Are Physical Activity, Screen Time, and Mental Health Related During Childhood, Preadolescence, and Adolescence?

11-Year Results From the German Motorik-Modul Longitudinal Study

Claudio R. Nigg; Kathrin Wunsch; Carina Nigg; Claudia Niessner; Darko Jekauc; Steffen C. E. Schmidt; Alexander Woll


Am J Epidemiol. 2021;190(2):220-229. 

In This Article


In the current study, we examined the relationships among PA, ST, and MH by sex in a sample of German schoolchildren from a longitudinal study over a period of 11 years. The behavioral portion of the results indicated that ST and PA did not predict each other—which replicates findings from another cohort study.[23] TV/video watching in childhood and preadolescence predicted PC/Internet use in preadolescence and adolescence, respectively. Although they were not the focus of our study, these results underline the relative independence of PA and ST, with an implied need to address both behaviors separately.

Somewhat surprisingly, PA only had a few relationships with MH. For female and to some extent male children, greater numbers of MH challenges negatively influenced preadolescent PA. The higher female children's PA level, the lower their preadolescent prosocial behavior, whereas the reverse was true (higher childhood PA, higher preadolescent prosocial behavior) for male children; in addition, the higher female preadolescents' PA level, the higher their prosocial behavior. Therefore, child and preadolescent PA may not be protective against self-reported MH challenges; rather, childhood MH may be a risk factor for lower preadolescent PA, especially in females. This is in line with the inconsistent results of previous studies, providing mixed evidence, especially among longitudinal research studies (e.g., Birkeland et al.;[38] for an overview, see Bauman[39]). Potential explanations may be that PA's MH benefits are context-dependent (e.g., group-/team-related[40,41]) and that PA is not always a positive experience for youth (e.g., possible teasing/bullying[42,43]). There is some evidence that PA plays an ambivalent role in MH (prosocial behavior), which needs further exploration. Furthermore, the effects of PA on MH are rather small, such that effect reversal is possible.

However, the reliability of self-reported PA is age-dependent;[34] the older participants are, the better the reliability of their PA self-reports. Therefore, reliability issues could be an explanation for this age-dependent result. The PA and MH results might also be due to the many changes that take place during this developmental period, which includes puberty and changes in the structure of the environment. Specifically, childhood and preadolescence have a structure (school) providing PA. School physical education classes are viewed as an ideal setting for the promotion of regular PA, because up to 97% of elementary school children participate in some sort of physical education program.[44] Adolescents have not yet completed the full social transition to adulthood and are in a period that has been termed the "demographically dense years," which incorporate different life changes regarding transitioning out of school to employment, higher education, and family formation, all of which influence PA.[45,46] Some maturation processes may account for MH inconsistencies.

On the other hand, ST (operationalized through TV/video watching at all 3 time points, PC/Internet use at T1 and T2, and PC/Internet use outside of computer games at T3) had more pronounced relationships with MH, although this was dependent on sex. For females, behavior influenced MH—more childhood and preadolescent ST negatively influenced self-reported preadolescent MH and adolescent MH, respectively, confirming results from other studies.[47,48] This finding is in line with literature showing consistently negative associations between sedentary behavior and self-reported MH.[16] Because there is growing evidence that an excessive amount of time devoted to sedentary behaviors (predominantly ST) is associated with a higher risk of diseases and all-cause mortality,[49,50] regardless of leisure-time PA level, sedentary behavior in children and preadolescents should be addressed in consideration of its presumed adverse effects on (mental) health over the course of maturation.[51]

Underlying mechanisms possibly explaining that sedentary behavior leads to self-reported MH issues are that 1) sedentary behavior (especially ST) decreases social interaction, and less social interaction or social support is related to MH issues;[52,53] that 2) ST is related to obesity[54,55] and other chronic diseases, which involve MH comorbidity;[56] and that 3) ST does not facilitate excretion of (neuroendocrine) biomarkers as PA does (like epinephrine or dopamine, for example[57]), thus possibly being implicated in self-reported MH issues.[58,59] Moreover, other studies have shown that a decrease in sedentary behavior is related to an increase in PA among nonobese children,[60] enabling them to profit more from the physiological as well as psychological health-promoting mechanism of PA.[61] However, it needs to be taken into account that ST demands change over time, as schoolchildren have to use PCs throughout school time as well, and employed adolescents often have to be available all day, which also increases ST. It is important to investigate associations between different forms of ST and subsequent MH due to varying characteristics of ST (e.g., differences between Internet surfing (an independent activity), social networking (socially interacting), and online computer gaming (employing tactical strategies in the games)). The rapidly changing technology also affected our 11-year cohort study, necessitating the change in ST assessment at T3. For males, MH seemed to influence behavior—with the most robust findings being that having more MH issues during preadolescence predicted more ST in adolescence. A potential coping strategy of young people with MH challenges during this developmental period may be to withdraw from live social interactions and turn to TV/video or PC/Internet use, which may be counterproductive. The fact that MH challenges are a risk factor for increased ST has been documented in the literature and implies that efforts to address MH challenges need to incorporate positive/productive alternate activities to ST.

