Negative and independent influence of apolipoprotein E on C-reactive protein (CRP) concentration in obese adults.: Potential antiinflammatory role of apoE in vivo

Background: Obesity is frequently associated with an increase in the early inflammation marker C-reactive protein (CRP), insulin resistance and changes in lipoprotein metabolism, Increased CRP is known as an independent cardiovascular risk factor. Since the apolipoproteins (apo) E and CIII components of HDL are associated with reduced cardiovascular risk and since apoE has in vitro anti-inflammatory effect, we have investigated the relationships between apoE, apoCIII (in apoB and non apoB containing lipoproteins) and CRP in obese adults. Methods: The following parameters from 34 healthy obese fasting women (age 22-64y, body mass index (BMI) 28-68 kg/m(2)) were measured: (1) ApoE and apoCIII, in total plasma, in apoB- (E LpB, CIII LpB) and non-apoB-containing lipoproteins (E LpnonB, CIII LpnonB); (2) CRP and cytokine secreted by adipose tissue (TNF-alpha and its soluble receptor TNFR2); (3) triglyceride, HDL-cholesterol, systolic blood pressure, diastolic blood pressure, waist and hip circumferences, insulin, glucose. HOMA, a marker of insulin sensitivity, and the ratio E/CIII in LpB and LpnonB were calculated. Results: CRP was positively correlated with BMI (P<0.05), waist circumference (WC, P<0.05), triglyceride (P<0.05) and negatively correlated with apoE (P<0.01) and E LpnonB (P<0.05). Two multiple regression models including parameters related to CRP with a P<0.25 were run stepwise to assess their independent contribution to CRP concentration. In the first model (including BMI, WC, HOMA, insulin, triglyceride, apoE, E LpnonB), apoE was the best predictor of CRP (P=0.01) together with triglyceride (P=0.02) and BMI (P=0.08), The second model took into account E/CIII LpnonB ratio with the parameters included in the first model. In this second model, E/CIII LpnonB was the best predictor of CRP (P=0.007), explaining 39% of CRP variance. Conclusion: ApoE is strongly correlated with CRP and could have an anti-inflammatory effect in vivo in obese subjects. This correlation could be limited to LpnonB lipoproteins, depending on their apoE and CIII relative content.