Multiple biomarker models for improved risk estimation of specific cardiovascular diseases related to metabolic syndrome: a cross-sectional study

Table 6 Adjusted ^a log-transformed joint association logistic regression base model results

CVD outcome (cases/n)	β x IQR	OR^b	(95% CI)	p-value
CHF (351/12,531)
Triglycerides	0.1654	1.180	0.961, 1.459	0.1141
CRP	0.6209	1.861	1.435, 2.432	<0.0001
Glycohemoglobin	0.1399	1.150	1.062, 1.246	0.0006
*Joint association*	0.9262	2.525	1.856, 3.437	<0.0001
Angina (377/12,520)
Triglycerides	0.2119	1.236	1.016, 1.504	0.0342
CRP	-0.0082	0.992	0.742, 1.326	0.9560
Glycohemoglobin	0.0974	1.102	1.021, 1.190	0.0123
*Joint association*	0.3012	1.352	1.006, 1.816	0.0456
MI (549/12,557)
Triglycerides	0.0834	1.087	0.923, 1.280	0.3164
CRP	0.0478	1.049	0.822, 1.339	0.7005
Glycohemoglobin	0.1304	1.139	1.058, 1.227	0.0006
*Joint association*	0.2617	1.299	0.997, 1.693	0.0527
Stroke (418/12,560)
Triglycerides	0.0686	1.071	0.888, 1.292	0.4743
CRP	0.3042	1.356	1.072, 1.715	0.0112
Glycohemoglobin	0.0812	1.085	1.010, 1.164	0.0248
*Joint association*	0.4538	1.575	1.194, 2.077	0.0013
CHD (508/12,518)
Triglycerides	0.1082	1.114	0.965, 1.287	0.1419
CRP	-0.0404	0.960	0.719, 1.284	0.7851
Glycohemoglobin	0.1560	1.169	1.092, 1.251	<0.0001
*Joint association*	0.2238	1.251	0.918, 1.705	0.1565

^aModels adjusted for age, gender, race/ethnicity, BMI, smoking status, and family income.
^bOR estimates calculated for an IQR change in the respective biomarker, or an IQR in each biomarker for the joint effect estimates. IQRs for ln(Triglycerides, mg/dL) = 0.798 [corresponding to a 2.22 fold increase in non-transformed Triglycerides], ln(CRP, mg/dL) = 2.22 [corresponding to a 9.21 fold increase in non-transformed CRP], ln(Glycohemoglobin, %) = 0.0935 [corresponding to a 1.10 fold increase in non-transformed glycohemoglobin].

ISSN: 1478-7954