Low testosterone is associated with decreased function and increased mortality risk: a preliminary study of men in a geriatric rehabilitation unit

J Am Geriatr Soc. 2004 Dec;52(12):2077-81. doi: 10.1111/j.1532-5415.2004.52562.x.


Objectives: To evaluate whether low testosterone levels are associated with greater depression or poorer function in a geriatric rehabilitation unit.

Design: Retrospective review.

Setting: Geriatric rehabilitation unit.

Measurements: Low testosterone levels were defined as total testosterone of 3.0 ng/mL or less or free testosterone of 9.0 pg/mL or less. Age, ethnicity, weight, depression, ambulation, length of rehabilitation, and 6-month rehospitalization and mortality rates were obtained. Overall illness severity was determined using the Cumulative Illness Rating Scale for Geriatrics.

Results: Low testosterone levels were present in 29 of 44 (65.9%) men. There were no significant differences between men with low and normal testosterone levels in ethnicity, age, weight, depression, and overall illness severity. Lower testosterone levels were correlated with decreased ability to ambulate and transfer (Spearman P>.34; P<.05). There were no significant differences between men with low and normal testosterone in length of stay on the rehabilitation unit (mean+/-standard deviation= 19.6+/-11.6 vs 17.7+/-17.5 days, P=.68) or rehospitalization rates (41.4% vs 26.7%; P=.34). Men with low testosterone had a trend toward increased 6-month mortality (31.0% vs 6.7%; chi(2)=3.3, P=.07) and shorter survival time (log rank=3.2; df 1, P=.07). After entering testosterone and variables with potential prognostic significance for mortality in a stepwise manner in a Cox regression analysis, there was a significant mortality risk associated with low testosterone (hazard ratio=27.9, 95% confidence interval=2.0-384.0; P=.01).

Conclusion: Low testosterone levels were correlated with decreased physical function and increased risk for 6-month mortality. Prospective studies with larger sample sizes and better standardized testosterone measures are needed to confirm these findings.

Publication types

  • Research Support, Non-U.S. Gov't
  • Research Support, U.S. Gov't, Non-P.H.S.

MeSH terms

  • Activities of Daily Living*
  • Aged
  • Aged, 80 and over
  • Depression / blood
  • Depression / epidemiology*
  • Humans
  • Male
  • Middle Aged
  • Mortality*
  • Proportional Hazards Models
  • Retrospective Studies
  • Statistics, Nonparametric
  • Survival Analysis
  • Testosterone / deficiency*
  • Washington / epidemiology


  • Testosterone