1. Introduction
Adolescence is a critical period of rapid neurological, cognitive, and behavioral development, during which lifestyle behaviors can have lasting impacts on physical, mental, and cognitive health outcomes (Skandsen et al., 2023). During this developmental phase, adolescents establish routines related to physical activity, sleep, and eating, which are essential in shaping long-term health trajectories (Cherewick et al., 2024; Spear, 2015). Simultaneously, the rise of digital technologies has created new behavioral contexts, introducing both opportunities and risks. While online engagement supports education, social interaction, and entertainment, excessive or poorly regulated internet use, termed problematic internet use (PIU), can disrupt academic performance, social relationships, daily functioning, and psychological health (Jelenchick & Christakis, 2014; Lopez-Fernandez & Kuss, 2020). Recent Indonesian data suggest that adolescents spend on average 5–8 hours per day online, with nearly 19% meeting criteria for potential PIU (Ramdhani Dwi Cesario, 2025), highlighting the need to better understand the interplay between digital behaviors and other lifestyle factors in shaping adolescent development.
Socioeconomic status (SES) is a key contextual factor influencing adolescents’ lifestyle behaviors and cognitive development. Previous research consistently shows that parental education, household income, and perceived financial resources are associated with opportunities for sport participation, diet quality, sleep conditions, and digital media use (Law & Singh, 2014; Rachmi et al., 2021). Adolescents from higher SES backgrounds are more likely to engage in structured extracurricular activities and report healthier behavioral patterns, while those from lower SES contexts often face limited access to sport facilities and health-promoting resources. SES is also linked to cognitive outcomes through differences in nutrition, educational stimulation, and home learning environments (Firk et al., 2018; Lestari et al., 2024; Zou et al., 2018). In Indonesia, socioeconomic disparities remain evident and may contribute to inequalities in both health behaviors and cognitive development during adolescence. Therefore, SES is an essential factor to consider when examining associations between extracurricular sport participation, lifestyle behaviors, and cognitive ability.
Physical activity (PA) is widely recognized as a foundational factor supporting physiological, cognitive, and psychosocial development during adolescence. Defined as any bodily movement produced by skeletal muscles that requires energy expenditure, PA contributes not only to physical fitness but also to enhanced cognitive performance, including improved in attention, executive function, working memory, and self-regulation (González-Pérez et al., 2023; Okely et al., 2001; Ramos-Campo & Clemente-Suárez, 2024). Participation in structured sport programs (extracurricular sport activity) may provide additional psychosocial benefits, such as discipline, teamwork, and academic engagement; however, this represents a specific context within the broader construct of physical activity rather than being synonymous with it. Despite these advantages, evidence from Indonesia indicates that a large proportion of children and adolescents fail to meet recommended PA levels. According to the (WHO, 2024), only 42% of children aged 10–14 years and 49.6% of adolescents aged 15–19 years meet the recommended daily activity guidelines. Similar findings are reported in national surveys, which highlight low engagement across multiple domains of activity, including organized sports, active play, and general physical movement (Andriyani et al., 2020; Mahendra, 2022; Setyaningsih et al., 2021). This insufficient engagement may limit the cognitive and social benefits of PA, while also interacting with other lifestyle behaviors such as sleep, eating, and internet use, creating compounding effects on adolescent development.
Sleep quality (SQ) is another crucial determinant of cognitive and overall adolescent well-being. Adequate sleep supports physiological restoration, memory consolidation, emotional regulation, and attention, whereas insufficient or irregular sleep is associated with impaired cognitive performance, mood disturbances, and diminished academic outcomes (Goldstone et al., 2020). In Indonesia, more than 60% of adolescents report insufficient or irregular sleep patterns, with nighttime screen exposure and digital media use as key contributing factors (Kementrian Kesehatan, 2023). Sleep behaviors, including duration, continuity, efficiency, timing, and subjective quality, interact closely with other lifestyle factors such as physical activity, eating habits, and internet engagement (Şimşek & Tekgül, 2019; van den Eijnden et al., 2021).
Eating habits (EH), including dietary patterns, meal frequency, and nutritional adequacy, play a fundamental role in supporting energy levels, cognitive function, and emotional regulation during adolescence. Healthy eating is associated with improved attention, working memory, and executive function, whereas poor diet quality particularly diets high in ultra-processed foods and low in fruits and vegetables can impair cognitive development and amplify behavioral difficulties (Rachmi et al., 2021). In Indonesia, shifting dietary patterns toward processed foods and reduced nutrient-dense intake have been observed among adolescents, raising concerns about long-term cognitive and health outcomes. Poor nutrition during adolescence can compound the effects of low PA and poor sleep, increasing vulnerability to PIU and other maladaptive behaviors, thereby creating a complex web of interrelated lifestyle factors affecting cognitive ability.
Cognitive ability (CA) refers to an individual’s capacity to acquire knowledge, process information, solve problems, and adaptively apply reasoning to novel situations (Chen et al., 2018; Farzanegan, 2020). Adolescence is characterized by rapid development in executive function, working memory, attention, inhibitory control, and fluid intelligence, which are critical for learning, problem-solving, and self-regulation (Spada & Marino, 2017). In Indonesia, evidence indicates that cognitive problems are prevalent among adolescents and are frequently linked to early-life nutritional deficiencies, stunting, and mental health issues. The Indonesian Family Life Survey reported that approximately 41% of children aged 7–14 years scored below age norms on cognitive assessments, with stunted children exhibiting reduced attention, memory, and reasoning abilities (Lestari et al., 2024; Setyaningsih et al., 2021). Complementary findings from the Indonesian National Adolescent Mental Health Survey indicate that one-third of adolescents experience attention deficits or other cognitive difficulties alongside mental health challenges such as anxiety, depression, or ADHD (Wahdi et al., 2023).
