Table 3.

Correlates of the Carotid and Aortic AIs

	Carotid AI			Aortic AI
Correlates	β ± SE	P	r2	β ± SE	P	r2
Longitudinal analysis
Being female	11.3 ± 2.2	<0.0001	10.3	14.9 ± 2.4	<0.0001	10.8
Age (+15 years)	12.4 ± 0.96	<0.0001	28.9	13.8 ± 1.14	<0.0001	29.6
Body height (+7 cm)	−3.22 ± 0.84	<0.0001	1.49	−1.75 ± 0.91	0.03	0.41
Mean arterial pressure (+10 mm Hg)	3.41 ± 0.79	0.0001	1.42	2.28 ± 0.87	0.01	0.66
Smoking at baseline	6.80 ± 1.76	0.0001	1.39	8.85 ± 1.91	<0.0001	2.14
Cross-sectional analysis
Being female	10.2 ± 2.2	<0.0001	2.23	13.0 ± 2.4	<0.0001	10.4
Age (+15 years)	10.9 ± 0.90	<0.0001	33.2	13.7 ± 0.97	<0.0001	34.8
Body height (+7 cm)	−3.85 ± 0.84	<0.0001	10.3	−2.45 ± 0.91	0.0053	0.64
Mean arterial pressure (+10 mm Hg)	5.59 ± 0.76	<0.0001	3.56	3.23 ± 0.84	<0.0001	1.46
Total cholesterol (+1 mmol/L)	…	…	…	2.54 ± 0.86	0.002	0.77
Smoking at follow-up	5.22 ± 1.69	0.002	0.79	8.23 ± 1.85	<0.0001	1.47

The AIs were standardized to a heart rate of 75 beats/min. The covariables considered for entry into the multiple regression model were sex, age, body height, mean arterial pressure, total cholesterol, current smoking and drinking, and antihypertensive drug treatment. β ± SE, P, and r² indicate the effect size ±SE, statistical significance, and the percentage variance of the dependent variable explained by a single covariable, respectively. Effect sizes are expressed for a ~1 SD higher value of the explanatory variables. The percentage of variance explained by the whole model was 50.1% and 49.5% for the carotid and aortic AIs in the cross-sectional analysis, respectively, and 43.5% and 43.6% in the longitudinal analysis, respectively. Abbreviation: AI, augmentation index.