Associations between physical activity, fitness, cognitive and academic performance in Swedish adolescents: Findings from a cross-sectional study

Karin Kjellenberg; Rui Wang; Jonna Nilsson Horre; Björg Helgadóttir; Örjan Ekblom; Amika Singh; Gisela Nyberg

doi:10.1371/journal.pone.0344087

Abstract

Background

Adolescence is a crucial phase for the development of cognitive abilities linked to academic performance. Factors such as physical activity (PA) and fitness have been hypothesized to be linked to these outcomes, although the evidence is inconclusive. This study aimed to examine the associations between PA, fitness, academic, and cognitive performance in the same sample of adolescents.

Methods

In this cross-sectional study, 1139 Swedish adolescents (mean age 13.4 years) participated. We assessed PA (accelerometry), cardiovascular fitness (submaximal ergometer test), cognitive performance (computer-based test assessing episodic and working memory with 3 tests per domain), and academic performance (grades in math and language, Swedish). We utilized multilevel mixed models to explore associations and structural equation modeling to perform mediation analyses.

Results

We found a weak positive association between cardiovascular fitness and math grades (b: 0.18, 95% CI: 0.09, 0.26 β: 0.18), language grades (b: 0.16, 95% CI: 0.07, 0.25 β: 0.13), and cognitive performance (b: 0.01, 95% CI: 0.00, 0.02 β: 0.11). Further, cognitive performance scores mediated 40% of the associations between fitness and math, and language grades. Adolescents with parents with short education (≤12 years of education) or foreign-born had lower fitness levels and grades in math and language (all p < 0.05).

Conclusion

Fitness was the only significant factor associated with cognitive and academic performance, although the association was weak. Given our findings of low fitness and academic performance among adolescents with parents with short education or parents born abroad, future studies should tailor their interventions to ensure the inclusion of these groups.

Citation: Kjellenberg K, Wang R, Horre JN, Helgadóttir B, Ekblom Ö, Singh A, et al. (2026) Associations between physical activity, fitness, cognitive and academic performance in Swedish adolescents: Findings from a cross-sectional study. PLoS One 21(3): e0344087. https://doi.org/10.1371/journal.pone.0344087

Editor: Metha Chanda, Kasetsart University Faculty of Veterinary Medicine Kamphaengsaen Campus, THAILAND

Received: August 14, 2024; Accepted: February 15, 2026; Published: March 9, 2026

Copyright: © 2026 Kjellenberg et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: The datasets cannot be publicly shared due to ethical restrictions. Some participating schools were small, and the dataset includes sensitive information, such as anthropometric measures and pubertal development data. Public disclosure could risk participant privacy. Data are available from the GIH Institutional Data Access (contact via gisela.nyberg@gih.se or gdpr@gih.se) for researchers who meet the criteria for access to confidential data.

Funding: This study is part of the project “Physical activity for healthy brain functions in school youth”, which was funded by the Knowledge Foundation (grant 20180040 received by GN and ÖE) and conducted in collaboration with Coop, IKEA, Skandia, Skanska, Stockholm Consumer Cooperative Society, and the Swedish Crown Princess Couple’s Foundation/Generation Pep. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Abbreviations: PA, Physical activity; MVPA, Moderate to vigorous physical activity; MPA, Moderate physical activity; VPA, Vigorous physical activity; SEM, Structural equation modeling; RMSEA, Root Mean Square Error of Approximation; CFI, Comparative Fit Index

1. Introduction

Adolescence is a crucial stage of development where important social, emotional, and higher cognitive skills are acquired. These skills are essential for preparing individuals to succeed in different aspects of society, such as in school or work life [1].

Therefore, adolescence is considered a critical period for brain development, during which environmental and behavioral factors are known to play a significant role [1,2]. Physical activity (PA) has been suggested as a potential factor that could benefit cognitive and academic performance; nevertheless, most adolescents do not meet the PA recommendations. For instance, only 32% of Swedish adolescents (aged 11–16 years) met the recommendations [3]. When reviewing previous studies in the field, it is important to acknowledge that while cognitive and academic performance are often discussed together, they are distinct concepts [4]. Cognitive abilities are shaped by many factors, such as age [5], parental and environmental factors [6], and genetics [7]. These skills include for example working memory and episodic memory, which are essential for academic performance [5,8]. For instance, episodic memory facilitates the storage and retrieval of information within specific contexts [8], while working memory aids in organizing ideas, problem-solving, planning, and goal achievement [5]. However, academic performance is also shaped by other factors, including a student’s engagement and participation in the school environment, as well as the quality and quantity of education received [4].

Several theories propose the mechanisms by which PA improves cognitive performance, such as alterations in grey and white matter, as well as improving neural properties like functional activation or connectivity, which are crucial for cognitive performance [9] and thereby, academic performance. However, the evidence in this area is primarily based on studies conducted on animals or adults. While a small-scale PA intervention in children study found that PA participation increased white matter integrity and that these improvements were associated with improved cognitive performance [10], there are insufficient studies to establish a link in adolescents.

Habitual PA participation, more specifically vigorous-intensity PA (VPA) enhances cardiorespiratory fitness. For example, a review found that school-based PA interventions improved cardiorespiratory fitness among adolescents [11]. Consequently, fitness is proposed to be a factor in the causal pathway linking PA to academic performance. Cross-sectional studies involving children have found that fitness plays a mediating role in the association between PA and academic performance [12,13]. One suggested mechanism is that high fit adolescents allocate attentional resources more efficiently (indicated by the monitoring of the event-related potential component P3, thought to reflect this ability), resulting in better cognitive performance, compared to their less-fit counterparts [14]. Cross-sectional studies have also found fitness to have a direct positive association with cognitive [15] and academic performance [16]. Although it is important to acknowledge that cardiorespiratory fitness is influenced by pubertal status [17], highlighting the need to control for puberty status among adolescents.

While single-session PA has been shown to produce beneficial effects on some aspects of cognitive performance [2], long-term PA interventions have demonstrated low or no effect on cognitive or academic performance [15,18]. Some of the potential limitations of these interventions are low fidelity (i.e., intervention was not executed as intended), and the failure to monitor the proposed improvements in participants’ PA during the interventions [18]. This makes it challenging to discern whether null results in PA interventions arise from PA not enhancing cognitive and academic performance, or from the intervention failing to increase PA levels, thus hindering investigation into the impact of PA [18].

While the inconclusive results could be due to the failure to induce changes in PA habits, cross-sectional studies investigating the association between habitual PA and cognitive or academic performance have also shown mixed results. Although positive associations have been found in some aspects of cognitive performance, mainly executive functions, as well as academic performance [2,15,19]. Previous studies have limitations, such as lack of robust measures for PA and cognitive performance and the omission of key confounders.

A review concluded that the accelerometer-based studies showed inconclusive associations between PA and academic performance, while self-reported studies yielded more consistent results, although many of them overlooked important confounding variables [19]. Yet, relying on self-reported PA limits the possibility to distinguish between intensities and adolescents tend to over-report their time spent in PA. For instance, one study showed that only 7–20% of self-reported MVPA and VPA could be confirmed using accelerometry [20].

Another aspect is that studies that encompass both children and adolescents are frequently discussed collectively although most studies have been performed in children, with limited research focusing specifically on adolescents. This is problematic considering age has been proposed as a moderating factor [21]. For instance, intervention studies suggest larger effects in participants under 12 years old [22]. Additionally, some research suggests that individuals with lower baseline performance may experience greater benefits from physical activity interventions on cognitive or academic performance [23,24]. Furthermore, positive effects of exercise have also been observed on cognitive behaviors in individuals with Attention Deficit Hyperactivity Disorder [25].

Moreover, there is a lack of studies examining the effects of different PA intensities separately. A cross-sectional study in children, for example, found a positive association between VPA and academic performance, but no such association between MPA and academic performance [26], indicating that the relationship may differ based on the intensity of PA. Lastly, many studies on cognitive performance rely on only one or two tasks, varying in format (e.g., letters, numbers, visual cues) and response mode (e.g., computer, oral, paper, and pencil). This limited approach may not accurately capture a cognitive skill but rather proficiency in specific tests. Moreover, most previous research has focused on executive functions [2], whereas other cognitive skills, such as episodic memory are less explored.

In conclusion, there is a scarcity of studies using detailed measures of PA and cognitive performance to investigate how PA at different intensities, along with fitness, are associated with cognitive and academic performance within the same sample. In addition, few studies have included key confounding factors, such as parental education, which is known to be strongly associated with both academic and cognitive outcomes. Furthermore, there is a need for studies exploring the mediating role of fitness and cognitive performance in these associations to gain deeper insight into how these factors covary among adolescents.

The aim of this study was therefore to investigate how accelerometer-based PA (MPA and VPA) and fitness are associated with cognitive (working memory and episodic memory) and academic performance (grades in math and language) among Swedish adolescents while adjusting for gender, puberty, parental education, and parental country of birth. Additionally, we aimed to explore whether fitness and cognitive performance mediated these associations. The hypothesis and theoretical framework are shown in Fig 1.

