We found in this sample of asymptomatic older adults ≥70 years that low subclinical vascular disease by the combination index (lowest quartiles of CAC and cIMT with a normal ABI) was independently associated with traditional risk factors of younger age, female gender, lower BMI, absence of hypertension and dyslipidemia, and never smoking. This confirms prior studies that low levels of multiple traditional risk factors contribute to successful cardiovascular aging. The Framingham Heart Study showed that traditional markers of low-risk status (fewer or no cigarettes smoked, lower BP, favorable lipid profile, absence of a FH of premature CHD, absence of diabetes) were associated with survival beyond 75 years.19
Non-white race was significantly associated with low CAC and cIMT with a non-significant trend for the combination index, whereas non-white race was inversely associated with normal ABI. Other MESA analyses have confirmed these racial/ethnic disparities with African-Americans having less CAC but also lower ABIs that are not explained by differences in coronary risk factors.20,21
We did not find any independent associations of diet, physical activity, or the several novel risk factors that we explored with subclinical vascular health after adjusting for traditional risk factors. However, other studies have shown an association between healthy aging and lifestyle factors such as the Mediterranean diet22
, physical activity23
, higher education6
, and modest wine consumption6
. The lack of association of healthy diet and physical activity with low subclinical disease in our study may be because the lifestyle practices of these individuals at age ≥70 years may not reflect their usual diet pattern or physical activity throughout their youth and middle ages. Those with more adverse subclinical profiles may have altered their diet and physical activity in response to knowledge of elevated risk factors.
In addition, studies have found associations between CRP, IL-6, homocysteine, and fibrinogen with CVD events.24,25
However similar to our study, other studies have not found any association of CRP with subclinical vascular disease.26,27
We found that homocysteine and IL-6 were associated with a decreased odds of the healthy combination index in unadjusted analyses, but we did not find any independent association of these markers with vascular health after covariate adjustment. Perhaps factors associated with subclinical disease, especially cross-sectionally, are different from those factors predicting clinical events.
It has been accepted for some time that vascular age is not the same as chronological age (“A man is only as old as his arteries” said Dr. Thomas Sydenham in 1689), and our study reinforces this message. The use of a combination index to identify the absence of subclinical vascular disease in multiple vascular territories is a strength of the present analysis and has precedent from the CHS cohort, which used an index based on ABI, carotid artery stenosis, cIMT, electrocardiographic and echocardiographic abnormalities, and angina;18
however, CAC was not part of their combination index. Because CAC is strongly associated with aging, low CAC scores in older individuals are uncommon, particularly for men,28
but are associated with low CVD event rates.29,30
The close association between CAC and vascular age has been applied to the Framingham Risk Score. It has been proposed that the amount of plaque burden measured by CAC scoring be used to modify the points assigned to chronological age when calculating 10-year CVD risk, so that older individuals with low CAC could potentially be assigned a lower chronological age in risk-prediction models.31
The Framingham Risk Score has been shown to be an imperfect tool in predicting subclinical atherosclerosis risk across multiple arterial beds.32
Although certain factors such as younger age, absence of dyslipidemia, and never smoking were consistently found to be associated with each separate parameter of healthy vasculature in addition to the combination index, there were some differences in the associations between some of the other risk factors and the individual measures of subclinical disease. For example lower BMI was inversely associated with normal ABI, which may reflect residual confounding such as with smoking quantity. Also, the absence of diabetes was not found to have a statistically significant association with low CAC or the combination index, despite its association with cIMT and ABI, although adjusting for CVD risk factors in the multivariable models had slightly attenuated any association with diabetes.
Additionally, a negative FH of MI is strongly associated with the lack of significant CAC even after adjusting for traditional risk factors, and supports the notion that genetic factors play a role as supported by other studies. However, we did not find lack of FH to be associated with low cIMT, although there was a non-significant trend for association of FH with normal ABI and the combination index in the multivariable analysis. This suggests that some factors leading to the development and progression of CAC may be different than those for cIMT. The MESA Family Study is evaluating genetic predictors of subclinical vascular disease, and differences in associations between risk factors and the various measures of subclinical vascular disease will be explored further in future MESA papers using the entire cohort. A recent study from MESA found that while CAC was a stronger predictor for MI and overall CVD events compared to cIMT, cIMT was marginally better at predicting stroke than CAC.33
Thus the predictive ability of subclinical atherosclerosis imaging for CVD events varies by the different vascular beds, and it remains to be seen whether a composite index of multiple vascular beds would be better than single bioimaging tests.
The conclusions about relationships drawn here are probably influenced by several aspects of our study design. All MESA participants were free of clinical CVD; the range of differences between the comparison group and those who aged successfully was only in subclinical measures, not clinical disease. Thus we studied healthy vascular aging in the context of generally fairly healthy aging (i.e. absence of clinical disease). The risk factors studied here might well have distinguished older MESA participants from those excluded because of clinical CVD. It is possible that some risk factors play a more important role for disease development in a younger middle-age cohort but are relatively less important in this cohort surviving to ≥70 years at entry into study free of clinical CVD at baseline. If our comparison group were elderly subjects in the general U.S. population, we would expect to see a much stronger association of the absence of the major CVD risk factors with low subclinical vascular disease.
This study is limited by the fact that these measurements were made cross-sectionally. We do not have information about risk factors in these participants measured earlier in life. Residual confounding, measurement imprecision, and survival bias may also be plausible explanation for some of the results. There may have been real but small effects that we were not able to detect statistically due to limitations of sample size. However, most of our findings were predicted by our a priori hypotheses and the biological bases have been confirmed in other studies.
This study is unique in its evaluation of an asymptomatic multi-ethnic group of older individuals who have not only survived to ≥70 years and are free of clinical CVD, but who also remained free of significant subclinical vascular disease. In summary, by evaluating factors associated with low subclinical vascular disease burden in healthy asymptomatic older adults, our findings are consistent with results from previous longevity cohort studies demonstrating lower levels of multiple traditional risk factors, but perhaps not several novel risk factors, are associated with successful vascular aging. Follow-up of the MESA cohort for future CVD events will allow us to determine which measures of subclinical vascular disease are the best predictors of cardiovascular health in older adults.