IJCS | Volume 33, Nº1, January / February 2019

62 Despite the great potential of the method, the use of CCTA for the establishment of a correlation between CAD and obesity is still little explored. Compared with catheterization, computed tomography angiography is a highly accurate, non-invasive method, with acceptable levels of patient radiation and contrast, that can be useful in the identification of coronary arterial narrowing by atherosclerotic plaques. Although the association of obesity with CAD is well documented, 15,16 there is evidence supporting that cardiovascular risk factors are not more common inMHO individuals compared with non-obese subjects. 17-19 In other words, obesity alone would not be determinant for increased incidence of CAD. This is corroborated by our findings on the prevalence of obstructive coronary disease, which was not different between obese and non-obese subjects. On the other hand, the higher values of coronary artery calcium score among obese individuals suggest a correlation between this condition with the development of subclinical atherosclerosis. Chang et al., 20 demonstrated that MHO patients have higher calcium score values than non-obese patients. However, after adjusting for metabolic risk factors, this association was attenuated and no longer statistically significant. The authors concluded that obesity is an additional risk for coronary atherosclerosis, including the subclinical form, mediated by metabolic changes whose thresholds are lower than those considered abnormal. In this context, one important factor is the influence of BMI on tomography imaging analysis. Obese individuals show a reduced signal-to-noise ratio in chest images, due to increased adipose tissue compared with non-obese individuals. The higher chest wall thickness in obese subjects attenuates the X-rays emitted from the tubes, allowing that a lower amount of photons reaches the detector for image construction, resulting in a more “grained” image. Such loss could be compensated by modulations in the tube voltage and in the X-ray tube current, improving the signal-to-noise ratio of these tests. However, the methods used for image acquisition for coronary artery calcium scoring do not allow adjustments in tube voltage of the tomography scanner, fixing it at 120 kilovolts. In practical terms, that implicates that images with lower signal-to-noise ratio are obtained from CCTA in obese patients. In parallel with the potential effect of obesity on coronary calcification, we believe that this change in the signal-to- noise ratio in obese patients image may have contributed to changes in the threshold for coronary artery calcium detection, artificially increasing calcium score levels in this population. In this regard, mean calcium score percentile in our patients was 61 according to the Multi- Ethnic Study of Atherosclerosis (MESA), 21 indicating a higher-than-average coronary calcification. However, these results are not comparable with those reported in the MESA study, which evaluated asymptomatic individuals, with not history of CAD, due to selection bias of our study population (patients referred for coronary tomography for investigation of CAD and hence more likely to have the disease). Limitations Our study has limitations inherent to its retrospective design. Since this was a cross-sectional study evaluating the association of obesity with CAD based on medical records, the results do not take into account some variables, such as the time of exposure to triggering factors of the disease. The definition of metabolically healthy obesity was based on the identification and exclusion of obesity-related metabolic abnormalities (hypertension, dyslipidemia, diabetes). Nevertheless, laboratory markers of insulin resistance, including the Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) were not used in patients recruitment in our study. Our study population was selected based on BMI, which, although is the most widely used anthropometric variable to characterize obesity, it does not provide information regarding body composition. Therefore, assuming that the percentage of body fat has a direct effect on insulin resistance, BMI alone does not give us any insight into this condition. In addition, other anthropometric measures known to provide a more accurate estimation of visceral fat (e.g. waist circumference and waist-to-hip ratio measurements) were not registered in the medical records, and hence could not be used in the analysis. Finally, the definition of CAD by CCTA may be controversial; although CCTA is a very robust method to define non-obstructive atherosclerosis by coronary artery calcium scoring, the method considered the gold- standard method to detect obstructive coronary disease is invasive coronary angiography combined or not with intracoronary ultrasound. Pereira et al. Obesity and coronary artery disease Int J Cardiovasc Sci. 2020;33(1):57-64 Original Article