Cardiovascular and metabolic diseases are the principal causes of mortality in developed countries [¹]. Therefore, cardiovascular diseases (CVD) and metabolic risk factors are very important concerns in public health policies [²]. To prevent atherosclerosis, type 2 diabetes, and the metabolic syndrome, it is necessary to develop actions from childhood and adolescence because CVD risk factors track towards adulthood [³]. Likewise, the increasing prevalence of CVD risk factors in children and adolescents might be, at least in part, a response to the worldwide trends in pediatric overweight and obesity [⁴,⁵].

Adequate habitual physical activity and a healthy diet are the basic cornerstones to prevent obesity and CVD [⁶]. However, chronic diseases result from the interaction among many factors [⁷], and the evaluation of other lifestyles could be relevant. In the last decades, the development of technologies has been related to sedentary behavior and obesity in children and adolescents [⁸]. Nowadays, there is a social trend to mistakenly consider physical activity and sedentary behaviors as opposite poles from the same continuous (e.g. if a child does not achieve enough physical activity, he or she may be categorized as sedentary or couch potato), but studies in adolescents have shown that time spent in sedentary behaviors must be considered independently of physical activity [⁹], and hence, both sedentary behavior and physical activity may have a different effect in the prevention and development of CVD.

Television (TV) viewing is a well-known sedentary behavior and too much time spent watching TV is considered an unhealthy behavior associated with obesity and other harmful health outcomes in youth [¹⁰]. Little is known about how TV viewing and CVD risk factors are linked in youth because the majority of the studies have been conducted in adults [^11-14]. To our knowledge, in children and adolescents, several studies have analyzed the associations between TV viewing and individual CVD risk factors-abdominal fat [¹⁵,¹⁶], lipid abnormalities [¹⁷,¹⁸], hypertension [¹⁹,²⁰], insulin resistance [²¹]- and only two studies have examined the associations using composite CVD risk scores [²²,²³]. The majority of the above mentioned studies did not show whether these associations were independent of body fat.

On the other hand, interventional studies reducing TV viewing in apparently healthy children and adolescents have shown small or no significant effects on obesity [⁸,²⁴], whereas interventions on obese children by decreasing time spent watching TV have shown a positive effect on obesity indicators [²⁵,²⁶]. Hence, these findings suggest that the effect of TV viewing on obesity might differ by weight status. Nevertheless, whether associations between TV viewing and CVD risks factors vary by weight status in adolescents have not been evaluated yet.

Therefore, the aim of the present study was to examine the independent association of excessive TV viewing with individual and clustered CVD risk factors in adolescents. Additionally, we also examined whether a different weight status modifies the associations between TV viewing and CVD risk factors.

Table 1 provides an overview of the characteristics of the study participants by sex. Adolescent boys were taller and heavier than girls but there was no significant difference in BMI levels. Sixty adolescent boys (28%) and thirty-nine adolescent girls (19%) were classified as overweight. For the CVD risk factors, adolescent boys had significantly higher abdominal body fat, and less favorable values of TG, HDL-C, glucose, and Apo A-1 than girls. In contrast, adolescent girls had less favorable values of TC, LDL-C and Apo B-100 than boys. Two hundred and twenty-five adolescents (53%) were classified in the "high TV viewing" group and a significantly higher percentage of adolescent boys than girls were classified in the high TV viewing group (p = 0.023).

Differences between weight status groups for CVD risk factors are shown in Table 2. Adolescents classified as overweight had less favorable values of WC, TG, HDL-C, LDL-C, Apo A-1, Apo B-100 than non-overweight adolescents controlling for age, sex, sexual maturation and race. Moreover, overweight adolescents also had a higher CVD risk score than non-overweight adolescents (p < 0.001).

Differences between TV viewing groups for CVD risk factors are shown in Table 3. Adolescents in the high TV viewing group had less favorable values of HDL-C, glucose and Apo A-1, controlling for potential confounders, as well as a lower CVD risk score (p < 0.001). When weight status was also included into the model as a confounder variable (model 2), the main results did not change (Table 3).

Finally, we analyzed whether the influence of TV viewing on CVD risk factors differ by weight status including a TV viewing ? weight status interaction term into the model (model 3). These analyses showed that differences between TV viewing groups for CVD risk factors remained significant, but there was a significant interaction effect of TV viewing ? weight status on WC (p = 0.002). Differences in WC across TV viewing and weight status groups in the adolescent sample are illustrated in Figure 1. Thus, the influence of TV viewing on WC in adolescents was only significant in the overweight group (p = 0.031), whereas this influence was attenuated in the non-overweight group (p > 0.05). All the results did not change when analyses were adjusted for SES as a confounder variable (data not shown).

The main results of our study suggest that excessive TV viewing is positively associated with CVD risk factors such as HDL-C, glucose, Apo A-1 and a continuous CVD risk score in adolescents, independently of potential confounders including weight status. These findings provide some evidence regarding the detrimental role of excessive TV viewing on CVD risk factors development in adolescence. In addition, further analyses revealed that weight status might modify the influence of TV viewing on abdominal body fat. Thus, the negative influence of TV viewing on WC remained in the overweight group but was attenuated in the non-overweight group.

