The mediated ��effects�� (ab) are assessed in all these 5000 datasets, and 95%CI are derived from the empirical distribution of these 5000 estimated ��effects��. We also quantified the magnitude of mediation <a href="http://www.selleckchem.com/products/gsk126.html">http://www.selleckchem.com/</a>
relative to the total ��effects�� in % by dividing the mediated ��effect�� by the overall association between adiposity measurement and serum C3 (i.e. ��ab/��c x 100). We also investigated whether all associations reported herein differed between sexes by adding interaction terms between the adiposity measure and sex to the (fully adjusted) models. We found no such effect modification and therefore all results are presented for men and women combined. A two-sided p-value of <?0.05 was considered statistically significant. All analyses were performed using the Statistical Package for Social Sciences for Windows, version 17.0 (SPSS Inc., Chicago, IL, USA). Characteristics of the study population across tertiles of serum C3 are shown in Table?1. Subjects in higher C3 tertiles <a href="http://www.selleckchem.com/products/Erlotinib-Hydrochloride.html">selleck
chemical</a> generally had increasing levels of body fat, inflammatory markers, and HOMA2-IR and a worse overall metabolic and cardiovascular risk profile. Pearson correlations of adiposity measures with each other, and with metabolic and inflammatory variables are shown in Table?2. All adiposity measurements, except the skinfold ratio, were significantly associated with serum C3 in linear regression models including age and sex (Table?3, paths cmodel1). After further adjustment for glucose metabolism status, prior CVD, eGFR, smoking status, alcohol use and medication, these associations were attenuated, <a href="http://en.wikipedia.org/wiki/Vasopressin_receptor">Vasopressin
Receptor</a> but remained statistically significant (paths cmodel 2). The sagittal diameter (��?=?0.412; 95%CI 0.333�C0.490) and the WC (��?=?0.383; 95%CI 0.302�C0.464) showed the strongest associations with C3. Mediation by low-grade inflammation and/or HOMA2-IR in the associations between all adiposity measures and C3 is shown in Table?3. For instance, the strength of the association between WC and C3 (��?=?0.383; 95%CI 0.302�C0.464), represented by path c in Fig.?1, was attenuated to ��?=?0.283 (95%CI 0.205-0.360) when further adjusted for low-grade inflammation, or to ��?=?0.175 (95%CI 0.085-0.265) when further adjusted for HOMA2-IR (Table?3; simple mediation analyses �C paths c��). The magnitude of the mediation in absolute values was ��?=?0.101 (95%CI 0.067-0.141) for low-grade inflammation and ��?=?0.208 (95%CI 0.151-0.268) for HOMA2-IR, explaining 26% and 54% of the association between WC and C3, respectively (paths ab). When both mediators were simultaneously added to the regression model, the association between WC and C3 was attenuated to 0.115 (95%CI 0.030-0.200) (Table?3; multiple mediation analyses �C path c��). The mediation by low-grade inflammation (��?=?0.090; 95%CI 0.060-0.126) and by HOMA2-IR (��?=?0.179; 95%CI 0.128-0.