This study showed no significant associations of total screen time and TV time with cardiometabolic risk in adolescents with overweight and obesity. Our findings are in agreement with recent findings among overweight and obese adolescents 
, but in contrast with previous findings among obese children 
and among normal weight children and/or adolescents 
The lack of an association of screen time and TV time with cardiometabolic risk in the present study may be attributed to several factors. First, the potential association between screen time and cardiometabolic risk factors could be masked by underreporting of screen time. Obese adolescents may have underreported their actual screen time, as also has been observed for food intake 
. Similarly, in the study of Slootmaker et al 
it was observed that overweight adolescents overrated their physical activity level. Second, mainly obese adolescents participated in the present study. Although variances of metabolic values were comparable to previous studies 
, the relatively high metabolic values could have masked a potential association between screen time and cardiometabolic risk. It is unclear why Pardee et al 
found, in a comparable cross-sectional study and among a comparable group of children, a positive association between self-reported TV time and hypertension, whereas we found no association between TV time systolic and diastolic blood pressure.
Our finding of a small but significant negative association between computer time and blood levels of TC and LDL-C is in contrast with the recent cross-sectional study of Goldfield 
. This study among 282 overweight and obese adolescents aged 14–18 years old found that self-reported time spent playing seated video games, but not recreational computer time, was negatively associated with cardiometabolic risk factors. Unfortunately, we could not differentiate between time spent playing video games and other computer activities, thereby limiting a detailed comparison of our findings with the findings of Goldfield 
In a study among 7–10 years old boys, Wang & Perry demonstrated that seated video gaming resulted in various physiological and metabolic responses, indicating that video gaming is not an entirely passive activity in children. Although our measure of computer time was broader than video gaming only, similar to Wang & Perry 
our results could suggest that computer use may have different cardiometabolic effects than TV viewing.
Limitations of this study include the self-reported screen time and MVPA using the AQuAA. Among normal weight adolescents, the reliability of the AQuAA was fair to moderate, but this questionnaire – similar to other self-report measures – considerably underestimates sedentary time and overestimates time spent on physical activity 
. Unfortunately, we currently do not have a gold standard for the measurement of screen time available. Accelerometry is increasingly used as an objectively measure of sedentary behaviour; however, it cannot distinguish between TV time and computer time. In addition, the cross-sectional design limits the ability to conclude on causality within the associations found. Finally, we did not have data on energy intake, cigarette smoking and alcohol consumption, which may have attenuated the associations. Strengths include the objective measures of all outcome variables, including trunk and total fat mass (determined using DEXA) and indicators of glucose and lipid metabolism (determined using fasting blood samples). In addition to fasting blood samples, this study also included an OGTT.