TY - JOUR
T1 - High-sensitivity C-reactive protein and all-cause mortality in four diverse populations
T2 - The CRONICAS Cohort Study
AU - Bernabe-Ortiz, Antonio
AU - Carrillo-Larco, Rodrigo M.
AU - Gilman, Robert H.
AU - Smeeth, Liam
AU - Checkley, William
AU - Miranda, J. Jaime
N1 - Publisher Copyright:
© 2021
PY - 2022/3
Y1 - 2022/3
N2 - Purpose: To assess the association between all-cause mortality and hs-CRP, based mainly on the cumulative burden approach. Methods: Cohort study with adults ≥35 years from general population, using hs-CRP at two timepoints: at baseline and 30 months later to establish different exposures: change over time, cumulative, and weighted cumulative hs-CRP. The outcome was all-cause mortality assessed 7 years later. Cox models were generated to quantify the association. Results: Data from 3,119 participants (mean age 55.6 years, and 51.2% females), were analyzed. During follow-up, 164 (5.6%) deaths occurred over 20,314.5 person-years, indicating an overall mortality rate of 8.1 per 1,000 person-years. In multivariable model, hs-CRP at baseline was associated with high risk of mortality (HR = 1.77; 95%CI: 1.28–2.46). Similarly, hs-CRP change over time (HR = 2.50; 95%CI: 1.46–4.29), as well as cumulative and weighted cumulative hs-CRP (HR = 2.05; 95%CI: 1.31–3.20) were associated with greater risk of all-cause mortality. The weighted cumulative hs-CRP had the best goodness-of-fit for mortality prediction. Conclusions: In this cohort across diverse geographical low-resource settings, high levels of hs-CRP were strongly associated with all-cause mortality. Two measurements of hs-CRP are better than one to predict mortality, and the weighted cumulative approach had the best prognostic fit.
AB - Purpose: To assess the association between all-cause mortality and hs-CRP, based mainly on the cumulative burden approach. Methods: Cohort study with adults ≥35 years from general population, using hs-CRP at two timepoints: at baseline and 30 months later to establish different exposures: change over time, cumulative, and weighted cumulative hs-CRP. The outcome was all-cause mortality assessed 7 years later. Cox models were generated to quantify the association. Results: Data from 3,119 participants (mean age 55.6 years, and 51.2% females), were analyzed. During follow-up, 164 (5.6%) deaths occurred over 20,314.5 person-years, indicating an overall mortality rate of 8.1 per 1,000 person-years. In multivariable model, hs-CRP at baseline was associated with high risk of mortality (HR = 1.77; 95%CI: 1.28–2.46). Similarly, hs-CRP change over time (HR = 2.50; 95%CI: 1.46–4.29), as well as cumulative and weighted cumulative hs-CRP (HR = 2.05; 95%CI: 1.31–3.20) were associated with greater risk of all-cause mortality. The weighted cumulative hs-CRP had the best goodness-of-fit for mortality prediction. Conclusions: In this cohort across diverse geographical low-resource settings, high levels of hs-CRP were strongly associated with all-cause mortality. Two measurements of hs-CRP are better than one to predict mortality, and the weighted cumulative approach had the best prognostic fit.
KW - C-reactive protein
KW - Cohort
KW - Inflammation
KW - Mortality
KW - Peru
UR - http://www.scopus.com/inward/record.url?scp=85122148042&partnerID=8YFLogxK
U2 - 10.1016/j.annepidem.2021.12.007
DO - 10.1016/j.annepidem.2021.12.007
M3 - Artículo
C2 - 34923118
AN - SCOPUS:85122148042
SN - 1047-2797
VL - 67
SP - 13
EP - 18
JO - Annals of Epidemiology
JF - Annals of Epidemiology
ER -