Change of self-rated physical health predicts mortality in aging individuals: results of a population-based cohort study

Abstract

Abstract Background Self-rated physical health (SRPH) is known as an important predictor of mortality. Previous studies mostly used baseline values of self-rated health to predict long-term mortality. The effect of change in self-rated physical health on mortality during the course of aging has rarely been researched. The present study aimed to determine SRPH over time in women and men of an aging population, assess whether and how change in SRPH affects mortality while adjusting for known determinants of mortality, and test effect modification by sex on the relation between course of SRPH and mortality. Methods Data of N = 12,423 respondents of the 5-year follow-up of the Gutenberg Health Study (GHS) with participation at the baseline assessment were analysed. All-cause mortality from 5-year follow-up onwards was defined as the primary outcome. SRPH was assessed by a single item. Cox proportional hazards models with adjustment for age, sex, socio-economic status and physical diseases were fitted to assess the predictive power of baseline score and course of SRPH. Additionally, effect modification by sex was assessed. Results During a median follow-up period of 7.3 years (quartiles 6.0-8.5 years), 618 (5%) participants died. Overall, 70.9% of the participants indicated good or very good SRPH at baseline (T1) and follow-up (T2), 6.9% rated their SRPH as not so good at T1 and T2, and 0.6% reported bad SRPH at T1 and T2. An improvement of SRPH was indicated by 9.6% and 12.0% indicated deterioration of their SRPH. Change in SRPH added substantial predictive information to the Cox proportional hazards models, when adjusting for relevant covariates. In men, deterioration and constantly bad SRPH were associated with the strongest increase in risk of mortality by 87%, resp. 228%. While improvements increased mortality risk in men (67%), women with an improved SRPH had a lower risk (57%). Conclusion A sizeable subgroup of aging participants reported deterioration of SRPH over five years. The association between change of SRPH and mortality is modified by sex. Deterioration of SRPH predicts mortality over baseline-assessment even when adjusted for relevant covariates. SRPH should be assessed regularly as part of an older individual’s health evaluation. Deterioration, constantly bad and improved SRPH should be taken seriously as unfavorable prognostic indicators, the latter only in men.