Structured sport-based after-school programs have been associated with numerous benefits, including increased PA, healthier eating habits, improved sleep hygiene, and enhanced cognitive performance (Lubans et al., 2016; Bailey, 2018). In contrast, non-sport programs may lack sufficient physical engagement, potentially leaving adolescents vulnerable to higher PIU, poor diet, and cognitive underperformance (Nurulfa et al., 2026). Despite these findings. However, few studies have simultaneously examined the relationships among physical activity, sleep quality, eating habits, problematic internet use, and cognitive ability within the context of sport and non-sport after-school programs. Moreover, the mechanisms by which lifestyle behaviors and digital media use interact to influence PA and cognitive outcomes remain poorly understood, representing a clear research gap.
To address this gap, we adopt an integrated lifestyle framework, conceptualizing physical activity, sleep quality, eating habits, and problematic internet use as interconnected behavioral systems that collectively influence cognitive development. Using an integrated lifestyle framework grounded in self-determination theory and ecological systems perspectives, this study conceptualizes physical activity, sleep quality, eating habits, and problematic internet use as interconnected behavioral systems that collectively shape adolescent cognitive development rather than as isolated factors (Tong & An, 2023). Physical activity may enhance sleep quality, reduce stress, and promote healthier eating, while adequate sleep supports executive function and better dietary choices, and nutritious eating sustains both cognitive performance and physical engagement. In contrast, problematic internet use may disrupt these positive pathways by increasing sedentary behavior, reducing time for health-promoting activities, impairing sleep, and contributing to unhealthy dietary patterns. These behaviors interact within adolescents’ daily routines and may be further influenced by socioeconomic conditions such as family resources, parental education, and access to supportive environments. Within structured after-school settings, sport-based programs may strengthen positive behavioral synergies through routine, peer support, and motivation, whereas non-sport programs may offer fewer protective factors, increasing vulnerability to clustered risk behaviors. Therefore, by examining adolescents in sport (SA) and non-sport (NS) after-school programs, this study aims to provide a more comprehensive understanding of how multiple lifestyle behaviors and program type collectively relate to cognitive outcomes.
This study aimed to examine differences in physical activity, sleep quality, eating habits, problematic internet use, and cognitive ability among Indonesian adolescents participating in sport and non-sport extracurricular programs. In addition, the study investigated the relationships among physical activity and sleep quality, eating habits, problematic internet use, and cognitive ability, while accounting for socioeconomic and demographic covariates. It was hypothesized that adolescents participating in sport programs would demonstrate higher physical activity, healthier eating habits, better sleep quality, lower problematic internet use, and higher cognitive ability compared with their peers in non-sport programs.
2. Method
2.1 Participants
Data collection was conducted during the regular school term across three public junior high schools in Jakarta, Indonesia. A two-stage sampling approach was used. At the institutional level, schools were recruited using a convenience sampling method based on accessibility, administrative approval, and willingness to participate during the data collection period. Multiple public junior high schools in Jakarta were approached, and the final sample included three schools that consented and were available to participate.
Following approval from school administrators, eligible students were identified based on their enrollment in either sport after-school (SA) or non-sport after-school (NS) programs. SA programs consisted of structured extracurricular sports such as football, basketball, volleyball, badminton, athletics, and martial arts, typically conducted 1–3 times per week for 60–90 minutes per session under the supervision of physical education teachers or coaches (Eime & Harvey, 2017; Janssen & LeBlanc, 2010). In contrast, NS programs included extracurricular activities focused on academic, artistic, or social development, such as science club, language club, music club, youth red cross, and student leadership activities. which did not involve structured exercise or sport-related movement, and typically met 1–2 times per week for 60–90 minutes (Feldman & Matjasko, 2005; Kemendikbud, 2020).
At the individual level, students in Grades 7 and 8 who met the inclusion criteria were invited to participate. Inclusion criteria were enrolment in either sport-based or non-sport extracurricular activities for at least 6 months, ability to complete the questionnaires independently, and physical and cognitive capability to participate in both the self-report assessments and cognitive ability test. Information sheets and parental consent forms were distributed prior to data collection. All assessments were administered during scheduled school hours in quiet classroom settings under the supervision of trained research assistants. Participants first completed self-report questionnaires assessing physical activity, sleep quality, eating habits, and problematic internet use. Standardized instructions were provided, and research staff were available to clarify questions to ensure accurate and complete responses.
A total of 450 consent forms were distributed, of which 360 students provided consent and participated in the study. Participant recruitment across schools was distributed as follows: School A (n = 116), School B (n = 105), and School C (n = 139). Following data screening, 20 participants (5.56%) were excluded due to incomplete data using listwise deletion, resulting in a final analytical sample of 340 students. Missing data across study variables ranged from 1.39% to 3.33%, including physical activity (PA = 1.67%, n = 6), problematic internet use (PIU = 2.22%, n = 8), cognitive ability (CA = 3.33%, n = 12), eating habits (EH = 1.39%, n = 5), and sleep quality (SQ = 3.06%, n = 11). Given the relatively low level of missingness, listwise deletion was considered appropriate for the final analyses. Of the final sample, 165 students were enrolled in NS after-school programs and 175 were enrolled in SA programs. The sample was approximately balanced by gender (boys: NS = 46.1%; SA = 53.1%), with participant ages ranging from 12.0 to 16.0 years (see Figure 1).
Figure 1
Flowchart of Participant enrolment.