Download:

Fig 1. Theoretical model outlining the proposed associations between physical activity and academic performance in adolescents.

The hypotheses of this study were pre-registered at: https://osf.io/ep74n. We hypothesize that physical activity of different intensities—moderate (H1) and vigorous (H2)—will be associated with cognitive performance, as well as academic performance (H3 and H4). Additionally, we propose that fitness will be associated with both cognitive (H5) and academic performance (H6). We further hypothesize that cognitive performance will serve as a mediator in the relationships between physical activity and academic performance (H7). Finally, we also hypothesize (H8) that fitness mediates the association between vigorous physical activity (VPA) and academic performance. This hypothesis was not specified in the pre-registration but was added later, based on findings from similar studies in children. The grey dotted arrows in the model represent potential reverse associations, which are not examined in this study.

https://doi.org/10.1371/journal.pone.0344087.g001

2. Methods

2.1. Sample and study design

This study is part of the larger cross-sectional study “Physical Activity for Healthy Brain Functions in School Youth”. Approval for the study was obtained from the Swedish Ethical Review Authority (reference: 2019−03579, date of approval: 2019-08-06). All participating adolescents and their guardians provided written informed consent. In cases where written consent could not be obtained by the guardians (for instance one guardian lived elsewhere), verbal consent was obtained by phone from that guardian. This verbal consent was carefully documented and witnessed by the research staff. The design, recruitment, and protocol of the study have been described in more detail elsewhere [27].

In brief, all schools within 2–3 hours’ driving distance from the Swedish School of Sport and Health Sciences (GIH) in Stockholm, Sweden were invited to participate in the study (n = 558). The invitation was sent out by e-mail to the principals or co-principals of the schools in February 2019. A total of 84 schools accepted the invitation; however, school inclusion was capped at 40 schools, ensuring diversity in geographical location, school type, and parental education levels for practical reasons. Six schools dropped out due to time constraints, leaving a final sample of 34 schools (see supporting information S1-S3 Figs).

The schools provided a list of contact information for students in the participating classes, allowing invitation letters and consent forms to be mailed to adolescents and their guardians. Teachers collected the consent forms and informed the researchers of the number of participants from each class and school. The inclusion criteria were grade seven students, with the ability to read or understand Swedish, as the questionnaires and cognitive test instructions were in Swedish. The students were recruited between August 12^th – December 6^th, 2019.

A total of 1139 grade seven students from these schools participated, representing a 73% participation rate out of 1556 eligible students. Of the 1556 eligible students, 63 were unable to participate due to illness or absence, 351 did not provide consent and therefore declined, and three were excluded because their consent forms were incomplete (e.g., missing a signature from one or both guardians).

2.2. Protocol

The data was collected at our research center at GIH in Stockholm, Sweden, between September 26^th – December 6^th, 2019. Each day, a maximum of 60 adolescents (two groups of 30) participated.

The visit typically lasted four to five hours. Upon arrival, adolescents were given a standardized breakfast and briefed on the day’s activities. They were then split into two groups, each accompanied by at least one teacher. Participants completed a questionnaire, were equipped with a hip-worn accelerometer performed a cognitive test battery and a submaximal cycle ergometer test. The adolescents received a 300 SEK gift card (approximately 30 USD) for their participation in the study.

2.3. Measurements

2.3.1. Cognitive performance (working memory and episodic memory).

A computer-based test battery with six tests to assess two cognitive domains: working memory (letter updating, numerical 3-back, spatial updating) and episodic memory (word recall, number-word recall and object-position recall). The test battery is described in more detail in the supporting information (S2-S6 Figs) and by [28]. The internal variability (Cronbach alpha) was 0.76 for cognitive performance (based on all tests), and 0.67 for the working memory and 0.73 for episodic memory domains.

Each test variable was transformed to T scores, centered to a mean of 50 and a SD of 10 to allow for meaningful comparisons between all outcomes. A structural equation model (SEM) was employed to generate a latent variable, incorporating loadings from all cognitive tests (available in supporting information S5 Fig). The model demonstrated a good fit, as evidenced by a Root Mean Square Error of Approximation (RMSEA) at 0.042, below the cut-off threshold of <0.06 and a Comparative Fit Index (CFI) of 0.990, exceeding the minimum criterion of 0.90. Subsequently, factor scores for each latent variable (global cognitive performance, working memory, and episodic memory) were extracted from the models and utilized as outcome measures for cognitive performance. Global cognitive performance was used as the primary outcome but findings on working memory and episodic memory are presented in supporting information S4 Table.

2.3.2. Academic performance (grades in math and language).

Academic performance data was obtained by official records from Statistics Sweden, using final grades from math and language (Swedish) in grade six. The language subject has two variations: Swedish for native speakers (88% in our sample) and Swedish as a second language (12%). Within the Swedish school system, an A signifies the highest grade attainable, while an F denotes a failing grade. These variables were converted into numeric values ranging from one to six, one representing F and six representing A. The variables were transformed to T scores, centered to a mean of 50 and a SD of 10 to allow for meaningful comparisons between all outcomes (cognitive and academic performance). Grades in math and language were used as the primary outcomes, while results from a combined academic performance outcome (average grade) are presented in supporting information S5 Table.

2.3.2. Physical activity (PA).

PA was assessed using a hip-worn triaxle accelerometer (model GT3X + , Actigraph, LCC, Pensacola, FL, USA) with a sampling rate of 30 Hertz. Participants were instructed to wear the accelerometer all waking hours, but not during water-based activities, for a period of seven days following the visit at the research center. The accelerometer data were processed in Actilife (v6.13.3) as uniaxial data, using 5 seconds epoch time intervals. Non-wear time was defined as 60 minutes of zero counts and no spike tolerance. An individual time filter for expected wear-time was applied to the data based on the participant’s reported awake/asleep time (extracted from the questionnaire). A minimum of 500 minutes of wear time was considered a valid day (after excluding non-wear time), and at least three valid days (at least one weekend day) were required to be included in the analyses [29]. The data were categorized using the following cut-offs counts per minute: MPA 2296–4011 and VPA ≥ 4012 [30]. Minutes spent in each intensity were divided by wear time and multiplied by 100 to create a variable representing the percentage in each intensity.

A variable for MVPA was created by combining MPA and VPA intensities. However, these results are only presented in the supporting information (S6 Table) for reference purposes, considering MVPA is commonly used in the research field but falls outside the scope of our study aim.

2.3.4. Cardiovascular fitness.

Cardiovascular fitness was measured through the Ekblom-Bak test, an incremental submaximal cycle ergometer test. To estimate maximal oxygen consumption (expressed as milliliters of oxygen per kilogram of body weight per minute, mL/kg/min) the change in heart rate response between two different workloads was used while accounting for gender and age. To account for pubertal status, prepubertal boys (Tanner stage 1–2) were estimated using the equation for females, as recommended by Björkman et al, [31].

2.3.5. Other variables.

Register data on parental education was provided by Statistics Sweden. The variable was dichotomized into short education (≤12 years of education, indicating completion of Swedish upper secondary school or less) or long education (>12 years of education, typically referring to university education), using the parent with the highest level of education.

In the questionnaire, adolescents provided information about their gender (girl, boy, or other) and their parental birth country (one parent born in Sweden/one parent born abroad, both parents born in Sweden, or both parents born abroad).

The adolescents’ body weight was assessed using a calibrated Tanita BC-418 scale (Tanita Corporation, Tokyo, Japan). Height was measured using a SECA 5123 stadiometer (SECA Weighing and Measuring Systems). Two weight and height measurements were taken and averaged, with the final values rounded to the nearest 0.1 kg or meters. Body Mass Index (BMI) was calculated as weight (in kilograms) divided by height squared (in meters). BMI classification was determined according to the International Obesity Task Force guidelines [32]. Pubertal status was self-reported by the adolescents using Tanner drawings [33]. Classroom noise levels during the cognitive test were rated as low, medium, or high by the research staff.

2.4. Statistical analysis

The statistical analysis plan and hypotheses were pre-registered at: https://osf.io/ep74n, as there were deviations from the statistical plan these have also been uploaded to the registry. Data analysis was conducted using Stata/SE version 18 (StataCorp LLC, College Station, TX, USA). Descriptive statistics were used to summarize the data, including means and standard deviations, or frequencies and proportions. Group differences (presented in Table 1, and in supporting information S2–S3 Tables) were assessed either through independent t-tests, ANOVAs, or Chi2-tests.

