Open AccessThe urine albumin-creatinine ratio is a predictor for incident long-term care in a general population
Author(s) -
Shuko Takahashi,
Fumitaka Tanaka,
Yuki Yonekura,
Kozo Tanno,
Masaki Ohsawa,
Kiyomi Sakata,
Makoto Koshiyama,
Akira Okayama,
Motoyuki Nakamura
Publication year - 2018
Publication title -
plos one
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.99
H-Index - 332
ISSN - 1932-6203
DOI - 10.1371/journal.pone.0195013
Subject(s) - medicine , hazard ratio , quartile , creatinine , proportional hazards model , biomarker , cohort , incidence (geometry) , renal function , natriuretic peptide , population , cohort study , confidence interval , heart failure , environmental health , biochemistry , physics , optics , chemistry
Background Several types of cardiovascular diseases (CVDs) impair the physical and mental status. The purpose of this study was to assess the predictive ability of several cardiovascular biomarkers for identifying the incidence of disability as future recipients of public long-term care (LTC) service. Methods The subjects of this study were community-dwelling elderly individuals ≥ 65 years of age without a history of CVD (n = 5,755; mean age, 71 years). The endpoint of this study was official certification as a recipient of LTC. The cohort was divided into quartiles (Qs) based on the levels of three CVD biomarkers: the urinary albumin-creatinine ratio (UACR), plasma B-type natriuretic peptide concentration (BNP), and serum high-sensitivity C-reactive protein concentration (hsCRP). A time-dependent Cox proportional hazard model was used to determine the multi-adjusted relative hazard ratios (HRs) for incident LTC among the quartiles of each biomarker. Results During the follow-up (mean 5.6 years), 710 subjects were authorized as recipients of LTC. The HR was only significantly increased in the higher Qs of UACR (Q3, p < 0.01; Q4, p < 0.001). However, other biomarkers were not significantly associated with the endpoint. The risk predictive performance for the incidence of LTC as evaluated by an essential model (i.e. age- and sex-adjusted) was significantly improved by incorporating the UACR (net reclassification improvement = 0.084, p < 0.01; integrated discrimination improvement = 0.0018, p < 0.01). Conclusions These results suggest that an increased UACR is useful for predicting physical and cognitive dysfunction in an elderly general population.