2.2 Instruments
2.2.1. Cognitive Ability (CA)
Cognitive ability was assessed using the short version of the Raven’s Standard Progressive Matrices (RSPM-Short), which contains 15 items selected from the original RSPM (Kramer & Huizenga, 2023; Langener et al., 2022). Each item presents a visual pattern with a missing piece, and participants must select the correct piece from multiple-choice options. The test was administered in supervised classroom settings either in a paper-and-pencil format or via a standardized digital version, depending on the availability of school resources. In both formats, administration followed identical instructions and timing conditions. Instructions emphasize the completion of all items, and participants may skip them if unsure. Completion typically takes 5–15 minutes, depending on the participant’s age and cognitive level. Scoring is straightforward: each correct response receives 1 point, with a maximum total score of 15. Raw scores are interpreted using percentile norms, with typical ranges as follows: 13–15 = high ability (≥90th percentile), 11–12 = above average (75th–89th percentile), 7–10 = average (25th–74th percentile), 4–6 = below average (10th–24th percentile), and 0–3 = substantially below average (<10th percentile). The RSPM-Short is norm-referenced, culture-fair, and has been validated as a reliable estimate of fluid intelligence comparable to the full version. (Langener et al., 2022; Raven, 2003).
2.2.2 Problematic Internet Use (PIU)
The Problematic and Risky Internet Use Screening Scale (PRIUSS) is a validated self-report tool designed to assess problematic internet use in adolescents. The PRIUSS-18 standard comprises 18 items, covering three domains: social behavior, emotional behavior, and risky internet use. Each item is rated on a 5-point scale (0 = Never to 4 = Very Often), with total scores ranging from 0 to 72. Higher scores indicate greater risk of problematic internet use, with cutoffs following standard recommendations: 0–20 = Low risk, 21–39 = Moderate risk, and ≥40 = High risk. This scale has demonstrated strong reliability and validity, and typically takes less than 10 minutes to complete (Klavina et al., 2023).
2.2.3 Physical Activity Questionnaire for Adolescents (PAQ-A)
A 7-day self-reported recall instrument to assess general physical activity levels. The instrument consists of 9 items rated on a 5-point Likert scale. The final activity score ranges from 1 (low activity) to 5 (high activity). The PAQ-A has been widely validated for use with adolescents (Crocker et al., 1997; Donen, 2005). The mean score represents the adolescent’s PA level: 1.00–1.99 = Low, 2.00–2.99 = Moderate, and 3.00–5.00 = High, consistent with prior studies of adolescent populations (Donen, 2005).
2.2.4 Pittsburgh Sleep Quality Index (PSQI)
Sleep quality was measured using the Pittsburgh Sleep Quality Index (PSQI) (Goldstone et al., 2020; Peerbhay et al., 2025) which contains 19 self-reported items + 5 additional questions for sleep partners and 7 domains: subjective sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbances, use of sleep medications, and daytime dysfunction, with scores ranging from 0 to 21. 0–4 = Good, ≥5 = Poor, >10 = Severe disturbance (Brown et al., 2025).
2.2.5 Adolescent Food Habits Questionnaire
The Adolescent Food Habits Checklist (AFHC) by Johnson et al. (Johnson et al., 2002) is a validated self-report tool that assesses healthy eating behavior in adolescents aged 11–18 years. The tool consists of 23 yes/no items covering healthy eating habits (e.g., fruit intake) and unhealthy habits (e.g., fried foods). Scoring is binary: 1 for healthy responses and 0 for unhealthy responses, with items containing negative wording scored inversely. The total score ranges from 0 to 23, with higher scores indicating healthier eating patterns. Generally, scores of 18–23 reflect healthy behaviour, 12–17 moderate, and 0–11 low (Johnson et al., 2002).
2.2.6. Socioeconomic Status (SES)
Socioeconomic Status (SES) was measured using a composite index consisting of four components: family income, affordability, parents’ education, and subjective financial condition, consistent with established multidimensional SES frameworks that combine objective and subjective indicators of social position (Adler et al., 2000; Lampert et al., 2018; Psaki et al., 2014). Family income was assessed based on self-reported household income level, affordability reflected perceived ability to meet basic needs, parents’ education was calculated as the average education level of both parents, and subjective financial condition reflected perceived household financial well-being, consistent with prior evidence emphasizing education, income, and perceived status as core SES dimensions (Duncan et al., 2016). Each component was rated on a 5-point Likert scale, with higher scores indicating higher SES. The overall SES score was computed by averaging all four components. Based on the composite score, participants were classified into low SES (1.00–2.33), middle SES (2.34–3.67), and high SES (3.68–5.00). However, in the multivariate general linear model, FI, AP, PE, and SFC were entered as separate covariates to allow examination of their independent associations with the outcomes and to avoid potential loss of information associated with composite scoring.
2.3 Data Collection Procedures
Participants were asked to complete the survey via a secure Google Forms link, accessible from smartphones, tablets, or computers. Responses were automatically recorded in a protected Google Sheet and were only accessible to the research team. Prior to participation, written informed consent was obtained from parents or legal guardians, and student assent was obtained from all participants. The consent form provided detailed information about the study’s purpose, procedures, potential risks, and benefits. Parents were assured that participation was voluntary, that data would be collected confidentially and anonymously, and that their child could withdraw at any time without consequences. Students received an age-appropriate assent form explaining the nature of the tasks, the approximate completion time, and their right to decline or discontinue participation.
2.4 Ethical Considerations
The measures used in this study include physical activity, sleep quality, eating habits, and risky internet use among adolescents and cognitive ability. This study has obtained ethical approval in accordance with the decision of the LSPA Ethics Committee (No. S/51813, 20 January 2024).
2.5 Statistical Analysis
All statistical analyses were conducted using SPSS. Data were first screened for missing values and outliers, confirming complete responses for all 340 participants across physical activity (PA), eating habits (EH), sleep quality (SQ), problematic internet use (PIU), and cognitive ability (CA). Descriptive statistics, including means and standard deviations, were computed for all study variables.
Group differences between non-sport (NS) and sport (SA) extracurricular participants were analyzed using a multivariate general linear model (GLM)/multivariate analysis of covariance (MANCOVA) framework. Physical activity, eating habits, sleep quality, problematic internet use, and cognitive ability were entered as dependent variables, while extracurricular activity group (NS vs SA) was included as the fixed factor. To adjust for socioeconomic influences, family income (FI), affordability (AP), parents’ education (PE), and subjective financial condition (SFC) were entered separately as covariates in the multivariate model.
