Dear Editor,

Recently Azemati and colleagues [1] reported the associations between junk food consumption and cardiometabolic outcomes in a large cross-sectional dataset of Iranian children and adolescents. This is one of the largest studies in such regions and extends on the current evidence of poor diet in countries where obesity is increasingly becoming an issue. However, we have identified gaps and it would be important to consider these when developing an evidence base. Firstly, the authors defined the metabolic syndrome (MetS) according to the Adult Treatment Panel III criteria modified for the pediatric population. To date, there is no clear consensus as to whether MetS should be defined in children and adolescents; and considering the progressive nature of the condition, controversy surrounds the dichotomization of continuous cardiometabolic variables, as such an approach results in a loss of data and statistical power and can underestimate future disease risk [2, 3]. Consequently, it is recommended that research in pediatric populations focus on the clustering of continuous cardiometabolic variables [2] and this should be considered by the authors when interpreting their findings, as although they conclude that junk food intake was not related to MetS, it is indeed possible that junk food intake in this age group contributes to the progression of MetS in later life.

Additionally, the authors studied an age group with great variation in sexual maturation; however, data regarding participant’s pubertal stage was not collected nor adjusted for in the statistical analyses. This may have influenced the authors’ findings, as during puberty, children develop transient insulin resistance—a known determinant of MetS [3]. Finally, there are limitations inherent in using BMI as a measure of obesity and MetS risk, as individuals with a high BMI can present without metabolic abnormalities, particularly when they have little abdominal visceral adipose tissue (VAT) [4]. As such, VAT is thought to drive the cardiometabolic health risks associated with obesity [4] and so we argue that VAT should be assessed in Iranian children as a more reliable marker of MetS risk. Although authors did assess WHtR, this measure is not enough to quantify VAT and as discussed above, this measure was not adjusted for pubertal stage and was dichotomised, thereby reducing its predictive reliability. In summary, whilst Azemati and colleagues (2020) do provide insights into the role of junk food on metabolic syndrome in countries with changing dietary patterns, such studies should assess cardiometabolic risk factors as continuous variables and need to carefully consider puberty and VAT to provide the most accurate and reliable estimates.