Dear Editor: We read with great interest the article by Tokgoz et al. [1] who obtained carotid intima–media thickness (CAIMT) in 6to 12-year-old epileptic children treated with valproate. In addition, they assessed plasma leptin, neuropeptide Y, ghrelin, and adiponectin levels of epileptic children treated with valproate and compared the data at the end of 6 and 12 months. In conclusion, they reported that plasma insulin, leptin, neuropeptide Y levels, and fasting insulin glucose ratio were increased, whereas plasma ghrelin and adiponectin levels, lipid profiles, and especially CAIMT did not change significantly at the end of 6 and 12 months. Dorsolateral aorta just before the bifurcation into the iliac arteries is the site of the earliest fatty streaks; the presence of these lesions is related to cardiovascular risk factors [2]. Moreover, involvement of the coronary and carotid arteries occurs later in life [2, 3]. Therefore, noninvasive measurement of the abdominal aortic wall thickness using highresolution ultrasonography is a better indicator for early atherosclerosis in children than CAIMT with low rates of missing data and reasonable reproducibility [3, 4]. We think that the results reported in this article should be confirmed by measurement of the abdominal aortic wall thickness. With an ever increasing number of studies for detecting development of subclinical atherosclerosis utilizing ultrasound evaluation of CAIMT as the primary endpoint, it is of vital interest to critically evaluate the validity of thesemeasurements [5]. In order to guide precise measurements of CAIMT, several studies reviewed the important methodologic aspects that exclude valid measurements of near-wall intima–media thickness, but permit reliable and valid measurements of far-wall intima–media thickness [4, 5]. The authors have not mentioned about this issue in section of materials and methods of their study. Moreover, some additional technical issues that deserve attention include a variety of acoustic settings inherent in the ultrasound imaging device, such as dynamic range, output and receiver gains, transducer frequency, etc. [6]. For example, for a fixed gain level, a 5-dB increase in dynamic range will result in an increased wall thickness of 0.003± 0.002 mm (p<0.001) [6]. We wonder that the authors took care of these parameters. It is recommended that for prospective clinical trials, one should analyze CAIMT measurements using semiautomated software systems. Because the older “on-screen” measurement methods appear dated and inadequate, due to limited quality-control issues [6], we think that it is especially important in children because of having thinner CAIMT. On the other hand, we had trouble understanding why the authors choose the left common carotid artery (CCA) for measuring. According to American Society of Echocardiography consensus statement, in measuring of CAIMT, ultrasound images of the distal 1 cm of the far wall of each CCA should be obtained and compared with values from a normative dataset [7]. Moreover, many well-known studies had measured bilateral CCA for evaluating effect of antiepileptic drug therapy to CAIMT [8, 9]. In conclusion, meticulous attention to quality control in image acquisition, measurement, interpretation, and reporting F. Ozkan (*) :M. F. Inci Department of Radiology Kahramanmaras, Kahramanmaras Sutcu Imam University, 46100 Yorukselim mah, Hastane Cad. No:32, Kahramanmaras, Turkey e-mail: drfozkan@yahoo.com