Univariate ANCOVA results were reported for each outcome variable, and estimated marginal means were used for adjusted group comparisons. Effect sizes were reported using partial eta squared (ηp2). Assumptions of the model, including homogeneity of variance and equality of covariance matrices, were assessed using Levene’s test and Box’s M test, respectively. Pairwise comparisons were adjusted using Bonferroni correction. Statistical significance was set at p < 0.05.
3. Results
The demographic characteristics of participants are presented in Table 1. The results indicate that there were no statistically significant differences between the non-sport (NS) and sport activity (SA) groups in age, height, weight, or body mass index (BMI), suggesting that the two groups were comparable in basic anthropometric and developmental characteristics. Similarly, socioeconomic status distribution did not differ significantly between groups (p = .260), with both NS and SA participants predominantly belonging to the middle SES category.
Table 1
Demographic and Anthropometric Characteristics of Participants.
| VARIABLE | NS (n = 165) FEMALE 53.9% | SA (n = 175) FEMALE 46.9% | p |
|---|---|---|---|
| Age (years) | 13.54 ± 0.87 | 13.61 ± 0.80 | .466 |
| Height (cm) | 159.41 ± 6.90 | 160.13 ± 13.80 | .550 |
| Weight (kg) | 51.12 ± 7.46 | 50.28 ± 7.40 | .301 |
| BMI (kg/m2) | 20.06 ± 2.28 | 19.77 ± 2.22 | .236 |
| SES score | 1.96 ± 0.68 | 2.22 ± 0.60 | <.001 |
| FI | 2.76 ± 0.86 | 3.09 ± 0.73 | <.001 |
| AP | 2.80 ± 0.96 | 3.18 ± 0.91 | <.001 |
| SFC | 2.85 ± 0.95 | 3.22 ± 0.83 | <.001 |
| PE | 3.35 ± 0.75 | 3.52 ± 0.72 | .032 |
| SES Level | N (%) | ||
| Low (1.00) | 26 (15.8%) | 19 (10.9%) | 0.260 |
| Middle (2.00) | 99 (60.0%) | 103 (58.9%) | |
| High (3.00) | 40 (24.2%) | 53 (30.3%) | |
| Type of Extracurricular | NS (n and %) | SA (n and %) | |
| Youth Red Cross 44 (26.7) | Athletics 69 (39.4) | <.001 | |
| Language Clubs 28 (17.0) | Basketball 46 (26.3) | ||
| Traditional Music 15 (9.1) | Martial Art 30 (17.1) | ||
| Science 15 (9.1) | Volleyball 20 (11.4) | ||
| Choir 12 (7.3) | Badminton 6 (3.4) | ||
| Other 51 (30.9) | Other 4 (2.3) |
[i] Notes. Values are presented as mean ± standard deviations; Non-Sports; SA = Sport Activity; p = significance level (two-tailed) < .05.
Regarding gender distribution, both groups showed a relatively balanced composition, although the proportion of females was slightly higher in the NS group (53.9%) compared with the SA group (46.9%). The comparison of extracurricular activity types revealed a statistically significant difference between groups (p < .001), reflecting the distinct nature of participation, with NS participants primarily engaged in non-sport structured activities such as Youth Red Cross, youth organizations, and arts-related programs, whereas SA participants were mainly involved in athletic and sport-based activities such as athletics, basketball, martial arts, volleyball, and badminton.
However, independent-samples t-tests indicated significant differences in continuous socioeconomic indicators between groups. The SA group reported higher family income, affordability, subjective financial condition (all p < .001), and parental education (p = .032) compared with the NS group. Consistently, the composite SES score was significantly higher among SA participants (2.22 ± 0.60) than NS participants (1.96 ± 0.68, p < .001), suggesting that adolescents participating in sport activities generally come from more advantaged socioeconomic backgrounds.
Table 2 summarizes group differences in behavioral and cognitive variables between adolescents participating in NS and SA after-school programs. Compared with adolescents in NS group, those engaged in SA programs reported significantly higher levels of physical activity 2.67 (0.83) p < .001) and healthier eating habits 15.12 (5.24) p < .001). Sleep quality also differed significantly between groups, with better sleep outcomes observed among SA participants 4.19 (2.75) p < .001) compare to NS. In addition, adolescents in SA programs demonstrated lower levels of problematic internet use 16.81 (8.02) p = .001) and higher cognitive ability scores 10.49 (2.41) p < .001) compare to NS.
Table 2
Group Comparison of Variables by After-School Program.
| VARIABLE | NS (n = 165) | SA (n = 175) | p | COHEN’S d |
|---|---|---|---|---|
| Physical Activity | 1.87 (0.68) | 2.67 (0.83) | <.001 | 0.76 |
| Eating Habits | 13.05 (5.44) | 15.12 (5.24) | <.001 | 0.39 |
| Sleep Quality | 6.72 (4.16) | 4.19 (2.75) | <.001 | 0.72 |
| Problematic Internet Use | 20.07 (9.08) | 16.81 (8.02) | .001 | 0.38 |
| Cognitive Ability | 8.44 (2.50) | 10.49 (2.41) | <.001 | 0.83 |
[i] Notes: NS = Non Sport, SA = Sport Activity. Cohen’s d: 0.20 = small, 0.50 = medium, 0.80 = large. Significance level (two-tailed) < .05.