Several studies in adults have shown similar findings to those found in the current study. In these studies the time spent watching TV has been positively associated with CVD risk factors [^11-14]. For example, Healy et al. [¹⁷] showed a detrimental dose-response association of TV viewing with WC, systolic blood pressure and 2-h plasma glucose in 4064 Australian adults [¹⁷]. Recently, Dustan et al. [³⁴] have shown that excessive TV viewing is associated with increased risk of all-cause and CVD mortality in 8800 Australian adults. Similar results were found in a men sample where the combination of time riding in a car and time spent watching TV were positively associated with CVD death [³⁵].

Despite of these findings in adult populations, the evidence linking TV viewing to CVD risk factors in children and adolescents is limited. For example, positive associations between TV viewing and abdominal body fat have been found in cross-sectional studies [¹⁵,¹⁶]. Nevertheless, several reviews have stated the weakness of these findings because poor results have been found in longitudinal and interventional studies [⁸]. Significant associations between TV viewing and blood markers (TC, insulin and glucose) [¹⁷,¹⁸] and blood pressure have also been found in children and adolescents [¹⁹,²⁰]. However, only in some cases body fat indicators were included as a confounder variable into analyses. For this reason, we cannot know whether the influence of TV viewing on CVD risk factors is independent of body fat-direct role- or whether body fat plays a pivotal role between TV viewing and a less favorable CVD risk factor profile-indirect role.

In the European Youth Heart Study, Ekelund et al. [²²] reported an independent association between TV viewing and metabolic syndrome risk factors in 1921 children and adolescents from 3 regions in Europe (Denmark, Estonia, and Portugal), but this association was mediated by adiposity (sum of 4 skinfold-thicknesses). Hence, the authors concluded the importance of reducing the time spent watching TV among children and adolescents in order to reduce directly body fat, and indirectly CVD risk factors. On the contrary, the associations between TV viewing and several CVD risk factors in the current study were not mediated by weight status. Interestingly, we also found that the influence of TV viewing on WC varied by weight status. The detrimental influence of TV viewing was maintained in the overweight group but not in the non-overweight group. These results suggest that reducing TV viewing time in overweight adolescents may have a beneficial influence on abdominal body fat. This interesting finding might explain the mixed effects found in intervention studies that decrease sedentary behaviors in apparently healthy and overweight children and adolescent [⁸,^24-26].

TV viewing is commonly used as a proxy to describe sedentary behavior even though the capacity to describe sedentary time using this approach is constantly questioned [³⁶,³⁷]. New research directions have used objective methods to assess daytime sedentary patterns and several reports have suggested the use of accelerometry for these purposes [³⁸]. Assessments using objective measures follow the idea of "inactivity physiology" posited by Hamilton et al. [³⁹]. In order to observe the differences between both methods, we have previously assessed sedentariness using sedentary time by accelerometry, and TV viewing and computer use by parent-report in a 3- to 8-year-old sample [²⁰]. On average, children spent 5 hrs/day in sedentary time and 1.5 hrs/day in screen time (TV + computer). These results show large differences between both methods. We consider that sedentary behavior with or without technology-mainly sitting time- and sedentary behavior using new technologies -mainly TV viewing- must be considered independently each other because both may be important in the development and prevention of CVD.

TV viewing in children and adolescents is usually associated with unhealthy behaviors. Firstly, TV viewing is associated with high consumption of soft drinks, salt, snacks, fat and low fruit and vegetable consumption [⁴⁰]. Secondly, TV viewing may contribute to the development of sleep problems from adolescence to adulthood [⁴¹]. Thirdly, TV viewing may restrict the possibility of youth to participate in physical activity [⁴²]. Fourthly, TV viewing contributes to inactive physiology as mentioned above. Hence, when reduction of time watching TV is suggested to prevent obesity and CVD risk factors, we may have influence on diet, sleep, physical activity and physical inactivity patterns that occur concurrently with TV viewing. Therefore, it is necessary to develop actions to reduce time watching TV from an early age. For children and adolescents, the American Academy of Pediatric recommends limits to no more than 2 hours per day of sedentary technologies, especially TV viewing [¹⁰]. Similarly, the Healthy People 2010 initiative promoted by the U.S. Department of Health and Human Services included the following objective: "to increase the proportion of adolescents who view television 2 or fewer hours on a school day" http://www.healthypeople.gov. As other authors [⁴³,⁴⁴], we could not use the 2 hours per day cut-off point. In our study, more than 50% adolescents spent >3 hrs/day watching TV, which is a similar prevalence to that found in a previous study in adolescents [⁴⁴].

Results found in our study confirm that an overweight status is positively associated with individual (WC, TG, HDL-C, LDL-C, apo A-1, apo B-100) and clustered CVD risk factors in adolescents. Therefore, a special focus of attention must be aimed at overweight adolescents who view too much TV. Moreover, future studies are encouraged to evaluate more health indicators in addition to obesity outcomes (e.g. BMI, % body fat, WC) because the evidence is scarce with other CVD risk factors. For example, in addition to the traditional CVD risk factors related to the metabolic syndrome, we have analyzed the associations of TV viewing with Apo A-1, Apo B-100, and lipoprotein(a). Thus, our results suggest that Apo A-1 is inversely and independently associated with TV viewing. Recently, childhood and adolescence Apo A-1 levels have been considered predictors of subclinical atherosclerosis in adulthood [⁴⁵].

