Best lung function equations for the very elderly selected by survival analysis

Area of Science:

• Gerontology
• Pulmonology
• Epidemiology

Background:

• Accurate prediction of lung function is crucial for assessing health in elderly populations.
• Existing lung function prediction equations may not accurately represent nonagenarians due to potential selection bias in older cohorts.

Purpose of the Study:

• To identify which lung function prediction equations best predict survival in a cohort of nonagenarians.
• To compare the predictive accuracy of equations derived from the US National Health and Nutrition Examination Survey (NHANES) III versus equations including elderly subjects.

Main Methods:

• A population-based cohort of 2262 Danes born in 1905 had lung function, grip strength, and Mini Mental State Examination (MMSE) scores measured in 1998.
• Mortality was tracked until 2011. Forced expiratory volume in 1 second (FEV1) standardized residuals (SR) were calculated using various prediction equations for 592 subjects with complete data.
• Cox proportional hazard regression was used to analyze the association between FEV1SR and mortality, controlling for MMSE, grip strength, and sex.

Main Results:

• Lower FEV1SR was a significant predictor of mortality in nonagenarians, even after adjusting for cognitive function, grip strength, and sex.
• Equations from the NHANES III (1999) demonstrated a better spread of median survival across FEV1SR quartiles compared to equations derived from elderly cohorts.
• The NHANES III equations yielded hazard ratios for death of 1, 1.16, 1.32, and 1.60 for increasing FEV1SR quartiles, indicating superior survival prediction.

Conclusions:

• Extrapolating lung function prediction from NHANES III equations provides more accurate survival predictions in nonagenarians than using equations derived from elderly subjects.
• The findings suggest that NHANES III equations may be more generalizable to very old populations, potentially mitigating selection bias present in elderly-specific cohorts.
• This study provides valuable insights for the development of future lung function prediction equations tailored for the elderly, emphasizing the need for representative cohort data.