Do changes in frailty, physical functioning, and cognitive functioning predict mortality in old age? Results from the...

Area of Science:

• Gerontology
• Public Health
• Biostatistics

Background:

• Accurate survival prediction is vital for clinical decisions in older adults.
• The value of health indicators and their changes for mortality prediction is unclear.
• This study investigates if changes in health indicators improve mortality predictions in old age.

Purpose of the Study:

• To assess the added value of current health status and changes in health indicators for mortality prediction in older adults.
• To compare the predictive power of various health indicators including frailty, physical performance, and cognitive function.
• To determine if changes in health indicators over time improve mortality predictions.

Main Methods:

• Population-based prospective cohort study of 995 community-dwelling older adults (aged 68-92).
• Two measurements of health indicators (frailty index, phenotype, grip strength, walking speed, MMSE) at a three-year interval.
• Cox regression analysis and Area Under the Curve (AUC) calculation to assess mortality prediction improvements.

Main Results:

• Age and sex alone predicted five-year mortality with an AUC of 72.8%, lowest in the oldest old.
• Current health indicators added 0.7-3.3% to AUC; three-year changes added 0.0-1.1%.
• Frailty index showed the most improvement in mortality prediction, followed by frailty phenotype, MMSE, grip strength, and walking speed.

Conclusions:

• Current health status significantly improves mortality predictions compared to changes in health.
• The predictive value of current health status is highest in the oldest old.
• The overall added value of these health indicators to age and sex models was limited.