State-Level Variation in Cardiovascular Inflammatory Biomarkers, Risk Factors, and Clinical Risk Scores.

Source: Women's Health Study. Abbreviations: hsCRP(Log) (natural log of high sensitivity C-reactive protein);
sICAM-1 (soluble intercelluar adhesion molecule); HDL-C (high-density lipoprotein cholesterol); LDL-C (low-density lipoprotein cholesterol); BMI(Log) (natural log of body mass index); ICC (intra-class correlation).
Multilevel linear regression random effect estimates and intraclass correlations estimate the presence non-zero variance in inflammatory biomarkers, cholesterol, BMI (log), or risk scores at the state level. P value ≤ 0.05 rejects null hypothesis of zero variance among states.
*Estimates presented as 1st to 99th percentile range.
†Models of cholesterol as the outcome variable do not control for cholesterol measures within the model. Models for BMI, Framingham and Reynolds risk scores are not adjusted for covariates.
Model I. Unadjusted.
Model II. Age-adjusted.
Model III. Adjusted for age, race, obesity/overweight, diabetes, systolic blood pressure, HDL-C†, LDL-C†
Model IV. Adjusted for age, race, obesity/overweight, diabetes, systolic blood pressure, HDL-C†, LDL-C†, exercise,
smoking status, daily caloric intake.