Body weight, bmi, percent fat and associations with mortality and incident mobility limitation in older men

Peggy M. Cawthon, Stephanie L. Harrison, Tara Rogers-Soeder, Katey Webber, Satya Jonnalagadda, Suzette L. Pereira, Nancy Lane, Jane A. Cauley, James M. Shikany, Samaneh Farsijani, Lisa Langsetmo

Research output: Contribution to journalArticlepeer-review


How different measures of adiposity are similarly or differentially related to mobility limitation and mortality is not clear. In total, 5849 community-dwelling men aged ≥65 years (mean age: 72 years) were followed mortality over 10 years and self-reported mobility limitations (any difficulty walking 2–3 blocks or with climbing 10 steps) at six contacts over 14 years. Baseline measures of adiposity included weight, BMI and percent fat by DXA. Appendicular lean mass (ALM, by DXA) was analyzed as ALM/ht2. Proportional hazards models estimated the risk of mortality, and repeated measures generalized estimating equations estimated the likelihood of mobility limitation. Over 10 years, 27.9% of men died; over 14 years, 48.0% of men reported at least one mobility limitation. We observed U-shaped relationships between weight, BMI, percent fat and ALM/ht2 with mortality. There was a clear log-linear relationship between weight, BMI and percent fat with incident mobility limitation, with higher values associated with a greater likelihood of mobility limitation. In contrast, there was a U-shaped relationship between ALM/ht2 and incident mobility limitation. These observational data suggest that no single measure of adiposity or body composition reflects both the lowest risk of mortality and the lowest likelihood for developing mobility limitation in older men.

Original languageEnglish (US)
Article number53
JournalGeriatrics (Switzerland)
Issue number2
StatePublished - Jun 2021


  • Mobility limitation
  • Mortality
  • Obesity

ASJC Scopus subject areas

  • Health(social science)
  • Aging
  • Gerontology
  • Geriatrics and Gerontology


Dive into the research topics of 'Body weight, bmi, percent fat and associations with mortality and incident mobility limitation in older men'. Together they form a unique fingerprint.

Cite this