Download:

Table 1. Descriptive characteristics of the study sample by gender (n = 1139).

https://doi.org/10.1371/journal.pone.0344087.t001

Multilevel mixed effect model was used to study the associations between PA (MPA or VPA), fitness and cognitive or academic performance (grades in math or language), modeling school as level 1 and individual adolescents as level 2. A random intercept was introduced for each school to account for variations at the school level while the slopes were fixed. This approach was based on the assumption that the associations between PA or fitness and cognitive or academic performance would be consistent across schools. Overall, the intraclass correlation coefficient showed that between 8–20% of the variance in the academic performance models could be attributed to the clustering of schools. The analyses only included cases with complete data, that is, participants with no missing values for any variables included in the models. As multiple models were estimated (e.g., MPA, VPA, and fitness across several outcomes), we adjusted for multiple comparisons using the Benjamini–Hochberg False Discovery Rate (FDR) approach, applying a false-discovery rate of 5% [34], which resulted in a corrected significance threshold of alpha below 0.018. This procedure was implemented to control for multiple testing and to reduce the risk of Type I error.

Extreme outliers in all variables were identified using the “Outlier labeling rule”, which calculates upper and lower cut-offs based on the formula: upper = Q3+(3*(Q3-Q1)) and lower = Q1-(3*(Q3-Q1)) [35]. However, no extreme outliers were identified in any of the variables, therefore, no observations were excluded based on this rule. The assumptions for mixed models were checked. In certain models, the assumptions of normal distribution and homoscedasticity were violated. To address this, we employed robust estimates. All models were run both crude and adjusted (including gender, puberty status, parental education, and parental country of birth, based on previous literature). The adjusted results will serve as the basis for interpretation, while the crude results are included for reference. Given the interconnected nature of these exposures and the study’s focus on examining the association of each exposure with the dependent variable, rather than their collective influence, we chose to test each exposure in separate models.

To further examine the associations, we performed path analysis using SEM. These models explored whether cognitive performance mediated the significant associations between the exposures and academic performance. In addition to the pre-registered hypotheses, we conducted an exploratory analysis to test a pathway (H8) suggested by prior literature, that fitness mediates the association between VPA and academic performance [12,13]. Although the association between VPA and academic performance was not statistically significant, we decided to explore fitness as a potential mediator, given the link between VPA and fitness. MPA was not tested as an exposure in the mediation analyses. We tested both mediation analyses within the same model: (1) the association between VPA and academic performance, mediated by fitness, and (2) the association between fitness and academic performance, mediated by cognitive performance, as the models were assumed to be correlated and not independent (see the theoretical outline in Fig 1). The mediation models indicated good model fit (RMSEA = 0.028 and CFI = 0.999 in the math model and RMSEA = 0.026 and CFI = 0.999 in the language model). To assess the indirect and direct effects we used the Medsem package in Stata, following the Barron and Kenny approach to test mediation, using the modified significance test by Iacobucci et al. using the Monte Carlo method with 2500 samples, utilizing a significance level of p < 0.05, this approach has been described in more detail by Mehmetoglu [36].

Sensitivity analyses were performed, excluding participants with the latest start in the cognitive test battery (as this group exhibited lower scores), comparing the results to those including all participants. Furthermore, sensitivity analyses were also conducted including participants with at least one valid day of accelerometer measurement, compared with the stricter criterion of at least three valid days. Sensitivity analyses were also conducted to explore the influence of classroom noise level by comparing the results between the models with and without this confounder.

3. Results

3.1. Participants

A total of 1139 adolescents (mean age 13.4 ± 0.3) participated in the study, 51% were girls (see Table 1 for characteristics of the participants). Regarding their parents, two-thirds (66%) had parents with more than 12 years of education, 59% had both parents born in Sweden and 26% had both parents born abroad (most were born outside Europe, 68% of the mothers and 63% of the fathers).

3.2. Descriptive statistics for cognitive scores, academic performance, physical activity, and fitness

With regard to the cognitive test battery, 1082 adolescents completed all tests. Descriptive details of the cognitive test parameters, including start times, test duration, and classroom noise levels, are provided in the supporting information, S1 Table. The correlation between cognitive and academic performance was moderate r(1067)=0.59 for math and r(1065)=0.55 for language.

To assess the confounders, we compared mean differences in the dependent and independent variables between groups based on gender (Table 1), parental education, and parental country of birth (supporting information S2-S3 Tables). Girls had higher scores in all cognitive tasks (except in special updating) and grades in language (but not math) compared to boys (see Table 1). Further, adolescents with parents with long education, or both parents born in Swedish had higher cognitive scores and higher grades (supporting information S2-S3 Tables).

With regards to fitness and PA, 1020 adolescents had valid fitness measurements, while 904 had at least three days of valid accelerometer measurements. Those with valid accelerometer measurements had higher scores on the cognitive tasks (p < 0.05 for all tests) and math grade, there were no significant differences in language grade or fitness level, compared to those with less than three valid days.

Furthermore, boys spent more time in MPA and VPA and had higher fitness, compared to girls (Table 1). Additionally, adolescents to parents with long education spent more time in MPA (but similar time in VPA) and had higher fitness, compared to adolescents with parents with short education (≤12 years of education). There was no difference in PA depending on parental country of birth. For fitness, adolescents to parents with long education (>12 years of education) or both parents born in Sweden had higher fitness, compared to adolescents with parents with short education and both parents born abroad.

3.3. Associations between fitness and cognitive and academic performance

There was a significant association between fitness and cognitive performance, in the adjusted models (see Table 2). Thus, confirming our hypothesis that fitness would be significantly associated with cognitive performance (H5).

Download:

Table 2. Associations between exposures and cognitive performance scores analyzed with multi-level linear regression with schools as a cluster.

https://doi.org/10.1371/journal.pone.0344087.t002

Furthermore, a significant association was observed between fitness and academic performance (both math and language grades) in the adjusted models (Table 3). This supports our hypothesis (H6), suggesting that adolescents with higher fitness levels also achieved higher grades.

Download:

Table 3. Associations between exposures and academic performance (math or language) analyzed with multi-level linear regression with schools as a cluster.

https://doi.org/10.1371/journal.pone.0344087.t003

3.4. Associations between physical activity (MPA or VPA) and cognitive and academic performance

As seen in the adjusted models (Table 2 and 3), there were no significant associations between MPA and cognitive or academic performance (grades in math or language). Therefore, we rejected our hypothesis that MPA would be significantly associated with cognitive (H1) or academic performance (H3).

Further, there were no significant associations found between VPA and cognitive or academic performance in the adjusted models (Table 2 and 3). Therefore, we rejected our hypothesis, that VPA would be significantly associated with cognitive (H2) or academic performance (H4).

3.5. Path analyses using fitness and cognition as mediators

Fitness and cognitive performance were explored as mediators using SEM, these results are presented in Fig 2a and b.

Download:

Fig 2. The mediating role of fitness on the association between VPA and grades in math (a) or language (b), and cognitive performance (cog score) on the association between fitness and grades in math (a) or language (b).

Mediation analyses were adjusted for gender, parental education, parental country of birth, pubertal status and gender. To account for the clustering of schools the vce(cluster) option was applied in Stata. Fitness was found to mediate the association between VPA and math (a), and language (b) grade. Cognitive performance was found to mediate the association between fitness and math (a) and language (b) grade. *** p < 0.001; * < 0.05, n = 688 math model, n = 687 language model.

https://doi.org/10.1371/journal.pone.0344087.g002

We found individual cognitive performance differences to significantly mediate the association between fitness and academic performance, supporting our hypothesis (H7). In the math models, cognitive performance accounted for 40% of the association (Fig 2a). Further, in the language model, cognitive performance accounted for 40% of the association (Fig 2b). In both models, the direct association between fitness and academic performance was still significant, after accounting for the mediators and confounders.

Furthermore, we found fitness to significantly mediate the association between VPA and academic performance (Fig 2b).This finding supports our exploratory hypothesis (H8).In the math model, fitness explained 77% of the association between VPA and math, but the direct association between VPA and math was not significant after accounting for the effects of the mediators and confounders (p = 0.796). Whereas in the language model, fitness explained 32% of the association, also in this model the association between VPA and language grade was no longer significant after taking the mediators and confounders into account (p = 0.215). Estimates from the crude version of the structural model are provided in the supporting information (S7 Table).

3.6. Sensitivity analyses

Given the differences in the outcomes between those with and without valid accelerometer measurements, sensitivity analyses were conducted, incorporating all participants with at least one valid day. The overall results aligned with those obtained from the models using the more stringent criterion of three valid days. Additionally, BMI was tested as a covariate in the models, but since its inclusion did not significantly affect the results, it was not included in the final model.

Sensitivity analyses were also conducted, excluding adolescents with a late start of the cognitive test battery, as this group had lower scores. The findings were consistent with those obtained when including all participants. Lastly, sensitivity analyses were run including class noise level as a confounder, these results were also in line with those excluding this confounder.

4. Discussion

In this study, we investigated the associations between PA (MPA, VPA), fitness, and cognitive or academic performance in Swedish adolescents. We found a positive association between fitness and both cognitive and academic performance. Furthermore, cognitive performance mediated the association with academic performance. Moreover, fitness served as a mediator in the association between VPA and academic performance, although this analysis was done post hoc and the direct association between VPA and academic performance was not significant.

