This study investigated stability and changes in screen-based sedentary behaviours in a group of Norwegian children in a transition phase between childhood and adolescence, and factors associated with stability and change in total screen time.
Findings based on both absolute and rank measures indicate that the SB investigated had fair to moderate tracking between the beginning of 6th and the end of 7th grade. The tracking patterns identified in this study are similar to those reported for children in similar age groups and over similar periods [10
], although comparability between tracking studies is complicated among other things by the duration of follow-up and by the method used to assess tracking including whether adjustment for potential covariates has been done or not. Tracking patterns were largely similar between boys and girls, except for a small difference in the tracking coefficients of electronic games/computer use. A review concluded that there seemed to be little evidence for any gender differences in tracking of SB [10
], and our study supports this. The findings that boys and girls watch the same amount of TV and that boys use more electronic games/computers than girls have similarly been documented in the literature previously [19
There was an increase in time spent on both SB during the studied period, but TV viewing seemed to start levelling off in boys. Similarly, over 70% of both boys and girls exceed the recommendations in some countries for a maximum daily total daily electronic media use time of 2 hours or less [31
]. These findings are in line with the reported increases in SB and in particular electronic media use that occurs in childhood and early adolescence [1
]. Accessibility to SB seems to be easier than that to PA in most people [11
], and this accessibility might increase as the children grow older because they spend more time alone at home and may also have TVs and computers in their own rooms. These factors might further increase screen time.
Higher screen time at BL and higher odds of tracking high TST among girls with an ethnic minority background compared to ethnic Norwegian girls might be due to factors such as low integration as well as cultural differences which might make other recreational activities less common among ethnic minority girls. A study among adolescents has previously shown that girls with an ethnic minority background are likely to be more inactive than ethnic Norwegian girls, whereas no difference was found in boys [39
]. Parental education was also found to be inversely associated with TST among girls. The inverse relationship between parental education and use of electronic media has previously been documented [19
Not living with married/cohabitating parents was positively associated with an increase in TST, and with tracking high TST among males. A review of correlates of TV viewing among youth concluded that young people in single-parent/guardian families consistently watch more TV than those from two parent/guardian families [19
]. Overweight/obese boys in the study were also more likely to tracking high TST compared to those not overweight. This might indicate a bi-directional relationship: high TST might influence children's BMI and high BMI might lead to children reverting to more SBs. As mentioned earlier, several studies have shown an association between screen time behaviours and in particular TV use and being overweight [2
]. One possible explanation for the presence of this association in boys only in this study might be the fact that boys engage in leisure time PA more often than girls and they also tend to engage in more strenuous exercises and focus more on achievement [42
]. Thus, overweight/obese boys might want to withdraw from such PA and instead revert to sedentary leisure time activities such as screen time activities because of potential physical limitations as well as psychological fears of lack of competence to engage in such PAs. Other studies have similarly found an association in male adolescents and none among their female counterparts [43
], in line with the findings from this study.
A variable consistently inversely associated with TST in both genders was self-efficacy related to barriers to PA, one barrier being screen time itself. Self-efficacy related to barriers to PA was investigated for its association with SB in this study because it is the strongest and most consistent correlate of PA in this age group [44
], also found previously in the HEIA study [46
]. The displacement hypothesis suggests that SB can displace time that could otherwise have been used for PA [47
]. Findings from several studies do not support this hypothesis, as indicated by a large meta-analysis [5
]. Negative associations between SB and PA have nevertheless been documented [12
]. Self-efficacy related to barriers to PA, by positively influencing levels of PA, can consequently potentially lead to decreased time spent on SB, which appears to be the case in this study. It has been suggested that targeting self-efficacy should have a prominent position in efforts to promote PA [45
]. The results of this paper suggest that such efforts might also be successful in lowering time spent on electronic media use.
It has been suggested that SB might be more strongly associated with socio-demographic factors than modifiable factors such as psychosocial or behavioural variables [19
]. It is nevertheless possible that variables from the psychological and behavioural domains play a greater role in influencing TST in this age group of children, and even mask the importance of socio-demographic factors. There is thus a need to look into such correlates of TST in children in this age group.
It is not likely that there was any difference in tracking and change patterns between participants and drop-outs as the attrition analyses did not show any difference in baseline characteristics between these two groups.
Strengths and weaknesses
This was a longitudinal study with a relatively large sample size. In addition, the rate of retention in this study was very high. The study also provides much needed information about screen time patterns in a Norwegian context. It is also one of a few studies looking at factors associated with tracking patterns of TST. Different screen time behaviours were included. Gender-related differences as well as similarities were also reflected.
The use of self-reported measures is associated with problems of validity and reliability, in particular in younger children. Nevertheless, there was a good test-retest reliability of the outcome measures. In addition, the trends in screen time obtained in this study were found to be in line with those found in other studies, including the gender differences noted [19
]. The study looked at stability and change in screen time behaviours, as well as associated factors and comparisons between genders, and we believe that the measures used are reasonably reliable and valid for such purposes. In general, the more unreliable the measures, the higher the chance of Type II errors; that is not finding differences and associations that in fact exist. Therefore, differences and associations in the study were probably underestimated rather than overestimated.
Finally, the study is restricted to a single geographical area, therefore generalizability is limited.