In this study, we observed a significant relationship between AIP and CAC as well as the progression of CAC over 4-year period in Korean adults without CVD. These findings are consistent with previous studies that showed strong associations between AIP and cardiovascular risk factors and CVD. Furthermore, to the best of our knowledge, this study is the first study to reveal a longitudinal association between AIP and CAC progression.
When we categorized the subjects into tertiles according to AIP, those at the highest tertile had the highest BP, BMI, CACS, and adverse lipid profiles. Also, the incidences of diabetes, hypertension, alcohol drinking, and smoking were highest in this group. In line with our results, prior studies demonstrated that AIP is an independent predictor of CAD among Chinese subjects, Chinese postmenopausal women, and very young adults[9, 10, 16]. Another recent study revealed that AIP predicts a plaque burden in intermediate CVD risk patients presented with chest pain[17]. In addition, AIP was associated with various metabolic disorders including fatty liver disease, hypertension, diabetes, and diabetic complications[18, 19]. While most of these studies were conducted on subjects with overt CAD or chest pain, or those diagnosed with diabetes, which is considered CAD equivalent, our study excluded people with CAD or cerebrovascular disease as well those on lipid lowering therapy. Despite the relatively low cardiovascular risk of our study population, the higher AIP was still associated with the higher CACS.
Furthermore, the higher AIP was associated with the progression of CAC. While 17.6% of subjects in the lowest AIP tertile showed the CAC progression, 32.5% of subjects in the highest tertile showed the progression. Moreover, the Δ √transformed CACS and the annual Δ √transformed CACS increased gradually across the tertile, indicating that the baseline AIP predicts the progression of CAD and possibly future coronary events.
However, after adjusting for various conventional cardiovascular risk factors such as blood pressure, glucose, LDL-C, exercise, alcohol, smoking, BMI, and the presence of hypertension and diabetes, the predictive value of AIP on CAC progression lost the significance. There are several possible reasons for this. First, a near normal, narrow range of lipid parameters of our study subjects may have attributed. Even the subjects in the highest tertile AIP had the mean TG level of 2.17 mmol/L and mean HDL-C level of 1.06 mmol/L. Previous studies show that a high TG and low HDL-C are closely related with CAC, even more than LDL-C[20, 21]. If the range of AIP in our study was wider with more extreme values, it may have resulted in a significant relationship. Also, while AIP is an independent risk factor for CAD in a cross section setting[9, 10], other factors may be more critical to the CAD progression. For example, insulin resistance is an important risk factor for atherosclerosis[22], and our previous study also demonstrated that triglyceride-glucose (TyG) index, which is a surrogate marker of insulin resistance, is an independent predictor of CAC progression[23].
Although AIP did not predict the progression of CAC, it does not imply that it is not a good predictor of CVD. Recently, there has been a controversy over the prognostic value of the repeated measure of CAC in predicting CVD[24]. While earlier studies suggested the additive contribution of changes in CAC in CV risk prediction, other studies showed that CAC change was only the fifth strongest risk marker for CHD, following baseline CAC, gender, SBP, and total cholesterol[25]. Also, MESA demonstrated that CAC change of a greater than > 100 U/y was associated with coronary heart disease independent of risk factors and baseline CAC score[25]. In other words, it may be likely that although AIP was not able to independently predict the progression of AIP, it does not mean that is a good predictive marker of future CVD, and that it may have a synergistic role with the baseline CACS.
There are several limitations in this study. Since it was a retrospective, longitudinal study, not all of the potential confounding factors were controlled. For example, medications such as antiplatelet agent, anti-diabetic, anti-hypertensive drugs that could affect the progression of atherosclerosis as well as diet, exercise, smoking and alcohol consumption patterns were not controlled or monitored during the follow-up period. Also, the follow-up period was variable. Secondly, our study results cannot be generalized. Not only because people with existing CAD were excluded, but those with no CVD risk are unlikely to be included in the study. We only included subjects who voluntarily took repeated coronary CT scans for a health check-up, and thus there may be a selection bias. Thirdly, we applied the same definition of CAC progression as our previous paper[23], but there is no consensus on the optimal way to quantify CAC change.
In spite of above limitations, this study has significant implications that are clinically relevant, as it is the first to investigate the association between AIP and CAC progression.