Table 1: Major hypercholesterolemia risk factors of the study population by race/ethnicity. 
 Variable  MA  OH  BNH  WNH  P value 
 Cholesterol checked (yes)  51.2 (1.6)^{a}  60.5 (2.2)^{b}  69.7 (1.3)^{c}  78.5 (0.9)^{d}  <.001 
 Inadequate nonHDLC (≥130 mg/dL)  25.8 (2.0)^{a}  22.1 (2.5)^{a}  33.1 (2.1)^{b}  30.6 (1.1)^{b}  .008  High cholesterol (≥200 mg/dL)  45.9 (1.4)^{a}  42.1 (2.2)^{a,b}  39.7 (1.3)^{b}  44.5 (0.9)^{a,b}  .019  Lipidlowering medications (yes)  50.7 (3.4)^{−}  44.1 (3.4)^{−}  51.7 (1.8)^{−}  54.7 (1.6)^{−}  .129  Age (years)  51.2 (1.2)^{a}  52.3 (1.1)^{a,b}  53.8 (0.5)^{b}  57.4 (0.3)^{c}  <.001  Education: <HS  51.5 (1.6)^{a}  37.7 (2.4)^{b}  26.1 (1.7)^{c}  13.8 (1.4)^{d}  <.001  No health insurance  52.2 (1.9)^{a}  36.3 (3.1)^{b}  26.0 (1.4)^{c}  13.7 (0.7)^{d}  <.001  Current smoker  11.4 (1.1)^{a}  14.5 (1.8)^{a}  21.7 (1.7)^{b}  19.1 (1.2)^{b}  <.001  BMI 30–34.9 (kg/m^{2}) obesity I  22.5 (1.6)^{a}  18.6 (1.4)^{b,c}  21.6 (1.3)^{a,c}  18.5 (0.6)^{b}  <.001  BMI ≥ 35(kg/m^{2}) obesity II  18.0 (0.8)^{a}  17.0 (1.3)^{a}  26.1 (1.2)^{b}  16.2 (0.6)^{a}  <.001  HTN (yes)  19.2 (1.4)^{a}  21.1 (1.6)^{a}  38.7 (1.3)^{b}  31.8 (1.0)^{c}  <.001  Diabetes (yes)  8.4 (0.8)^{a}  8.5 (0.8)^{a}  13.5 (0.8)^{b}  7.6 (0.6)^{c}  <.001  CHD (yes)  1.8 (0.3)^{a}  1.4 (0.3)^{a}  1.6 (0.2)^{a}  3.7 (0.2)^{b}  <.001 


MA: Mexican American; OH: other Hispanic, not Mexican; BNH: Black nonHispanic; WNH: White nonHispanic; nonHDLC: nonHDL cholesterol (total cholesterol minus HDLC); BMI: body mass index; HTN: hypertension; CHD: coronary heart disease. The data represent the weighted percent and standard error [% (SE)]. Significance is based on the adjusted F, a variant of the secondorder RaoScott adjusted chisquare statistic and its degrees of freedom. Each superscript letter denotes a subset of race/ethnicity categories whose column proportions do not differ significantly from each other at the 0.05 level for each row (variable). Significant differences between each group were measured by complex logistic regression analyses across race/ethnicity separately for each dependent binary variable. Age was tested by complex general linear models, recoding ethnicity to measure differences between groups.