Table 3 presents the distribution of behavioural and cognitive categories among adolescents in non-sport (NS) and sport activity (SA) programs. Significant group differences were observed across all variables. SA were more likely to report high physical activity levels, with 47.4% classified as highly active compared with 17.0% of NS. Healthier eating habits were also more prevalent among SA participants (38.3%) compared with NS adolescents. Adolescents participating in sport activities demonstrated substantially better sleep profiles than those in non-sport programs. A larger proportion of sport participants were classified in the good sleep quality category (PSQI 0–4; 76.6%), compared with 46.7% of non-sport adolescents, while severe sleep problems were considerably less common among sport participants (3.4% vs. 23.0%). Moderate-risk problematic internet use was slightly lower among SA participants (33.7%) compared with NS adolescents. Cognitive ability also differed significantly between groups, with 24.0% of SA participants classified as having high cognitive ability compared with 8.5% of NS adolescents.
Table 3
Distribution of Behavioral and Cognitive Categories among Adolescents in Non-Sport and Sport After-School Programs.
| VARIABLE | LEVEL | NS (n = 165) | SA (n = 175) | TOTAL (n = 340) | χ2 | DF | p | phi/CRAMER’S V |
|---|---|---|---|---|---|---|---|---|
| Physical Activity | Low (1–1.99) | 50 (30.3%) | 2 (1.1%) | 52 | 71.38 | 2 | <.001 | 0.458 |
| Moderate (2–2.99) | 87 (52.7%) | 90 (51.4%) | 177 | |||||
| High (3–5) | 28 (17.0%) | 83 (47.4%) | 111 | |||||
| Eating Habits | Low (0–11) | 72 (43.6%) | 44 (25.1%) | 116 | 13.98 | 2 | .001 | 0.203 |
| Moderate (12–17) | 52 (31.5%) | 64 (36.6%) | 116 | |||||
| Healthy (18–23) | 41 (24.8%) | 67 (38.3%) | 108 | |||||
| Sleep Quality | Good (0–4) | 77 (46.7%) | 134 (76.6%) | 211 | 41.06 | 2 | <.001 | 0.348 |
| Poor (5–10) | 50 (30.3%) | 35 (20.0%) | 85 | |||||
| Severe (>10) | 38 (23.0%) | 6 (3.4%) | 44 | |||||
| Problematic Internet Use | Low risk (0–20) | 95 (57.6%) | 116 (66.3%) | 211 | 6.19 | 2 | .045 | 0.135 |
| Moderate risk (21–39) | 66 (40.0%) | 59 (33.7%) | 125 | |||||
| High risk (>40) | 4 (2.4%) | 0 (0%) | 4 | |||||
| Cognitive Ability | Below Average (4–6) | 32 (19.4%) | 3 (1.7%) | 35 | 41.79 | 3 | <.001 | 0.351 |
| Average (7–10) | 94 (57.0%) | 89 (50.9%) | 183 | |||||
| Above Average (11–12) | 25 (15.2%) | 41 (23.4%) | 66 | |||||
| High Ability (13–15) | 14 (8.5%) | 42 (24.0%) | 56 |
[i] Notes: Data are presented as frequency (percentage). Group comparisons were conducted using chi-square tests. Statistical significance was set at p < .05. Percentages reflect valid cases within each extracurricular activity group. Non-Sports; SA = Sport Activity.
Pearson correlation analyses (Figure 2) revealed significant associations among physical activity, eating habits, sleep quality, problematic internet use, and cognitive ability in both groups. Among NS participants, physical activity was positively correlated with eating habits (r = 0.39, p < .01) and cognitive ability (r = 0.44, p < .01), and negatively correlated with sleep problems (r = –0.35, p < .01) and problematic internet use (r = –0.61, p < .01). Eating habits were also positively associated with cognitive ability (r = 0.65, p < .01) and negatively associated with problematic internet use (r = –0.62, p < .01); the association with sleep problems was not significant (r = –0.14, p = .069). Problematic internet use was strongly negatively associated with cognitive ability (r = –0.79, p < .01).
Figure 2
Correlation Network of Physical Activity, Eating Habits, Sleep Quality, Problematic Internet Use, and Cognitive Ability Among Non-Sport and Sport After-School Participants.
Note. The Figure 2 presents Pearson correlation networks separately for Non-Sport and Sport after-school program participants (side-by-side comparison). Abbreviations: Physical Activity (PA), Eating Habits (EH), Sleep Quality (SQ), Problematic Internet Use (PIU), Cognitive Ability (CA). Positive correlations are indicated by blue lines, and negative correlations are indicated by orange lines.
Among SA participants, physical activity was positively associated with eating habits (r = 0.51, p < .01) and cognitive ability (r = 0.41, p < .01), and negatively associated with sleep problems (r = –0.30, p < .01) and problematic internet use (r = –0.62, p < .01). Eating habits were positively associated with cognitive ability (r = 0.74, p < .01) and negatively associated with problematic internet use (r = –0.71, p < .01); associations with sleep problems were not significant. Problematic internet use was strongly negatively associated with cognitive ability (r = –0.77, p < .01). Additionally, partial correlation analyses controlling for family income, affordability, parents’ education, and subjective financial condition were conducted and are presented in Supplementary Table S1.
Table 4 presents the mean and standard deviation of eating habits, sleep quality, problematic internet use, and cognitive ability across physical activity levels within each extracurricular activity group. Among non-sport adolescents, the largest subgroup was observed at PA level 2 (n = 87), with mean scores of 14.41 ± 5.26 for eating habits, 6.71 ± 2.28 for sleep quality, 16.74 ± 6.96 for problematic internet use, and 9.23 ± 2.58 for cognitive ability. In the sport group, the largest subgroup was also observed at PA level 2 (n = 92), with mean scores of 12.87 ± 5.19 for eating habits, 6.21 ± 1.35 for sleep quality, 20.79 ± 6.88 for problematic internet use, and 9.68 ± 2.27 for cognitive ability. Notably, the highest physical activity level (level 4) was observed only among sport participants, who demonstrated the highest eating habit scores (19.38 ± 2.67) and cognitive ability scores (12.18 ± 2.05), as well as the lowest problematic internet use scores (8.08 ± 3.55).
