The significant interactions between both atherosclerosis severity measures and age in models predicting CCA diameter are consistent with age being a marker for length of exposure to the damaging risk factor effects. However, a significant interaction persisted only between age and plaques after adjustment for risk factors suggesting that much of the greater age effect in the pre-existing disease and high risk groups occur because of the effect of atherosclerotic damage evidenced by the presence of plaques.

The low risk subset was selected so that subjects had risk factor values lower than those typically used to identify cardiovascular risk. The reduction in the parameter estimate for age's impact on CCA diameter in the low risk group after adjusting for continuous atherosclerotic risk factors, indicates that even risk factor levels not commonly classified as detrimental can impact CCA diameters. This is not surprising given previous findings for BP ³³. Thus, at least part of the CCA diameter differences seen with older age in this low risk group was likely due to higher (though still normal) atherosclerosis risk factor levels at older ages. The persistence of a significant, positive association between age and CCA diameter in the low risk population, even after adjustment for atherosclerotic risk factors, suggests that damage or weakening of the arterial wall associated with aging also contributes to diameter enlargement. Since the primary goal of the ARIC ultrasound examination was not to identify plaques, we cannot exclude the possibility that some of the CCA diameter enlargement, even in the low risk group, was due to underlying atherosclerosis which was not identified.

This difference in diameters among the low risk group and higher risk groups has important implications for defining normal values in general. Since the prevalence of obesity and associated risk factors is high in many populations ^34-36, using population based samples could result in incorrect estimations of optimal values for CCA diameter and other risk factors influenced by obesity. Thus, persons at risk could be misclassified as being normal, and the possibility of misclassification could be largest in populations with the greatest obesity and risk.

Because the study is cross-sectional in design, the associations that were identified do not necessarily indicate a causal mechanism. The division into low risk, high risk and pre-existing disease groups is somewhat arbitrary; and so, results could reflect differences because of the ages represented within the risk groups. However, given that substantial proportions of all age categories 45–49, 50–54, 55–59, and 60–64 were represented within each risk group, age differences are unlikely to be driving the findings. Persons who had missing information leading to their exclusion in the current study were different in a number of ways from the persons available for study. However, since the associations for age and CCA diameter were quantitatively similar to those reported by for a smaller "normal" male population and qualitatively similar to those reported in a prospective study ^11,13, our results are likely to provide reasonable estimates for the differences that would be found in populations with disparate atherosclerotic risk.

Conclusion
In conclusion, consistent with our hypothesis and previous studies, our cross-sectional results indicate that persons with pre-existing atherosclerotic disease or atherosclerosis risk factors have larger diameters than persons without those attributes. More importantly, we also found that the impact of each year of older age on CCA diameter was significantly larger among populations with pre-existing cardiovascular disease and among persons with high risk factor levels or plaques compared to those at low risk of atherosclerotic disease and we were able to quantify the expected effect. Future investigations should consider the relationship between age, length of risk factor exposure, diameter, wall thickness, plaque, and disease development to better understand the progression of atherosclerosis and development of cardiovascular disease.

List of abbreviations
CCA: common carotid artery ARICLAD: Atherosclerosis Risk in Communities Limited Access Data CHD: coronary heart disease MI: myocardial infarction LDL: low density lipoprotein HDL: high density lipoprotein SBP: systolic blood pressure DBP: diastolic blood pressure NHLBI: National Heart, Lung, and Blood Institute