TY - JOUR
T1 - CRP level and HDL cholesterol concentration jointly predict mortality in a Korean population
AU - Kim, Kwang Il
AU - Oh, Se Won
AU - Ahn, Soyeon
AU - Heo, Nam Ju
AU - Kim, Sejoong
AU - Chin, Ho Jun
AU - Na, Ki Young
AU - Chae, Dong Wan
AU - Kim, Cheol Ho
AU - Kim, Suhnggwon
PY - 2012/8
Y1 - 2012/8
N2 - Background: C-reactive protein (CRP) and high-density lipoprotein (HDL) cholesterol are well-known cardiovascular predictors. However, the joint effect of these parameters on long-term mortality has not been established. Methods: We studied a total of 92,500 subjects older than 20 years who underwent routine health examination at the three health care centers affiliated with Seoul National University. High-sensitivity CRP and the lipid profile were obtained at baseline. Subjects were followed for a median of 45.5 months. Mortality data were obtained from the National Statistics Office of Korea. Results: There were 649 deaths (0.7%) during the follow-up. The leading cause of death was cancer. The subjects who died were significantly older, had a male predominance, and had increased levels of inflammatory markers. A significant mortality difference was identified according to the CRP and HDL cholesterol levels. Considering both parameters jointly, subjects with a CRP <1.4 mg/L (highest quartile) and HDL cholesterol <45 mg/dL (lowest quartile) were at the highest risk for all-cause mortality, even after adjusting for covariates (hazard ratio 2.29, 95% confidence interval, 1.83∼2.87). After matching on the propensity score, 6304 subjects with a high CRP and low HDL cholesterol were at high risk of death (hazard ratio 2.52, 95% confidence interval, 1.59∼4.01). Interestingly, the joint effect of CRP and HDL cholesterol was observed for cardiovascular as well as cancer-related mortality prediction. Conclusions: Elevated CRP and low HDL cholesterol jointly contribute to the prediction of all-cause, cancer, and cardiovascular mortality in Koreans. The interactive relationship between them in mediating inflammatory processes might explain these results.
AB - Background: C-reactive protein (CRP) and high-density lipoprotein (HDL) cholesterol are well-known cardiovascular predictors. However, the joint effect of these parameters on long-term mortality has not been established. Methods: We studied a total of 92,500 subjects older than 20 years who underwent routine health examination at the three health care centers affiliated with Seoul National University. High-sensitivity CRP and the lipid profile were obtained at baseline. Subjects were followed for a median of 45.5 months. Mortality data were obtained from the National Statistics Office of Korea. Results: There were 649 deaths (0.7%) during the follow-up. The leading cause of death was cancer. The subjects who died were significantly older, had a male predominance, and had increased levels of inflammatory markers. A significant mortality difference was identified according to the CRP and HDL cholesterol levels. Considering both parameters jointly, subjects with a CRP <1.4 mg/L (highest quartile) and HDL cholesterol <45 mg/dL (lowest quartile) were at the highest risk for all-cause mortality, even after adjusting for covariates (hazard ratio 2.29, 95% confidence interval, 1.83∼2.87). After matching on the propensity score, 6304 subjects with a high CRP and low HDL cholesterol were at high risk of death (hazard ratio 2.52, 95% confidence interval, 1.59∼4.01). Interestingly, the joint effect of CRP and HDL cholesterol was observed for cardiovascular as well as cancer-related mortality prediction. Conclusions: Elevated CRP and low HDL cholesterol jointly contribute to the prediction of all-cause, cancer, and cardiovascular mortality in Koreans. The interactive relationship between them in mediating inflammatory processes might explain these results.
KW - C-reactive protein
KW - HDL cholesterol
KW - mortality
UR - http://www.scopus.com/inward/record.url?scp=84864274510&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84864274510&partnerID=8YFLogxK
U2 - 10.1016/j.amjmed.2012.02.013
DO - 10.1016/j.amjmed.2012.02.013
M3 - Article
C2 - 22840665
AN - SCOPUS:84864274510
SN - 0002-9343
VL - 125
SP - 787-795.e4
JO - American Journal of Medicine
JF - American Journal of Medicine
IS - 8
ER -