Several limitations must be also mentioned in the current study. Our results are limited due to its cross-sectional design and causal directionality cannot be inferred. Moreover, one blood sample was used and that might not accurately reproduce long-term lipid and metabolic abnormalities. Unfortunately, blood pressure was not available in the AVENA Study, and therefore, we cannot compare findings from previous studies. TV viewing was measured using a single-response item. This type of questions to assess TV viewing has been widely used in large sample studies because objective measurements (direct observation, video, TV time manager) are not usually feasible in population studies. In children and adolescents, an adequate reliability and acceptable validity of these single 1-item questions to assess TV viewing have been observed [⁴⁶]. In spite of this, single-response questions are preferred rather than combined questions (e.g. How many hours do you usually spend watching TV and using computer per day?) because computer use and video game play may have a different effect on metabolic and physiologic parameters [⁴⁷]. Finally, we did not assess dietary patterns during TV viewing. Ekelund et al. [²²] found that associations between TV viewing and CVD risk factors and adiposity were attenuated when dietary behavior while viewing TV was controlled into the model although results were not shown.

Excessive TV viewing might aggravate several CVD risk factors in adolescents independently of their weight status. Furthermore, reducing the time spent watching TV might improve abdominal body fat in overweight adolescents. Specific intervention strategies are necessary in overweight adolescents who spent too much time daily watching TV. Experimental studies examining the role of TV viewing on traditional and new CVD risk factors in children and adolescents are warranted.

The authors declare that they have no competing interests.

Statistical analysis: DMG; Draft the manuscript: DMG; Funding and overall concept and design: AM, LAM, MGF and MD. Interpretation and acquisition of data: DMG, JPRL, PC, SGM, GVR, MMM, MGF, MD, LAM, OLV, JCE and AM. All authors read and approved the final manuscript.

The pre-publication history for this paper can be accessed here:

http://www.biomedcentral.com/1471-2458/10/274/prepub

The authors express their sincere gratitude to the adolescents who participated in this study. The AVENA Study was supported by the Spanish Ministry of Health, FIS (00/0015) and grants from Panrico S.A., Madaus S.A. and Procter & Gamble S.A. DMG was supported by a grant from the Spanish Ministry of Education and Science (AP2006-02464). JPRL, GVR, MGF and LAM were also supported by a grant from the Spanish Ministry of Health: Maternal, Child Health and Development Network (number RD08/0072).

AVENA Study Group:

Coordinator: A Marcos, Madrid. Local coordinators: MJ Castillo, Granada; A Marcos, Madrid; S Zamora, Murcia; M Garc?a Fuentes, Santander; M Bueno, Zaragoza. Granada: MJ Castillo, MD Cano, R Sola (Biochemistry); A Guti?rrez, JL Mesa, JR Ruiz (Physical fitness); M Delgado, P Tercedor, P Chill?n (Physical activity), M Mart?n-Matillas, F Carre?o, FB Ortega, GV Rodr?guez, R Castillo, F Arellano (Collaborators). Universidad de Granada. E-18071 Granada. Madrid: A Marcos, M Gonz?lez-Gross, J W?rnberg, S Medina, F S?nchez-M?niz, E Nova, A Montero, B de la Rosa, S G?mez, S Samart?n, J Romeo, R ?lvarez, (Coordination, immunology) A ?lvarez (Cytometric analysis) L Barrios (Statistical analysis) A Leyva, B Pay? (Psychological assessment). L Mart?nez, E Ramos, R Ortiz, A Urzanqui. (Collaborators). Instituto de Nutrici?n y Bromatolog?a. Consejo Superior de Investigaciones Cient?ficas (CSIC). E-28040 Madrid. Murcia: S Zamora, M Garaulet, F P?rez-Llamas, JC Baraza, JF Mar?n, F P?rez de Heredia, MA Fern?ndez, C Gonz?lez, R Garc?a, C Torralba, E Donat, E Morales, MD Garc?a, JA Mart?nez, JJ Hern?ndez, A Asensio, FJ Plaza, MJ L?pez (Diet analysis). Dpto. Fisiolog?a. Universidad de Murcia. E-30100 Murcia. Santander: M Garc?a Fuentes, D Gonz?lez-Lamu?o, P de Rufino, C Redondo-Figuero, MJ Noriega, R Lanza-Saiz, T Amigo (Genetic study). Dpto. Pediatr?a. Universidad de Cantabria. E - 19003 Santander. Zaragoza: M Bueno, LA Moreno, A Sarri?, J Fleta, G Rodr?guez, CM Gil, MI Mesana, JA Casaj?s, Vicente Blay, Mar?a Guadalupe Blay. (Anthropometric assessment). Escuela Universitaria de Ciencias de la Salud. Universidad de Zaragoza. E-50009 Zaragoza.