Translating Mortality Hazard Ratios into Remaining-Lifespan Measures in Older Adults Using Japanese Complete Life...

Background/Objectives: Mortality hazard ratios (HRs) do not directly express remaining lifespan or fixed-time survival. We constructed Japanese age-sex-HR reference tables and examined uncertainty, time patterns, and age-based translation. Methods: Using the 23rd Complete Life Tables for Japan, 2020, we transformed annual death probabilities at index ages of 65-90 years with HRs of 1.00-4.00. Outcomes were median remaining lifespan; mean remaining years; 1-, 3-, 5-, and 10-year survival; and a baseline-equivalent age shift obtained by matching each transformed median to the HR 1.00 median curve. Results: At the age of 75 years, the median remaining lifespans at HRs of 1.00, 1.50, 2.00, 2.50, and 4.00 were 12.6, 10.1, 8.4, 7.3, and 5.1 years in men and 16.8, 14.3, 12.6, 11.3, and 8.8 years in women. An HR of 2.00 corresponded to baseline-equivalent age shifts of 5.8 and 4.8 years. A hypothetical HR interval of 1.50-2.50 translated to median ranges of 7.3-10.1 and 11.3-14.3 years. Relative to an HR of 2.00 throughout, a one-year front-loaded HR of 4.00 followed by an HR of 2.00 reduced median lifespan by 0.42 and 0.19 years but reduced 5-year survival by 3.8 and 1.9 percentage points. Conclusions: Life tables provide reproducible translations of HRs. Applying an external HR requires proportionality and transportability assumptions; the outputs are not validated patient-specific prognoses.