Table 4
Eating Habits, Sleep Quality, Problematic Internet Use, and Cognitive Ability Across Physical Activity Levels by Extracurricular Activity Participation.
| PA LEVEL | EATING HABITS | SLEEP QUALITY | PROBLEMATIC INTERNET USE | COGNITIVE ABILITY | ||||
|---|---|---|---|---|---|---|---|---|
| NS (n) | SA (n) | NS (n) | SA (n) | NS (n) | SA (n) | NS (n) | SA (n) | |
| 1 | 9.54 ± 4.04 (50) | 12.00 ± 7.07 (2) | 6.98 ± 1.71 (50) | 4.00 ± 4.24 (2) | 29.12 ± 7.33 (50) | 20.00 ± 9.90 (2) | 6.54 ± 1.01 (50) | 10.00 ± 4.24 (2) |
| 2 | 14.41 ± 5.26 (87) | 12.87 ± 5.19 (92) | 6.71 ± 2.28 (87) | 6.21 ± 1.35 (92) | 16.74 ± 6.96 (87) | 20.79 ± 6.88 (92) | 9.23 ± 2.58 (87) | 9.68 ± 2.27 (92) |
| 3 | 15.11 ± 5.34 (28) | 16.24 ± 4.31 (42) | 4.32 ± 1.89 (28) | 3.88 ± 1.84 (42) | 14.29 ± 5.16 (28) | 16.05 ± 6.65 (42) | 9.39 ± 2.27 (28) | 10.69 ± 2.15 (42) |
| 4 | — | 19.38 ± 2.67 (39) | — | 5.10 ± 1.93 (39) | — | 8.08 ± 3.55 (39) | — | 12.18 ± 2.05 (39) |
[i] Notes: Non-Sports; SA = Sport Activity; p = significance level (two-tailed) < .05. PA = Physical Activity.
Table 5 Revealed the multivariate general linear model indicated that extracurricular activity, socioeconomic variables, gender, and BMI level showed differentiated associations with adolescent health-related outcomes. Participation in sport activities was significantly associated with higher physical activity (F = 20.565, p < .001, ηp2 = .060), better sleep quality (F = 11.330, p = .001, ηp2 = .034), and higher cognitive ability (F = 17.265, p < .001, ηp2 = .051), while no significant effects were observed for eating habits or problematic internet use. Gender and BMI level did not show significant associations across most outcomes.
Table 5
Multivariate General Linear Model Examining Associations of Extracurricular Activity, Gender, BMI, and Socioeconomic Factors with Adolescent Health Behaviors and Cognitive Ability.
| PREDICTOR/FACTOR | OUTCOME | F | df | p | PARTIAL η2 |
|---|---|---|---|---|---|
| Extracurricular Activity (SA vs NS) | Physical Activity | 20.565 | 1.324 | <.001 | .060 |
| Eating Habits | 1.454 | 1.324 | .229 | .004 | |
| Sleep Quality | 11.330 | 1.324 | .001 | .034 | |
| Problematic Internet Use | 3.404 | 1.324 | .066 | .010 | |
| Cognitive Ability | 17.265 | 1.324 | <.001 | .051 | |
| Gender | Physical Activity | 0.560 | 1.324 | .455 | .002 |
| Eating Habits | 0.966 | 1.324 | .326 | .003 | |
| Sleep Quality | 0.229 | 1.324 | .633 | .001 | |
| Problematic Internet Use | 0.802 | 1.324 | .371 | .002 | |
| Cognitive Ability | 1.003 | 1.324 | .317 | .003 | |
| BMI | Physical Activity | 0.513 | 2.324 | .599 | .003 |
| Eating Habits | 2.853 | 2.324 | .059 | .017 | |
| Sleep Quality | 1.063 | 2.324 | .347 | .007 | |
| Problematic Internet Use | 1.537 | 2.324 | .217 | .009 | |
| Cognitive Ability | 1.357 | 2.324 | .259 | .008 | |
| FI | Physical Activity | 0.409 | 1.324 | .523 | .001 |
| Eating Habits | 1.070 | 1.324 | .302 | .003 | |
| Sleep Quality | 2.737 | 1.324 | .099 | .008 | |
| Problematic Internet Use | 2.120 | 1.324 | .146 | .007 | |
| Cognitive Ability | 2.062 | 1.324 | .152 | .006 | |
| AP | Physical Activity | 8.950 | 1.324 | .003 | .027 |
| Eating Habits | 6.471 | 1.324 | .011 | .020 | |
| Sleep Quality | 8.122 | 1.324 | .005 | .024 | |
| Problematic Internet Use | 7.475 | 1.324 | .007 | .023 | |
| Cognitive Ability | 4.443 | 1.324 | .036 | .014 | |
| SFC | Physical Activity | 0.009 | 1.324 | .924 | .000 |
| Eating Habits | 1.180 | 1.324 | .278 | .004 | |
| Sleep Quality | 0.984 | 1.324 | .322 | .003 | |
| Problematic Internet Use | 1.262 | 1.324 | .262 | .004 | |
| Cognitive Ability | 1.181 | 1.324 | .278 | .004 | |
| PE | Physical Activity | 7.193 | 1.324 | .008 | .022 |
| Eating Habits | 21.948 | 1.324 | <.001 | .063 | |
| Sleep Quality | 18.147 | 1.324 | <.001 | .053 | |
| Problematic Internet Use | 41.321 | 1.324 | <.001 | .113 | |
| Cognitive Ability | 67.315 | 1.324 | <.001 | .172 |
[i] Note. Statistical significance was set at p < .05. Abbreviations: FI = Family Income; AP = Affordability of Parents; SFC = Subjective Financial Condition; PE = Parental Education.