One unique aspect of our study is that we assessed PA, fitness, cognitive, and academic performance within the same sample of adolescents. When examining the results, fitness stands as the only significant factor, showing a positive association with both cognitive and academic performance (grades in math and language). Given the close link between cognitive and academic performance [5,8], it is reasonable to expect that robust cognitive skills would improve comprehension and problem-solving, ultimately resulting in better school grades. Therefore, we hypothesized cognitive performance to act as a mediator. This hypothesis (H7) was supported, as cognitive performance was found to have a mediating role in the association between fitness and academic performance. This result highlights that fitness might not only be an important marker of physical health but also of cognitive performance, which in turn may have implications for academic performance [37].

The mediating role of cognitive performance in the association between fitness and academic performance has been reported in previous cross-sectional studies involving older adolescents [37], and children, for instance in a study by Van der Niet et al. [38]. Van der Niet et al., suggested that such findings prompt the question of whether sufficient executive functioning is necessary before fitness can effectively enhance academic achievement [38]. However, the observational nature of these studies prevents causal claims. Intervention studies in younger children have shown that the effect of acute PA on cognitive performance varies depending on baseline cognitive performance. Specifically, those with lower cognitive performance demonstrated greater improvement, while no change was observed among those with higher baseline performance [24]. Furthermore, a school PA intervention in children found no overall effects on academic performance, yet subgroup analysis showed a positive intervention effect among those with low numeracy skills at baseline [23]. These findings highlight the need to consider cognitive performance when examining the link between fitness or PA and academic performance.

However, it is also important to acknowledge the complexity of these associations and the potential influence of other factors. According to Cognitive Load Theory, information overload can impair working memory, hindering learning, even for individuals with high cognitive abilities [39]. Effective teaching, however, can mitigate this by for instance connecting new information with prior knowledge stored in long-term memory, helping to avoid overwhelming cognitive resources and improving academic performance [39]. Thus, there is an interplay between cognitive abilities, such as working memory and long-term memory, and teaching strategies that influence academic outcomes. Furthermore, the socio-ecological framework states that individual outcomes, such as academic performance or physical activity habits are shaped by multiple, nested environmental systems [40,41]. In the academic environment, these systems range from microsystems (e.g., peer and teacher interactions) to macrosystems (e.g., educational policies). Our model, focusing solely on student-level characteristics, does not account for these broader influences or the bidirectional relationship between academic and cognitive performance. Therefore, future studies should incorporate variables from multiple ecological levels to provide a more nuanced understanding of the diverse factors influencing academic performance, while also investigating causal inferences through longitudinal or randomized controlled trial designs.

Given the numerous factors influencing cognitive and academic performance, such as age [5], parental and environmental factors [6], genetics [7], and engagement and the quality of education received [4] it is not surprising that the associations between fitness and cognitive or academic performance have been reported to be weak [4], especially after controlling for confounding variables [16]. In our study, the largest standardized beta was found in the association between fitness and academic performance, with a 0.16 increase in standardized language grade or 0.02 on the unstandardized grade variable in response to a one-unit increase in fitness. However, it is important to emphasize that the units of fitness are small (1 mL/kg/min). To illustrate, if a student with low fitness (35 ml/kg/min) were to increase fitness by 15 units, this would correspond to a 0.33 increase in language grade, ranging from 1–6. This stresses caution in attributing significant academic improvement solely to fitness enhancements.

Previous studies often lack detailed measures of PA, especially among adolescents [15]. In our study, we addressed this by utilizing accelerometer-based PA, which enabled analysis of different PA intensities (MPA and VPA). Coe et al. have suggested that MPA might not reach the intensity threshold necessary for improvements in academic performance [26]. Therefore, when MPA and VPA are combined into MVPA, such distinction could get lost. Regardless, in the current study, no significant associations between either MPA or VPA and cognitive or academic performance was found. These results contrast with some previous studies, that reported VPA (but not MPA) to be positively related to academic performance in children [26] and older adolescents, although this was only observed among girls [42]. The discrepancy between our findings and the study in older adolescents could be that the associations were gender specific, given that significant associations were only found among girls [42].

Considering that engaging in regular PA typically leads to improved fitness levels [11], it might be surprising that fitness was found to be a significant exposure (despite the weak association), whereas VPA did not. One possible reason could be that PA was measured for one week, offering a snapshot of habits, while fitness, specifically cardiorespiratory endurance, refers to “the ability of the circulatory and respiratory systems to supply fuel during sustained PA” [43], and is therefore an ability developed through consistent PA over time. Further, grades were obtained at the end of grade 6, two to six months before measuring fitness and PA, reflecting the efforts and abilities demonstrated by adolescents throughout the entirety of grade 6. This discrepancy in timing may partly explain why the findings were more conclusive for fitness.

Furthermore, it is important to acknowledge that VPA and fitness represent different constructs. VPA reflects a behavior, while fitness represents a physical ability, also influenced by genetics, especially before puberty [44]. It is plausible that children and adolescents with a genetic disposition towards high fitness, also share a predisposition to specific structural brain features that contribute to enhanced cognitive performance. For example, studies have found that children with high fitness levels had larger volumes of brain structures (such as the basal ganglia [45] and hippocampus [46] associated with higher cognitive performance [45,46]. Furthermore, it has been suggested that high-fit children [47] and adolescents [14] are able to perform better on cognitive tasks through the allocation of more attentional resources, compared to their less-fit counterparts [14,47].

Considering the link between VPA and fitness, it could be inferred that an individual’s fitness level would partially explain the association between VPA and academic performance. The exploratory hypothesis (H8) in this study, that fitness mediates the association between VPA and academic performance, was supported. Although this hypothesis was not pre-registered and therefore represents post hoc theorizing, it was informed by evidence from previous studies in children and adolescents [12,13]. The results showed that in the associations between VPA and academic performance, the mediation models showed that fitness accounted for 77% in math and 32% in language. These findings are in line with previous studies among children [48] and adolescents, although this was only found in boys, [49]. However, these studies used self-reported PA which limits the possibility of distinguishing between PA intensities. The results of the current study indicate that adolescents engaging in more VPA exhibit higher fitness levels and achieve higher grades. However, the direct association between VPA and academic performance was not significant, suggesting that VPA may only be important if considered in the context of fitness. While the cross-sectional design of this study prevents us from determining causal relationships, previous longitudinal studies have shown that students who maintain or enhance their fitness over time tend to exhibit higher academic performance compared to those with lower fitness levels [15]. Although there has been discussion regarding the potential to improve fitness before puberty, a meta-review concluded that PA interventions improved cardiorespiratory fitness among adolescents [11].

Our findings showed that fitness was significantly associated with both cognitive and academic performance, although the associations were weak. Additionally, it was observed that the associations between VPA and academic performance operated through fitness. Therefore, it is concerning that we also found that adolescents having parents with short education, or parents born abroad had lower levels of fitness, PA, cognitive, and academic performance, compared to those having parents with long education or born in Sweden. Therefore, future PA interventions should prioritize the inclusion of these groups. Moreover, to investigate whether enhancing PA and fitness levels among these groups could also improve their academic performance.

A strength of the current study is the relatively large sample, using robust estimates of PA (accelerometry), fitness, and cognitive performance (a comprehensive test battery). Another strength is that we investigated the associations between PA, fitness, cognitive, and academic performance in the same sample, which facilitated direct comparisons across various outcomes. While this is a strength, it is also a limitation as the increased number of models raises the risk of Type I errors. Therefore, we adjusted the p-values for multiple comparisons, however this may increase the risk of Type II errors. Concerning generalizability, this study included adolescents with an average age of 13 years, limiting the possibility to extend these findings to other age groups due to the rapid mental and physical development occurring during this period. A limitation of our study is the potential for selection bias, as participants may not fully represent the broader adolescent population. However, by recruiting adolescents through schools and employing an inclusive strategy, such as translating materials and creating a recruitment film, we aimed to reduce this bias, achieving a diverse sample with a 73% response rate. Regarding measurement bias, we used accelerometers to measure PA, a method considered more reliable than self-reports, as adolescents tend to overestimate their PA [20]. However, accelerometers depend on wear-time compliance, in this study, we required at least three valid days of wear, including one weekend day, as commonly used in previous research [29]. While necessary, this criterion reduced the sample size by 200–300 participants, which may have increased the risk of selection bias. This reduction in sample size likely diminished the statistical power of the study, potentially affecting the reliability of the results. However, it is important to emphasize that this criterion may have led to the non-random exclusion of less compliant or less active adolescents, potentially introduced bias to the results. To address this concern, we conducted sensitivity analyses including all participants with at least one valid day of accelerometer data. These analyses showed results consistent with the primary models. Another consideration is that the use of percentage time spent in PA, may not fully capture meaningful differences in absolute PA levels given the variability in accelerometer wear time across participants, and this should be considered when interpreting the results. Finally, it is essential to acknowledge that this study has a cross-sectional design. Although mediation analyses were conducted within a theoretical framework, causality cannot be inferred, as temporal ordering among fitness, cognition, and academic performance cannot be established. Nonetheless, similar cross-sectional studies in younger populations have used mediation approaches to explore potential pathways [12], providing valuable preliminary insights. Our analyses should therefore be interpreted as exploratory and hypothesis-generating. Longitudinal and experimental studies are needed to clarify temporal and causal relationships among these variables.