Taken together, these results suggest that for females, higher TV/video and PC/Internet use was related to higher MH issues over the course of maturation. In males, there was some support for the finding that a greater number of MH challenges in preadolescence increased ST (both TV/video time and PC/Internet time) in adolescence. Thus, for females in this age group, greater ST is a risk factor for poorer MH, and for males, poorer MH is a risk factor for greater ST. These sex differences have to be taken into account when translating current results to public health policy.

This longitudinal cohort study had a sufficiently large sample size and sufficient statistical power to adequately examine the influences of PA and ST on self-reported MH, and external validity was estimated as high. Results for mean PA and mean ST are rather difficult to compare with those for other cohorts, since most other studies used the percentage of participants who met a particular guideline (at least 1 hour of PA per day, less than 2 hours of ST per day, etc.[62]). Nevertheless, in the analyzed cohort, descriptive results for number of days per week with at least 60 minutes of PA showed a decrease of about 0.7 days of PA from childhood to preadolescence and a decrease of another 0.25 days from preadolescence to adolescence. Regarding ST, there was an increase in TV/video time of 15 minutes per day from childhood to preadolescence and about 40 minutes per day from preadolescence to adolescence. These trends are comparable to those seen in other studies carried out in the United States[63] or across Europe.[64,65]

Furthermore, the current study covered an important phase of life, including the developmental transitions from childhood to preadolescence and from preadolescence to adolescence.[66] Moreover, different aspects of MH were examined. Limitations to consider when interpreting these results are that the study used self-report measures, which may lead to social desirability in response behavior and potentially less reliable estimates in children and preadolescents.[67] Relatedly, the MH measures have low-to-moderate internal consistencies. The duration of time between the assessments may have masked some of the changes taking place during these developmental periods. Physical growth in childhood and the maturation process in preadolescents may also have implications for PA and ST. Therefore, the time periods between measurement phases should be kept shorter in future cohort studies in order to observe any evidence of shorter-term effects. The ST item about computer and Internet use was slightly revised from T2 to T3, which may have influenced the comparability of the measures over time. However, the similarity of the trends across time and the congruence of direction with the TV/video indicator lends some credibility to this indicator. The study did not have an indicator for sleep quality or duration across all time points, so we were unable to draw conclusions about the 24-hour (full-day) period. Several models were fitted, and only significant path coefficients were included in the models. The problem of possible α coefficient inflation cannot be ruled out. Therefore, these results should be treated as exploratory findings which need to be confirmed in further studies. The strength of the nationally representative sample also limits our conclusions to this geographic region (Central Europe). For future studies, cross-cultural investigations are needed to gain further insights into possible mechanisms of the relationships among PA, ST, and MH in children and adolescents.

In conclusion, this study is one of the first to have examined the relationships of PA and ST with self-reported MH outcomes from childhood to adolescence in a longitudinal setting in Germany. The current examination provides some evidence of a negative impact of ST in childhood and preadolescence on self-reported MH in preadolescence and adolescence for females, whereas MH in preadolescence seems to be a risk factor for adolescent ST for males. These findings are salient because of the importance of being in good MH when pursuing higher education, starting a family, or embarking on a career—all hallmarks of late adolescence. In future studies, researchers should explore innovative and effective methods for reducing ST in childhood, especially among females with early signs of addiction and/or MH issues, and for addressing MH challenges in preadolescent males. Although the current study does not provide strong evidence for a protective effect of childhood and preadolescent PA on MH, it is prudent to promote PA because of its numerous other benefits.[35] Taken together, the current results should encourage public health policy-makers to thoroughly engage in child, preadolescent, and adolescent health promotion using interventional approaches which integrate adolescents' media usage needs with increases in PA and decreases in prolonged sedentary behavior.