Among socioeconomic indicators, affordability emerged as a consistent predictor across all domains, showing significant associations with physical activity (p = .003), eating habits (p = .011), sleep quality (p = .005), problematic internet use (p = .007), and cognitive ability (p = .036), with small effect sizes. In contrast, family income and subjective financial condition were not significant predictors of any outcome. Parental education demonstrated the strongest and most consistent effects, significantly predicting all outcomes, including physical activity (p = .008), eating habits (p < .001), sleep quality (p < .001), problematic internet use (p < .001), and cognitive ability (p < .001), with effect sizes ranging from small to large, particularly for cognitive ability (ηp2 = .172) and problematic internet use (ηp2 = .113).
4. Discussion
The present study examined differences in physical activity (PA), eating habits (EH), sleep quality (SQ), problematic internet use (PIU), and cognitive ability (CA) among adolescents participating in sport activity (SA) and non-sport (NS) extracurricular programs. In addition, the study explored the associations among these behavioral and cognitive outcomes while accounting for socioeconomic and demographic covariates. Overall, the findings supported the study hypotheses, indicating that adolescents in SA programs demonstrated higher PA, healthier EH, better SQ, lower PIU, and higher CA compared with those in NS programs. These results highlight the potential role of structured sport participation in promoting healthier lifestyle behaviors and cognitive functioning during adolescence.
One of the most prominent findings of the present study is that adolescents participating in SA reported significantly higher levels of physical activity compared with those in the NS group. This result is consistent with previous research demonstrating that organized sport participation is strongly associated with higher levels of moderate-to-vigorous physical activity among adolescents (Eime et al., 2015; Janssen & LeBlanc, 2010; Westerbeek & Eime, 2021). Organized sport programs provide structured opportunities that are associated with higher levels of adolescents’ physical activity. In contrast, adolescents in non-sport settings may experience greater discretionary time availability, increasing susceptibility to sedentary behavioral displacement, particularly toward screen-based activities (Nurulfa & Klavina, 2026). Global health reports indicate that a large proportion of adolescents fail to meet recommended physical activity guidelines, highlighting the importance of school- and community-based sport programs in promoting active lifestyles (Okely et al., 2021; World Health Organization, 2020). The large effect size observed in the present study further emphasizes the substantial difference in activity levels between adolescents involved in sport and those engaged in non-sport extracurricular activities.
Beyond physical activity itself, adolescents participating in sport activities also reported healthier eating habits and better sleep quality. These findings are consistent with evidence suggesting that health-related behaviors tend to cluster during adolescence, with physically active individuals being more likely to engage in other positive lifestyle practices (Leech et al., 2014; Pearson et al., 2019). Previous studies have similarly reported associations between physical activity and healthier dietary behaviors (Mora-Gonzalez et al., 2019; Voráčová et al., 2018; Zhu et al., 2019). Participation in sport settings may expose adolescents to health-promoting norms through coaches, teammates, and sport-related educational experiences. Likewise, the better sleep quality observed among sport participants aligns with evidence linking physical activity to healthier sleep patterns through mechanisms such as increased energy expenditure, improved circadian regulation, and reduced psychological stress (Kredlow et al., 2015; Li et al., 2021; Xu et al., 2019). Adolescents engaged in organized sport may also benefit from more structured daily routines that support consistent sleep schedules.
The clustering of health-related behaviors observed in the present study may also help explain the lower levels of problematic internet use reported among adolescents participating in sport activities. Excessive internet use has become a growing concern among adolescents and has been associated with sedentary behavior, sleep disturbances, and adverse psychosocial outcomes (Lopez-Fernandez & Kuss, 2020; Pontes & Macur, 2021; Throuvala et al., 2018). Adolescents involved in sport activities are observed to have lower levels of problematic internet use, potentially due to differences in daily time-use patterns. Previous studies have also reported that higher levels of physical activity are associated with lower levels of internet addiction and screen dependency among adolescents (An et al., 2014; Sundar et al., 2018). The present findings therefore suggest that participation in sport participation may be associated with lower levels of problematic internet use. However, although statistically significant differences in problematic internet use were observed between groups, the small effect size suggests that these findings should be interpreted with caution in terms of practical significance.
In addition to behavioral differences, adolescents participating in sport activities demonstrated higher cognitive ability scores than their non-sport peers. This finding is consistent with a growing body of literature linking physical activity and sport participation with cognitive functioning in children and adolescents (Donnelly et al., 2016; Lubans et al., 2016). Rather than being explained through nonspecific physiological mechanisms, this association can be better understood through executive function and cognitive control processes. Sport participation requires continuous engagement in goal-directed behavior, including rapid decision-making, task switching, and the regulation of competing behavioral demands. These demands are closely linked to core executive function components such as inhibitory control, working memory updating, and cognitive flexibility. Many sport activities involve strategic thinking and decision-making, which may be related to cognitive engagement, as suggested in previous literature (Hillman et al., 2008). Participation in team sports may also strengthen executive functions, including attention, working memory, and cognitive flexibility (Baxter-Jones et al., 2020; A. F. Silva et al., 2020). Although the cross-sectional nature of the present study precludes conclusions regarding causality, the findings suggest that adolescents engaged in sport activities exhibit more favorable cognitive outcomes than those participating in non-sport extracurricular programs.
The correlation analyses further reinforced the interconnected nature of adolescent lifestyle behaviors and cognitive outcomes. Physical activity was positively associated with healthier eating habits and higher cognitive ability, while showing negative associations with sleep problems and problematic internet use. These patterns support the concept of behavioral clustering, whereby multiple health-promoting or health-risk behaviors co-occur rather than operating independently (Leech et al., 2014; Pearson et al., 2019). Consequently, interventions targeting a single lifestyle behavior may have broader implications for adolescent health if they contribute to positive behavioral patterns across multiple domains.
Importantly, the multivariate analyses revealed that socioeconomic factors played a significant role in shaping adolescent behavioral and cognitive outcomes. Among SES indicators, affordability emerged as a consistent predictor across all domains, significantly associated with PA, EH, SQ, PIU, and CA. This suggests that adolescents’ ability to access resources and opportunities may directly influence both health behaviors and cognitive development.
