The epidemiology of pain during the last 2 years of life

Ann Intern Med. 2010 Nov 2;153(9):563-9. doi: 10.7326/0003-4819-153-9-201011020-00005.


Background: The epidemiology of pain during the last years of life has not been well described.

Objective: To describe the prevalence and correlates of pain during the last 2 years of life.

Design: Observational study. Data from participants who died while enrolled in the Health and Retirement Study were analyzed. The survey interview closest to death was used. Each participant or proxy was interviewed once in the last 24 months of life and was classified into 1 of 24 cohorts on the basis of the number of months between the interview and death. The relationship between time before death and pain was modeled and was adjusted for age, sex, race or ethnicity, education level, net worth, income, terminal diagnosis category, presence of arthritis, and proxy status.

Setting: The Health and Retirement Study, a nationally representative survey of community-living older adults (1994 to 2006).

Participants: Older adult decedents.

Measurements: Clinically significant pain, as indicated by a report that the participant was "often troubled" by pain of at least moderate severity.

Results: The sample included 4703 decedents. Mean age (SD) of participants was 75.7 years (SD, 10.8); 83.1% were white, 10.7% were black, 4.7% were Hispanic; and 52.3% were men. The adjusted prevalence of pain 24 months before death was 26% (95% CI, 23% to 30%). The prevalence remained flat until 4 months before death (28% [CI, 25% to 32%]), then it increased, reaching 46% (CI, 38% to 55%) in the last month of life. The prevalence of pain in the last month of life was 60% among patients with arthritis versus 26% among patients without arthritis (P < 0.001) and did not differ by terminal diagnosis category (cancer [45%], heart disease [48%], frailty [50%], sudden death [42%], or other causes [47%]; P = 0.195).

Limitation: Data are cross-sectional; 19% of responses were from proxies; and information about cause, location, and treatment of pain was not available.

Conclusion: Although the prevalence of pain increases in the last 4 months of life, pain is present in more than one quarter of elderly persons during the last 2 years of life. Arthritis is strongly associated with pain at the end of life.

Primary funding source: National Institute on Aging, National Center for Research Resources, National Institute on Musculoskeletal and Skin Diseases, and National Palliative Care Research Center.

Publication types

  • Research Support, N.I.H., Extramural
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Aged
  • Aged, 80 and over
  • Arthritis / complications
  • Chronic Disease
  • Cross-Sectional Studies
  • Death*
  • Female
  • Humans
  • Male
  • Pain / epidemiology*
  • Pain / ethnology
  • Palliative Care
  • Prevalence
  • Quality of Life
  • Socioeconomic Factors
  • Terminally Ill
  • Time Factors