5. Conclusion

In summary, our results showed that adolescents with higher fitness levels also had higher cognitive and academic performance (grades in math and language). Moreover, cognitive performance had a mediating role in this association, while fitness acted as a mediator in the association between VPA and academic performance. The results also showed lower fitness, PA, cognitive, and academic performance levels among adolescents having parents with short education or born abroad. Therefore, these groups should be considered when planning future interventions aiming at improving academic performance.

Supporting information

S1 Fig. Overview of recruitment of participating schools and adolescents.

https://doi.org/10.1371/journal.pone.0344087.s001

(TIF)

S2 Fig. Overview of the working memory task.

https://doi.org/10.1371/journal.pone.0344087.s002

(TIF)

S3 Fig. Visual overview of the numerical 3-back task.

https://doi.org/10.1371/journal.pone.0344087.s003

(TIF)

S4 Fig. Visual overview of the spatial updating task.

https://doi.org/10.1371/journal.pone.0344087.s004

(TIF)

S5 Fig. Visual overview of object-position recall task.

https://doi.org/10.1371/journal.pone.0344087.s005

(TIF)

S6 Fig. Visual representation of the SEM measurement model of memory using standardized estimates.

https://doi.org/10.1371/journal.pone.0344087.s006

(TIF)

S1 Table. Descriptive characteristics of cognitive test parameters for the whole sample and stratified by gender (n = 1139).

https://doi.org/10.1371/journal.pone.0344087.s007

(DOCX)

S2 Table. Descriptive characteristics of the study sample by parental education group (mean ± SD unless otherwise specified).

https://doi.org/10.1371/journal.pone.0344087.s008

(DOCX)

S3 Table. Descriptive characteristics of the study sample, by parental country of birth (mean ± SD unless otherwise specified).

https://doi.org/10.1371/journal.pone.0344087.s009

(DOCX)

S4 Table. Associations between predictors and memory (episodic memory and working memory) analyzed with multi-level linear regression treating schools as a cluster.

https://doi.org/10.1371/journal.pone.0344087.s010

(DOCX)

S5 Table. Associations between predictors and overall academic performance (grades in language and math) analyzed with multi-level linear regression treating schools as a cluster.

https://doi.org/10.1371/journal.pone.0344087.s011

(DOCX)

S6 Table. Associations between MVPA and cognitive performance (working memory and episodic memory) and academic performance (grade in math and language) analyzed with multi-level linear regression treating schools as a cluster.

https://doi.org/10.1371/journal.pone.0344087.s012

(DOCX)

S7 Table. Crude structural model showing direct and indirect associations between VPA, fitness, cognitive performance, and academic performance (grades in math, n = 729 and language n = 755) using fitness and cognitive performance as mediators.

https://doi.org/10.1371/journal.pone.0344087.s013

(DOCX)

Acknowledgments

We would like to thank all the participating adolescents and schools for making this study possible.