In contrast, family income and subjective financial condition were not significant predictors, indicating that objective household income alone may not fully capture adolescents’ lived socioeconomic conditions. Notably, parental education showed the strongest and most consistent associations across all outcomes, including a large effect on cognitive ability and PIU. These findings align with previous research showing that parental education is a key determinant of adolescent health behaviors, likely through mechanisms such as health literacy, parenting practices, and structured home environments (Pearson et al., 2019; Leech et al., 2014).
Within the Indonesian context, these findings should be considered alongside educational policies that promote participation in structured extracurricular activities as part of holistic student development. Programs such as Kartu Jakarta Pintar (KJP) seek to reduce economic barriers to educational participation among students from lower-income families (Nurdin, 2019). While such initiatives may help improve access to developmental opportunities, the continued associations observed for parental education and affordability suggest that school-based programs alone may not fully eliminate socioeconomic disparities in health and cognitive outcomes. Nevertheless, structured sport activities may provide accessible opportunities for physical, cognitive, and social engagement that support adolescent development across diverse socioeconomic backgrounds.
Strength and Practical Implications
This study has several strengths. It examined multiple lifestyle and cognitive variables simultaneously, providing a comprehensive assessment of adolescent health and behavioral outcomes. The relatively large sample size and use of validated instruments enhanced the reliability of the findings. In addition, multivariable analyses, including ANCOVA and partial correlation procedures, were conducted while controlling for key socioeconomic indicators (family income, affordability, parental education, and subjective financial condition), thereby strengthening the internal validity of the findings beyond simple bivariate associations.
The findings highlight the potential value of structured extracurricular sport programs as supportive environments for promoting positive health-related behaviors and cognitive development among adolescents. Participants involved in extracurricular sport activities demonstrated higher physical activity levels and more favorable behavioral profiles, including healthier eating habits, better sleep quality, and lower levels of problematic internet use. Although these findings represent adjusted associations rather than causal relationships, they suggest that participation in structured sport settings may be linked to healthier lifestyle patterns during adolescence.
Importantly, the present study also demonstrated that parental education was one of the strongest correlates of cognitive ability and problematic internet use, indicating that family educational background remains an important influence on adolescent development. While school-based sport programs cannot fully offset disparities associated with socioeconomic circumstances, they may provide equitable opportunities for physical, cognitive, and social engagement across diverse student populations. In this context, extracurricular sport programs may help mitigate socioeconomic disparities in adolescent health-related behaviors and cognitive development by providing structured opportunities for participation, although they cannot fully overcome broader family- and community-level influences. Therefore, schools, communities, and policymakers may consider strengthening extracurricular sport opportunities while simultaneously supporting family engagement and broader educational initiatives that address socioeconomic inequalities affecting adolescent health and cognitive development.
Study Limitations and Future Research
Several limitations should also be acknowledged. The cross-sectional design limits causal inference, and self-report measures may introduce recall and social desirability bias. Physical activity was assessed using the PAQ-A rather than objective measures, limiting precision in activity intensity estimation. Group allocation was not randomized, introducing potential self-selection bias. Although additional adjusted analyses, including multivariate models and partial correlations controlling for socioeconomic covariates, were conducted, residual confounding from unmeasured factors such as baseline physical activity, motivation, mental health, academic performance, pubertal status, and family support cannot be excluded. In addition, cognitive ability was assessed using both paper-based and digital versions of the RSPM-Short. Although identical test content, instructions, administration procedures, and time limits were used across formats, the potential influence of administration mode on cognitive performance was not formally examined and may have introduced additional measurement variability. Furthermore, sport type, intensity, frequency, and duration were not differentiated, limiting interpretation of potential dose–response relationships. Consequently, the generalizability of the findings may be restricted to adolescents from similar cultural and educational contexts.
Future research should employ longitudinal, intervention-based, or multilevel designs to better understand whether sport participation is associated with long-term changes in lifestyle behaviors and cognitive outcomes. Further studies should also incorporate objective physical activity measures (e.g., accelerometer-derived MVPA, step counts, or energy expenditure) and broader confounding variables such as gender, mental health, nutritional status, baseline academic ability, pubertal status, and school-level clustering. In addition, pathway-based analytical approaches, such as mediation analysis, path analysis, or Structural Equation Modeling (SEM), may help clarify potential direct and indirect relationships among physical activity, sleep quality, problematic internet use, and cognitive functioning in adolescent development.
5. Conclusion
In conclusion, adolescents participating in sport activity (SA) extracurricular programs were associated with healthier lifestyle behaviors and higher cognitive ability compared with those involved in non-sport (NS) activities. SA participants reported higher physical activity, healthier eating habits, better sleep quality, lower problematic internet use, and higher cognitive ability scores. Physical activity was significantly associated with multiple behavioral and cognitive outcomes, highlighting the interconnected nature of adolescent lifestyle behaviors. Socioeconomic factors, particularly parental education and affordability, were consistently associated with all outcomes and contributed meaningfully to adolescent behavioral and cognitive ability.
Additional File
The additional file for this article can be found as follows:
Table S1
Partial Correlations among Study Variables (controlling for family income, affordability, parents’ education, and subjective financial condition). DOI: https://doi.org/10.5334/paah.567.s1
Acknowledgements
The authors would like to express their sincere gratitude to all students who participated in this study. We also thank the school principals and Physical Education teachers for their support and collaboration throughout the research. Our appreciation extends to our colleagues at Faculty of Sport Science, Universitas Negeri Jakarta for their valuable insights and constructive feedback, as well as to all volunteers who contributed to the data collection and other research activities.
Author Contributions
RN conceived the study and performed the analyses. RN drafted the manuscript. All authors contributed to the study design, interpretation of the data, critically revised the manuscript and supervision. All authors approved the final version of the manuscript.