Table 7.

Multiple regression models of serum triglycerides predicting changes in peripheral blood leukocyte counts between men and women in NHANES 1999–2004.

		Men				Women
Cell Count	Model	β	SE	% Δ	p	β	SE	% Δ	p
Lymphocytes	unadjusted	0.097	0.012	0.9	<0.0001	0.135	0.014	1.3	<0.0001
	adjusted	0.112	0.013	1.1	<0.0001	0.145	0.015	1.4	<0.0001
Monocytes	unadjusted	0.025	0.013	0.2	0.06	0.100	0.017	1.0	<0.0001
	adjusted	0.000	0.014	0.0	1.00	0.067	0.019	0.6	<0.001
Neutrophils	unadjusted	0.123	0.018	1.2	<0.0001	0.161	0.017	1.6	<0.0001
	adjusted	0.075	0.020	0.7	<0.001	0.124	0.020	1.2	<0.0001
Eosinophils	unadjusted	0.016	0.006	0.2	<0.05	0.012	0.005	0.1	<0.05
	adjusted	0.010	0.006	0.1	0.10	0.006	0.005	0.1	0.21
Basophils	unadjusted	0.010	0.002	0.1	<0.0001	0.013	0.002	0.1	<0.0001
	adjusted	0.008	0.002	0.1	<0.01	0.011	0.002	0.1	<0.0001

β: slope; SE: standard error; p: p-value. % Δ = the predicted % change in blood leukocyte count (10³ cells/μL) for a 10% increase in LDL-C level (mg/dL). Adjusted model is adjusted for age, serum cotinine, BMI, waist circumference, race/ethnicity, education, statin use, and survey cycle. To best approximate linearity, both predictor and outcome were log transformed (base e); thus slope and standard errors are on log scale.