References

1. UNICEF. The Adolescent Brain: A second window of opportunity - A compendium. UNICEF; 2017.
2. Lubans DR, Leahy AA, Mavilidi MF, Valkenborghs SR. Physical activity, fitness, and executive functions in youth: Effects, moderators, and mechanisms. Sensitive Periods of Brain Development and Preventive Interventions. Springer. 2021. pp. 103–30.
3. Nyberg G, Kjellenberg K, Fröberg A, Lindroos AK. A national survey showed low levels of physical activity in a representative sample of Swedish adolescents. Acta Paediatr. 2020;109(11):2342–53. pmid:32266736
- View Article
- PubMed/NCBI
- Google Scholar
4. Keeley TJH, Fox KR. The impact of physical activity and fitness on academic achievement and cognitive performance in children. Int Rev Sport Exerc Psychol. 2009;2(2):198–214.
- View Article
- Google Scholar
5. Diamond A. Executive functions. Annu Rev Psychol. 2013;64:135–68. pmid:23020641
- View Article
- PubMed/NCBI
- Google Scholar
6. Zhang H, Lee ZX, White T, Qiu A. Parental and social factors in relation to child psychopathology, behavior, and cognitive function. Transl Psychiatry. 2020;10(1):80. pmid:32102994
- View Article
- PubMed/NCBI
- Google Scholar
7. Haworth CMA, Wright MJ, Luciano M, Martin NG, de Geus EJC, van Beijsterveldt CEM, et al. The heritability of general cognitive ability increases linearly from childhood to young adulthood. Mol Psychiatry. 2010;15(11):1112–20. pmid:19488046
- View Article
- PubMed/NCBI
- Google Scholar
8. Ghetti S, Bunge SA. Neural changes underlying the development of episodic memory during middle childhood. Dev Cogn Neurosci. 2012;2(4):381–95. pmid:22770728
- View Article
- PubMed/NCBI
- Google Scholar
9. Stillman CM, Cohen J, Lehman ME, Erickson KI. Mediators of physical activity on neurocognitive function: a review at multiple levels of analysis. Front Hum Neurosci. 2016;10:626. pmid:28018195
- View Article
- PubMed/NCBI
- Google Scholar
10. Krafft CE, Schaeffer DJ, Schwarz NF, Chi L, Weinberger AL, Pierce JE, et al. Improved frontoparietal white matter integrity in overweight children is associated with attendance at an after-school exercise program. Dev Neurosci. 2014;36(1):1–9. pmid:24457421
- View Article
- PubMed/NCBI
- Google Scholar
11. Minatto G, Barbosa Filho VC, Berria J, Petroski EL. School-based interventions to improve cardiorespiratory fitness in adolescents: systematic review with meta-analysis. Sports Med. 2016;46(9):1273–92. pmid:26846429
- View Article
- PubMed/NCBI
- Google Scholar
12. Visier-Alfonso ME, Sánchez-López M, Martínez-Vizcaíno V, Jiménez-López E, Redondo-Tébar A, Nieto-López M. Executive functions mediate the relationship between cardiorespiratory fitness and academic achievement in Spanish schoolchildren aged 8 to 11 years. PLoS One. 2020;15(4):e0231246. pmid:32275676
- View Article
- PubMed/NCBI
- Google Scholar
13. Kvalø SE, Dyrstad SM, Bru E, Brønnick K. Relationship between aerobic fitness and academic performance: the mediational role of executive function. J Sports Med Phys Fitness. 2019;59(8):1397–404. pmid:30722648
- View Article
- PubMed/NCBI
- Google Scholar
14. Stroth S, Kubesch S, Dieterle K, Ruchsow M, Heim R, Kiefer M. Physical fitness, but not acute exercise modulates event-related potential indices for executive control in healthy adolescents. Brain Res. 2009;1269:114–24. pmid:19285042
- View Article
- PubMed/NCBI
- Google Scholar
15. Donnelly JE, Hillman CH, Castelli D, Etnier JL, Lee S, Tomporowski P, et al. Physical activity, fitness, cognitive function, and academic achievement in children: a systematic review. Med Sci Sports Exerc. 2016;48(6):1197–222. pmid:27182986
- View Article
- PubMed/NCBI
- Google Scholar
16. Santana CCA, Azevedo LB, Cattuzzo MT, Hill JO, Andrade LP, Prado WL. Physical fitness and academic performance in youth: a systematic review. Scand J Med Sci Sports. 2017;27(6):579–603. pmid:27714852
- View Article
- PubMed/NCBI
- Google Scholar
17. Goswami B, Roy AS, Dalui R, Bandyopadhyay A. Impact of pubertal growth on physical fitness. Am J Sport Sci Med. 2014;2(5A):34–9.
- View Article
- Google Scholar
18. Singh AS, Saliasi E, van den Berg V, Uijtdewilligen L, de Groot RHM, Jolles J, et al. Effects of physical activity interventions on cognitive and academic performance in children and adolescents: a novel combination of a systematic review and recommendations from an expert panel. Br J Sports Med. 2019;53(10):640–7. pmid:30061304
- View Article
- PubMed/NCBI
- Google Scholar
19. Marques A, Santos DA, Hillman CH, Sardinha LB. How does academic achievement relate to cardiorespiratory fitness, self-reported physical activity and objectively reported physical activity: a systematic review in children and adolescents aged 6-18 years. Br J Sports Med. 2018;52(16):1039. pmid:29032365
- View Article
- PubMed/NCBI
- Google Scholar
20. Slootmaker SM, Schuit AJ, Chinapaw MJ, Seidell JC, van Mechelen W. Disagreement in physical activity assessed by accelerometer and self-report in subgroups of age, gender, education and weight status. Int J Behav Nutr Phys Act. 2009;6:17. pmid:19320985
- View Article
- PubMed/NCBI
- Google Scholar
21. Voss MW, Nagamatsu LS, Liu-Ambrose T, Kramer AF. Exercise, brain, and cognition across the life span. J Appl Physiol (1985). 2011;111(5):1505–13. pmid:21527670
- View Article
- PubMed/NCBI
- Google Scholar
22. Liu S, Yu Q, Li Z, Cunha PM, Zhang Y, Kong Z, et al. Effects of acute and chronic exercises on executive function in children and adolescents: a systemic review and meta-analysis. Front Psychol. 2020;11:554915. pmid:33391074
- View Article
- PubMed/NCBI
- Google Scholar
23. Resaland GK, Aadland E, Moe VF, Aadland KN, Skrede T, Stavnsbo M, et al. Effects of physical activity on schoolchildren’s academic performance: The Active Smarter Kids (ASK) cluster-randomized controlled trial. Prev Med. 2016;91:322–8. pmid:27612574
- View Article
- PubMed/NCBI
- Google Scholar
24. Drollette ES, Scudder MR, Raine LB, Moore RD, Saliba BJ, Pontifex MB, et al. Acute exercise facilitates brain function and cognition in children who need it most: an ERP study of individual differences in inhibitory control capacity. Dev Cogn Neurosci. 2014;7:53–64. pmid:24309300
- View Article
- PubMed/NCBI
- Google Scholar
25. Ng QX, Ho CYX, Chan HW, Yong BZJ, Yeo W-S. Managing childhood and adolescent attention-deficit/hyperactivity disorder (ADHD) with exercise: a systematic review. Complement Ther Med. 2017;34:123–8. pmid:28917364
- View Article
- PubMed/NCBI
- Google Scholar
26. Coe DP, Pivarnik JM, Womack CJ, Reeves MJ, Malina RM. Effect of physical education and activity levels on academic achievement in children. Med Sci Sports Exerc. 2006;38(8):1515–9. pmid:16888468
- View Article
- PubMed/NCBI
- Google Scholar
27. Nyberg G, Ekblom Ö, Kjellenberg K, Wang R, Larsson H, Thedin Jakobsson B, et al. Associations between the school environment and physical activity pattern during school time in Swedish adolescents. Int J Environ Res Public Health. 2021;18(19):10239. pmid:34639539
- View Article
- PubMed/NCBI
- Google Scholar
28. Nevalainen N, Riklund K, Andersson M, Axelsson J, Ögren M, Lövdén M, et al. COBRA: a prospective multimodal imaging study of dopamine, brain structure and function, and cognition. Brain Res. 2015;1612:83–103. pmid:25239478
- View Article
- PubMed/NCBI
- Google Scholar
29. Cain KL, Sallis JF, Conway TL, Van Dyck D, Calhoon L. Using accelerometers in youth physical activity studies: a review of methods. J Phys Act Health. 2013;10(3):437–50. pmid:23620392
- View Article
- PubMed/NCBI
- Google Scholar
30. Evenson KR, Catellier DJ, Gill K, Ondrak KS, McMurray RG. Calibration of two objective measures of physical activity for children. J Sports Sci. 2008;26(14):1557–65. pmid:18949660
- View Article
- PubMed/NCBI
- Google Scholar
31. Björkman F, Eggers A, Stenman A, Bohman T, Ekblom B, Ekblom Ö. Sex and maturity status affected the validity of a submaximal cycle test in adolescents. Acta Paediatr. 2018;107(1):126–33.
- View Article
- Google Scholar
32. Cole TJ, Lobstein T. Extended international (IOTF) body mass index cut-offs for thinness, overweight and obesity. Pediatr Obes. 2012;7(4):284–94. pmid:22715120
- View Article
- PubMed/NCBI
- Google Scholar
33. Tanner JM. Growth at adolescence: With a general consideration of the effects of hereditary and environmental factors upon growth and maturation from birth to maturity. Oxford: Blackwell Scientific Publications; 1962.
34. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc B. 1995;57(1):289–300.
- View Article
- Google Scholar
35. Hoaglin DC, Iglewicz B. Fine-tuning some resistant rules for outlier labeling. J Am Stat Assoc. 1987;82(400):1147–9.
- View Article
- Google Scholar
36. Mehmetoglu M. Medsem: A Stata package for statistical mediation analysis. 2018.
37. Park S, Chun H, Etnier JL, Yun D. Exploring the mediating role of executive function in the relationship between aerobic fitness and academic achievement in adolescents. Brain Sci. 2023;13(4):614. pmid:37190579
- View Article
- PubMed/NCBI
- Google Scholar
38. van der Niet AG, Hartman E, Smith J, Visscher C. Modeling relationships between physical fitness, executive functioning, and academic achievement in primary school children. Psychol Sport Exerc. 2014;15(4):319–25.
- View Article
- Google Scholar
39. Chandler P, Sweller J. Cognitive load theory and the format of instruction. Cogn Instr. 1991;8(4):293–332.
- View Article
- Google Scholar
40. Bronfenbrenner U. The ecology of human development: Experiments by nature and design. Harvard University Press; 1979.
41. Trinh L, Wong B, Faulkner GE. The Independent and Interactive Associations of Screen Time and Physical Activity on Mental Health, School Connectedness and Academic Achievement among a Population-Based Sample of Youth. J Can Acad Child Adolesc Psychiatry. 2015;24(1):17–24. pmid:26336376
- View Article
- PubMed/NCBI
- Google Scholar
42. Kwak L, Kremers SPJ, Bergman P, Ruiz JR, Rizzo NS, Sjöström M. Associations between physical activity, fitness, and academic achievement. J Pediatr. 2009;155(6):914–918.e1. pmid:19643438
- View Article
- PubMed/NCBI
- Google Scholar
43. Caspersen CJ, Powell KE, Christenson GM. Physical activity, exercise, and physical fitness: definitions and distinctions for health-related research. Public Health Rep. 1985;100(2):126–31. pmid:3920711
- View Article
- PubMed/NCBI
- Google Scholar
44. Schutte NM, Nederend I, Hudziak JJ, Bartels M, de Geus EJC. Twin-sibling study and meta-analysis on the heritability of maximal oxygen consumption. Physiol Genomics. 2016;48(3):210–9. pmid:26787216
- View Article
- PubMed/NCBI
- Google Scholar
45. Chaddock L, Erickson KI, Prakash RS, VanPatter M, Voss MW, Pontifex MB, et al. Basal ganglia volume is associated with aerobic fitness in preadolescent children. Dev Neurosci. 2010;32(3):249–56. pmid:20693803
- View Article
- PubMed/NCBI
- Google Scholar
46. Chaddock L, Erickson KI, Prakash RS, Kim JS, Voss MW, Vanpatter M, et al. A neuroimaging investigation of the association between aerobic fitness, hippocampal volume, and memory performance in preadolescent children. Brain Res. 2010;1358:172–83. pmid:20735996
- View Article
- PubMed/NCBI
- Google Scholar
47. Chaddock L, Erickson KI, Prakash RS, Voss MW, VanPatter M, Pontifex MB, et al. A functional MRI investigation of the association between childhood aerobic fitness and neurocognitive control. Biol Psychol. 2012;89(1):260–8. pmid:22061423
- View Article
- PubMed/NCBI
- Google Scholar
48. Muntaner-Mas A, Mazzoli E, Abbott G, Mavilidi MF, Galmes-Panades AM. Do Physical Fitness and Executive Function Mediate the Relationship between Physical Activity and Academic Achievement? An Examination Using Structural Equation Modelling. Children (Basel). 2022;9(6):823. pmid:35740760
- View Article
- PubMed/NCBI
- Google Scholar
49. Kyan A, Takakura M, Miyagi M. Mediating effect of aerobic fitness on the association between physical activity and academic achievement among adolescents: A cross-sectional study in Okinawa, Japan. J Sports Sci. 2019;37(11):1242–9. pmid:30547722
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. UNICEF. The Adolescent Brain: A second window of opportunity - A compendium. UNICEF; 2017.

[ref2] 2. Lubans DR, Leahy AA, Mavilidi MF, Valkenborghs SR. Physical activity, fitness, and executive functions in youth: Effects, moderators, and mechanisms. Sensitive Periods of Brain Development and Preventive Interventions. Springer. 2021. pp. 103–30.

