of the kindred, who generally died in their fifties, were uniformly recorded with "broncho-pneumonia" as the cause of death (Italian law or custom at the time allowed only one cause of death to be mentioned). During life, affected family members mostly received a diagnosis of "General Paralysis of the Insane." As late as 1973, after having examined a first cousin of our histologically proven AD proband, who displayed identical symptoms, the senior psychiatrist in charge, to whom I was explaining the interest of this case for research on AD, answered, "Thank you, my dear colleague, for bringing to my attention this most interesting instance of hereditary GPI."


    . [Alzheimer's presenile dementia transmitted in an extended kindred]. Rev Neurol (Paris). 1985;141(3):194-202. PubMed.

Make a Comment

To make a comment you must login or register.


Paper Citations

  1. . Death certificate reporting of dementia and mortality in an Alzheimer's disease research center cohort. J Am Geriatr Soc. 1995 Aug;43(8):890-3. PubMed.

External Citations

  1. National Center for Health Statistics

Further Reading

No Available Further Reading

Primary Papers

  1. . Reporting dementia on the death certificates of nursing home residents dying with end-stage dementia. JAMA. 2008 Dec 10;300(22):2608-10. PubMed.