Interactions between C-reactive protein and traditional risk factors in predicting mortality of older adults

Abstract

Abstract. Background: Elevated levels of C-reactive protein (CRP) are known to be associated with cardiovascular (CV) morbidity and mortality in older adults, however, there seems to be heterogeneity of this association across subsets of individuals. We aim to assess the effects of interactions between CRP and one of the following traditional CV risk factors regarding all-cause mortality in unselected elderly men and women: age, sex, body mass index, diabetes, and hypertension. Patients and methods: Three hundred and forty-four general practitioners all over Germany enrolled 6,817 unselected participants, aged 65 years or older, and performed thorough examinations, including CRP measurement at baseline (getABI study). All-cause mortality was determined in the following seven years. Cox regression analyses were done using uni- and multivariable models. Results: At baseline 4,172 participants of this cohort had a CRP value of ≤ 3 mg/L (low level CRP group), 2,645 participants had a CRP value of > 3 mg/L (high level CRP group). The unadjusted hazard ratio for all-cause death of the high level CRP group compared to the low level CRP group was 1.49 (95 % confidence interval [95 %CI] 1.34 to 1.66). After adjustment for sex, age, education, peripheral artery disease/media sclerosis, other prior vascular events, smoking status, diabetes, systolic blood pressure, antihypertensive medication, body mass index, cholesterol, and statin use, the hazard ratio was 1.34 (95 %CI 1.20 to 1.50). Significant interactions with CRP were found for sex (adjusted hazard ratio 1.38, 95 %CI 1.11 to 1.72), age (0.75, 95 %CI 0.60 to 0.94), and baseline systolic blood pressure (0.64, 95 % CI 0.51 to 0.81). The interactions of CRP with body mass index and of CRP with diabetes were not significant. Conclusions: In older German adults, there seem to be effect modifications by age, sex, and arterial hypertension regarding the effect of CRP in the prediction of all-cause mortality.