[ref3] 3. Nyberg G, Kjellenberg K, Fröberg A, Lindroos AK. A national survey showed low levels of physical activity in a representative sample of Swedish adolescents. Acta Paediatr. 2020;109(11):2342–53. pmid:32266736
View Article
PubMed/NCBI
Google Scholar

[4] View Article

[5] PubMed/NCBI

[6] Google Scholar

[ref4] 4. Keeley TJH, Fox KR. The impact of physical activity and fitness on academic achievement and cognitive performance in children. Int Rev Sport Exerc Psychol. 2009;2(2):198–214.
View Article
Google Scholar

[8] View Article

[9] Google Scholar

[ref5] 5. Diamond A. Executive functions. Annu Rev Psychol. 2013;64:135–68. pmid:23020641
View Article
PubMed/NCBI
Google Scholar

[11] View Article

[12] PubMed/NCBI

[13] Google Scholar

[ref6] 6. Zhang H, Lee ZX, White T, Qiu A. Parental and social factors in relation to child psychopathology, behavior, and cognitive function. Transl Psychiatry. 2020;10(1):80. pmid:32102994
View Article
PubMed/NCBI
Google Scholar

[15] View Article

[16] PubMed/NCBI

[17] Google Scholar

[ref7] 7. Haworth CMA, Wright MJ, Luciano M, Martin NG, de Geus EJC, van Beijsterveldt CEM, et al. The heritability of general cognitive ability increases linearly from childhood to young adulthood. Mol Psychiatry. 2010;15(11):1112–20. pmid:19488046
View Article
PubMed/NCBI
Google Scholar

[19] View Article

[20] PubMed/NCBI

[21] Google Scholar

[ref8] 8. Ghetti S, Bunge SA. Neural changes underlying the development of episodic memory during middle childhood. Dev Cogn Neurosci. 2012;2(4):381–95. pmid:22770728
View Article
PubMed/NCBI
Google Scholar

[23] View Article

[24] PubMed/NCBI

[25] Google Scholar

[ref9] 9. Stillman CM, Cohen J, Lehman ME, Erickson KI. Mediators of physical activity on neurocognitive function: a review at multiple levels of analysis. Front Hum Neurosci. 2016;10:626. pmid:28018195
View Article
PubMed/NCBI
Google Scholar

[27] View Article

[28] PubMed/NCBI

[29] Google Scholar

[ref10] 10. Krafft CE, Schaeffer DJ, Schwarz NF, Chi L, Weinberger AL, Pierce JE, et al. Improved frontoparietal white matter integrity in overweight children is associated with attendance at an after-school exercise program. Dev Neurosci. 2014;36(1):1–9. pmid:24457421
View Article
PubMed/NCBI
Google Scholar

[31] View Article

[32] PubMed/NCBI

[33] Google Scholar

[ref11] 11. Minatto G, Barbosa Filho VC, Berria J, Petroski EL. School-based interventions to improve cardiorespiratory fitness in adolescents: systematic review with meta-analysis. Sports Med. 2016;46(9):1273–92. pmid:26846429
View Article
PubMed/NCBI
Google Scholar

[35] View Article

[36] PubMed/NCBI

[37] Google Scholar

[ref12] 12. Visier-Alfonso ME, Sánchez-López M, Martínez-Vizcaíno V, Jiménez-López E, Redondo-Tébar A, Nieto-López M. Executive functions mediate the relationship between cardiorespiratory fitness and academic achievement in Spanish schoolchildren aged 8 to 11 years. PLoS One. 2020;15(4):e0231246. pmid:32275676
View Article
PubMed/NCBI
Google Scholar

[39] View Article

[40] PubMed/NCBI

[41] Google Scholar

[ref13] 13. Kvalø SE, Dyrstad SM, Bru E, Brønnick K. Relationship between aerobic fitness and academic performance: the mediational role of executive function. J Sports Med Phys Fitness. 2019;59(8):1397–404. pmid:30722648
View Article
PubMed/NCBI
Google Scholar

[43] View Article

[44] PubMed/NCBI

[45] Google Scholar

[ref14] 14. Stroth S, Kubesch S, Dieterle K, Ruchsow M, Heim R, Kiefer M. Physical fitness, but not acute exercise modulates event-related potential indices for executive control in healthy adolescents. Brain Res. 2009;1269:114–24. pmid:19285042
View Article
PubMed/NCBI
Google Scholar

[47] View Article

[48] PubMed/NCBI

[49] Google Scholar

[ref15] 15. Donnelly JE, Hillman CH, Castelli D, Etnier JL, Lee S, Tomporowski P, et al. Physical activity, fitness, cognitive function, and academic achievement in children: a systematic review. Med Sci Sports Exerc. 2016;48(6):1197–222. pmid:27182986
View Article
PubMed/NCBI
Google Scholar

[51] View Article

[52] PubMed/NCBI

[53] Google Scholar

[ref16] 16. Santana CCA, Azevedo LB, Cattuzzo MT, Hill JO, Andrade LP, Prado WL. Physical fitness and academic performance in youth: a systematic review. Scand J Med Sci Sports. 2017;27(6):579–603. pmid:27714852
View Article
PubMed/NCBI
Google Scholar

[55] View Article

[56] PubMed/NCBI

[57] Google Scholar

[ref17] 17. Goswami B, Roy AS, Dalui R, Bandyopadhyay A. Impact of pubertal growth on physical fitness. Am J Sport Sci Med. 2014;2(5A):34–9.
View Article
Google Scholar

[59] View Article

[60] Google Scholar

[ref18] 18. Singh AS, Saliasi E, van den Berg V, Uijtdewilligen L, de Groot RHM, Jolles J, et al. Effects of physical activity interventions on cognitive and academic performance in children and adolescents: a novel combination of a systematic review and recommendations from an expert panel. Br J Sports Med. 2019;53(10):640–7. pmid:30061304
View Article
PubMed/NCBI
Google Scholar

[62] View Article

[63] PubMed/NCBI

[64] Google Scholar

[ref19] 19. Marques A, Santos DA, Hillman CH, Sardinha LB. How does academic achievement relate to cardiorespiratory fitness, self-reported physical activity and objectively reported physical activity: a systematic review in children and adolescents aged 6-18 years. Br J Sports Med. 2018;52(16):1039. pmid:29032365
View Article
PubMed/NCBI
Google Scholar

[66] View Article

[67] PubMed/NCBI

[68] Google Scholar

[ref20] 20. Slootmaker SM, Schuit AJ, Chinapaw MJ, Seidell JC, van Mechelen W. Disagreement in physical activity assessed by accelerometer and self-report in subgroups of age, gender, education and weight status. Int J Behav Nutr Phys Act. 2009;6:17. pmid:19320985
View Article
PubMed/NCBI
Google Scholar

[70] View Article

[71] PubMed/NCBI

[72] Google Scholar

[ref21] 21. Voss MW, Nagamatsu LS, Liu-Ambrose T, Kramer AF. Exercise, brain, and cognition across the life span. J Appl Physiol (1985). 2011;111(5):1505–13. pmid:21527670
View Article
PubMed/NCBI
Google Scholar

[74] View Article

[75] PubMed/NCBI

[76] Google Scholar

[ref22] 22. Liu S, Yu Q, Li Z, Cunha PM, Zhang Y, Kong Z, et al. Effects of acute and chronic exercises on executive function in children and adolescents: a systemic review and meta-analysis. Front Psychol. 2020;11:554915. pmid:33391074
View Article
PubMed/NCBI
Google Scholar

[78] View Article

[79] PubMed/NCBI

[80] Google Scholar

[ref23] 23. Resaland GK, Aadland E, Moe VF, Aadland KN, Skrede T, Stavnsbo M, et al. Effects of physical activity on schoolchildren’s academic performance: The Active Smarter Kids (ASK) cluster-randomized controlled trial. Prev Med. 2016;91:322–8. pmid:27612574
View Article
PubMed/NCBI
Google Scholar

[82] View Article

[83] PubMed/NCBI

[84] Google Scholar

[ref24] 24. Drollette ES, Scudder MR, Raine LB, Moore RD, Saliba BJ, Pontifex MB, et al. Acute exercise facilitates brain function and cognition in children who need it most: an ERP study of individual differences in inhibitory control capacity. Dev Cogn Neurosci. 2014;7:53–64. pmid:24309300
View Article
PubMed/NCBI
Google Scholar

[86] View Article

[87] PubMed/NCBI

[88] Google Scholar

[ref25] 25. Ng QX, Ho CYX, Chan HW, Yong BZJ, Yeo W-S. Managing childhood and adolescent attention-deficit/hyperactivity disorder (ADHD) with exercise: a systematic review. Complement Ther Med. 2017;34:123–8. pmid:28917364
View Article
PubMed/NCBI
Google Scholar

[90] View Article

[91] PubMed/NCBI

[92] Google Scholar

[ref26] 26. Coe DP, Pivarnik JM, Womack CJ, Reeves MJ, Malina RM. Effect of physical education and activity levels on academic achievement in children. Med Sci Sports Exerc. 2006;38(8):1515–9. pmid:16888468
View Article
PubMed/NCBI
Google Scholar

[94] View Article

[95] PubMed/NCBI

[96] Google Scholar

[ref27] 27. Nyberg G, Ekblom Ö, Kjellenberg K, Wang R, Larsson H, Thedin Jakobsson B, et al. Associations between the school environment and physical activity pattern during school time in Swedish adolescents. Int J Environ Res Public Health. 2021;18(19):10239. pmid:34639539
View Article
PubMed/NCBI
Google Scholar

[98] View Article

[99] PubMed/NCBI

[100] Google Scholar

[ref28] 28. Nevalainen N, Riklund K, Andersson M, Axelsson J, Ögren M, Lövdén M, et al. COBRA: a prospective multimodal imaging study of dopamine, brain structure and function, and cognition. Brain Res. 2015;1612:83–103. pmid:25239478
View Article
PubMed/NCBI
Google Scholar

[102] View Article

[103] PubMed/NCBI

[104] Google Scholar

[ref29] 29. Cain KL, Sallis JF, Conway TL, Van Dyck D, Calhoon L. Using accelerometers in youth physical activity studies: a review of methods. J Phys Act Health. 2013;10(3):437–50. pmid:23620392
View Article
PubMed/NCBI
Google Scholar

[106] View Article

[107] PubMed/NCBI

[108] Google Scholar

[ref30] 30. Evenson KR, Catellier DJ, Gill K, Ondrak KS, McMurray RG. Calibration of two objective measures of physical activity for children. J Sports Sci. 2008;26(14):1557–65. pmid:18949660
View Article
PubMed/NCBI
Google Scholar

[110] View Article

[111] PubMed/NCBI

[112] Google Scholar

[ref31] 31. Björkman F, Eggers A, Stenman A, Bohman T, Ekblom B, Ekblom Ö. Sex and maturity status affected the validity of a submaximal cycle test in adolescents. Acta Paediatr. 2018;107(1):126–33.
View Article
Google Scholar

[114] View Article

[115] Google Scholar

[ref32] 32. Cole TJ, Lobstein T. Extended international (IOTF) body mass index cut-offs for thinness, overweight and obesity. Pediatr Obes. 2012;7(4):284–94. pmid:22715120
View Article
PubMed/NCBI
Google Scholar

[117] View Article

[118] PubMed/NCBI

[119] Google Scholar

[ref33] 33. Tanner JM. Growth at adolescence: With a general consideration of the effects of hereditary and environmental factors upon growth and maturation from birth to maturity. Oxford: Blackwell Scientific Publications; 1962.

[ref34] 34. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc B. 1995;57(1):289–300.
View Article
Google Scholar

[122] View Article

[123] Google Scholar

[ref35] 35. Hoaglin DC, Iglewicz B. Fine-tuning some resistant rules for outlier labeling. J Am Stat Assoc. 1987;82(400):1147–9.
View Article
Google Scholar

[125] View Article

[126] Google Scholar

[ref36] 36. Mehmetoglu M. Medsem: A Stata package for statistical mediation analysis. 2018.

[ref37] 37. Park S, Chun H, Etnier JL, Yun D. Exploring the mediating role of executive function in the relationship between aerobic fitness and academic achievement in adolescents. Brain Sci. 2023;13(4):614. pmid:37190579
View Article
PubMed/NCBI
Google Scholar

[129] View Article

[130] PubMed/NCBI

[131] Google Scholar

[ref38] 38. van der Niet AG, Hartman E, Smith J, Visscher C. Modeling relationships between physical fitness, executive functioning, and academic achievement in primary school children. Psychol Sport Exerc. 2014;15(4):319–25.
View Article
Google Scholar

[133] View Article

[134] Google Scholar

[ref39] 39. Chandler P, Sweller J. Cognitive load theory and the format of instruction. Cogn Instr. 1991;8(4):293–332.
View Article
Google Scholar

[136] View Article

[137] Google Scholar

[ref40] 40. Bronfenbrenner U. The ecology of human development: Experiments by nature and design. Harvard University Press; 1979.

[ref41] 41. Trinh L, Wong B, Faulkner GE. The Independent and Interactive Associations of Screen Time and Physical Activity on Mental Health, School Connectedness and Academic Achievement among a Population-Based Sample of Youth. J Can Acad Child Adolesc Psychiatry. 2015;24(1):17–24. pmid:26336376
View Article
PubMed/NCBI
Google Scholar

[140] View Article

[141] PubMed/NCBI

[142] Google Scholar

[ref42] 42. Kwak L, Kremers SPJ, Bergman P, Ruiz JR, Rizzo NS, Sjöström M. Associations between physical activity, fitness, and academic achievement. J Pediatr. 2009;155(6):914–918.e1. pmid:19643438
View Article
PubMed/NCBI
Google Scholar

[144] View Article

[145] PubMed/NCBI

[146] Google Scholar

[ref43] 43. Caspersen CJ, Powell KE, Christenson GM. Physical activity, exercise, and physical fitness: definitions and distinctions for health-related research. Public Health Rep. 1985;100(2):126–31. pmid:3920711
View Article
PubMed/NCBI
Google Scholar

[148] View Article

[149] PubMed/NCBI

[150] Google Scholar

[ref44] 44. Schutte NM, Nederend I, Hudziak JJ, Bartels M, de Geus EJC. Twin-sibling study and meta-analysis on the heritability of maximal oxygen consumption. Physiol Genomics. 2016;48(3):210–9. pmid:26787216
View Article
PubMed/NCBI
Google Scholar

[152] View Article

[153] PubMed/NCBI

[154] Google Scholar

[ref45] 45. Chaddock L, Erickson KI, Prakash RS, VanPatter M, Voss MW, Pontifex MB, et al. Basal ganglia volume is associated with aerobic fitness in preadolescent children. Dev Neurosci. 2010;32(3):249–56. pmid:20693803
View Article
PubMed/NCBI
Google Scholar

[156] View Article

[157] PubMed/NCBI

[158] Google Scholar

[ref46] 46. Chaddock L, Erickson KI, Prakash RS, Kim JS, Voss MW, Vanpatter M, et al. A neuroimaging investigation of the association between aerobic fitness, hippocampal volume, and memory performance in preadolescent children. Brain Res. 2010;1358:172–83. pmid:20735996
View Article
PubMed/NCBI
Google Scholar

[160] View Article

[161] PubMed/NCBI

[162] Google Scholar

[ref47] 47. Chaddock L, Erickson KI, Prakash RS, Voss MW, VanPatter M, Pontifex MB, et al. A functional MRI investigation of the association between childhood aerobic fitness and neurocognitive control. Biol Psychol. 2012;89(1):260–8. pmid:22061423
View Article
PubMed/NCBI
Google Scholar

[164] View Article

[165] PubMed/NCBI

[166] Google Scholar

[ref48] 48. Muntaner-Mas A, Mazzoli E, Abbott G, Mavilidi MF, Galmes-Panades AM. Do Physical Fitness and Executive Function Mediate the Relationship between Physical Activity and Academic Achievement? An Examination Using Structural Equation Modelling. Children (Basel). 2022;9(6):823. pmid:35740760
View Article
PubMed/NCBI
Google Scholar

[168] View Article

[169] PubMed/NCBI

[170] Google Scholar

[ref49] 49. Kyan A, Takakura M, Miyagi M. Mediating effect of aerobic fitness on the association between physical activity and academic achievement among adolescents: A cross-sectional study in Okinawa, Japan. J Sports Sci. 2019;37(11):1242–9. pmid:30547722
View Article
PubMed/NCBI
Google Scholar

[172] View Article

[173] PubMed/NCBI

[174] Google Scholar

Figures

Abstract

Background

Methods

Results

Conclusion

1. Introduction

2. Methods

2.1. Sample and study design

2.2. Protocol

2.3. Measurements

2.3.1. Cognitive performance (working memory and episodic memory).

2.3.2. Academic performance (grades in math and language).

2.3.2. Physical activity (PA).

2.3.4. Cardiovascular fitness.

2.3.5. Other variables.

2.4. Statistical analysis

3. Results

3.1. Participants

3.2. Descriptive statistics for cognitive scores, academic performance, physical activity, and fitness

3.3. Associations between fitness and cognitive and academic performance

3.4. Associations between physical activity (MPA or VPA) and cognitive and academic performance

3.5. Path analyses using fitness and cognition as mediators

3.6. Sensitivity analyses

4. Discussion

5. Conclusion

Supporting information

S1 Fig. Overview of recruitment of participating schools and adolescents.

S2 Fig. Overview of the working memory task.

S3 Fig. Visual overview of the numerical 3-back task.

S4 Fig. Visual overview of the spatial updating task.

S5 Fig. Visual overview of object-position recall task.

S6 Fig. Visual representation of the SEM measurement model of memory using standardized estimates.

S1 Table. Descriptive characteristics of cognitive test parameters for the whole sample and stratified by gender (n = 1139).

S2 Table. Descriptive characteristics of the study sample by parental education group (mean ± SD unless otherwise specified).

S3 Table. Descriptive characteristics of the study sample, by parental country of birth (mean ± SD unless otherwise specified).

S4 Table. Associations between predictors and memory (episodic memory and working memory) analyzed with multi-level linear regression treating schools as a cluster.

S5 Table. Associations between predictors and overall academic performance (grades in language and math) analyzed with multi-level linear regression treating schools as a cluster.

S6 Table. Associations between MVPA and cognitive performance (working memory and episodic memory) and academic performance (grade in math and language) analyzed with multi-level linear regression treating schools as a cluster.

S7 Table. Crude structural model showing direct and indirect associations between VPA, fitness, cognitive performance, and academic performance (grades in math, n = 729 and language n = 755) using fitness and cognitive performance as mediators.

Acknowledgments

References