Physical Activity, Fitness and Working Memory: How Teen Exercise Links to Mathematics Achievement Across Socioeconomic Lines

Table of Contents

1. Key Highlights
2. Introduction
3. A metropolitan sample and a layered measurement approach
4. What the model found: core pathways and effect sizes
5. Interpreting the pathways: converging biological and behavioral explanations
6. Parental education, SES and the social shaping of activity and achievement
7. Practical classroom and school strategies grounded in the findings
8. Real-world examples of movement-to-learning programs
9. Limitations that matter for interpretation
10. Directions for future research
11. What this means for educators and policymakers
12. FAQ

Key Highlights

• A study of 310 Turkish ninth-graders found that physical activity predicts both physical fitness and working memory; working memory is a strong predictor of mathematics achievement.
• Maternal and paternal education levels independently predicted mathematics scores; maternal education also predicted adolescents’ physical activity. Socioeconomic status (SES) showed no direct link to either activity or math in this sample.
• Physical fitness partially mediates the effect of physical activity on working memory, but most of the physical-activity → working-memory relationship is direct, indicating multiple pathways from movement to cognition.

Introduction

Schools and families push math drills, tutoring and study time as the primary route to stronger grades. A growing body of research suggests another lever: movement. Executive control and, specifically, working memory underpin students’ ability to hold intermediate steps, manipulate symbols and solve multistep problems. Simultaneously, physical activity produces physiological and motor adaptations that affect brain systems involved in those same executive operations.

A recent cross-sectional study of 310 fifteen-year-old students across Istanbul set out to map these relationships in a single structural model. The investigators measured socioeconomic indicators and parental education, asked students about weekly physical activity, tested physical fitness with field protocols, assessed working memory using n‑back tasks, and used centralized mathematics exam scores as the academic outcome. Structural equation modeling examined direct links and whether physical fitness serves as a physiological bridge between physical activity and working memory.

The results refine our understanding of how movement and context combine to shape cognitive and academic performance during a pivotal developmental window. They also offer practical signals for educators and policymakers: not all socioeconomic signals operate the same way, and boosting activity and fitness can support the cognitive capacities students use most in mathematics.

A metropolitan sample and a layered measurement approach

Researchers recruited 310 ninth-grade students (mean age 15) from five public high schools selected to represent different socioeconomic districts of Istanbul. The sample was balanced for gender and restricted to students medically cleared to perform physical fitness tests. The research design prioritized multiple, validated measures across environmental, behavioral, physiological and cognitive domains.

Key measures:

• Socioeconomic status (SES): A multidimensional index incorporating household assets, parental education and occupation, income and durable goods. For modeling, SES was treated as a continuous observed variable.
• Parental education: Maternal and paternal education were entered as separate variables to capture cognitive-cultural capital distinct from the broader SES index.
• Physical activity (PA): Collected via the Turkish short form of the International Physical Activity Questionnaire (IPAQ‑SF), producing MET‑minute estimates for walking, moderate and vigorous activities over seven days.
• Physical fitness (PF): Assessed with the ALPHA test battery—20 m shuttle run for cardiorespiratory endurance, standing long jump and handgrip for muscular strength, and the 4 × 10 m shuttle for speed‑agility. PF was modeled as a formative construct with these components entering the latent variable as distinct indicators.
• Working memory (WM): Measured individually with an n‑back protocol (1‑back practice then 2‑back visual, auditory and combined tasks) using Brain Workshop software; performance on 2‑back trials represented working memory capacity.
• Mathematics achievement: Students’ scores on the national High School Transition Exam (LGS) mathematics section provided a standardized outcome.

Data collection took place in classrooms and quiet testing rooms under medical supervision and standardized procedures. Analyses used partial least squares structural equation modeling (PLS‑SEM) with bootstrapping to test direct effects and mediation.

What the model found: core pathways and effect sizes

The structural model produced three broad sets of findings: how socioeconomic and parental-education signals map onto behavior and achievement; how physical activity, fitness and working memory interconnect; and how much of the activity → cognition link operates through fitness.

1. Parental education and SES

• Maternal education positively predicted adolescent physical activity (β = 0.364, p < 0.001). Mothers with higher education levels were associated with higher PA participation in their children.
• Both maternal (β = 0.198, p < 0.05) and paternal (β = 0.273, p < 0.01) education positively predicted mathematics achievement.
• The composite SES index, however, did not predict physical activity (β = −0.155, p = 0.25) or mathematics achievement (β = −0.013, p = 0.915) within this sample.

2. Activity, fitness and working memory

• Physical activity strongly predicted physical fitness (PA → PF β = 0.615 in the final model; earlier staging showed β = 0.505 when mediation tested).
• Working memory robustly predicted mathematics achievement (WM → Math β = 0.663, p < 0.001), so students with higher working memory capacity scored substantially higher in math.
• Physical activity also had a direct positive association with working memory (PA → WM β = 0.689 when PF was not in the model; reduced to β = 0.505 with PF included), meaning activity supports working memory both directly and via fitness.

3. Mediation: fitness as a partial bridge

• Physical fitness had a significant direct positive effect on working memory (PF → WM β = 0.300, p = 0.001).
• Bootstrapped mediation analysis returned an indirect effect (PA → PF → WM) that was statistically significant yet modest, with a Variance Accounted For (VAF) of 0.21. This indicates weak partial mediation: fitness explains some, but not most, of the relationship between activity and working memory.

Model explanatory power was moderate to strong: the predictors accounted for 52% of variance in physical fitness, 48% in working memory and 45% in mathematics achievement. Fit indices were within acceptable ranges for PLS‑SEM.

Interpreting the pathways: converging biological and behavioral explanations

The pattern of results supports a layered interpretation. Physical activity, as a behavioral exposure, associates with better working memory both directly and through the physiological adaptations captured by physical fitness. Several mechanisms explain these linkages.

Cardiovascular and neurotrophic pathways

• Repeated moderate‑to‑vigorous activity elevates cardiorespiratory capacity and cerebral perfusion. Better aerobic fitness correlates with increased hippocampal volume and prefrontal cortex efficiency—regions central to working memory.
• Exercise stimulates synthesis and release of brain‑derived neurotrophic factor (BDNF) and other growth factors that facilitate synaptic plasticity and neurogenesis. These molecular processes improve neuronal signaling and information maintenance capacity.

Motor coordination and psychomotor development

• Agility, balance and visuomotor integration are not incidental to cognition. Psychomotor skills require integration of perceptual, motor and executive systems. Training that sharpens motor coordination can therefore sharpen attentional control and the ability to manipulate visual‑spatial information—elements crucial for numerical reasoning.
• The study’s use of a formative PF model—keeping agility, strength and aerobic indicators separate—captures this multidimensionality. Agility adds behavioral information that aerobic measures do not.

Behavioral and psychosocial pathways

• Physical activity influences motivation, arousal, mood and fatigue levels. Acute and chronic PA can raise alertness, reduce mental fatigue and increase persistence on cognitively demanding tasks, thereby improving working-memory performance in practice.
• Social and environmental factors surrounding activity—team sports, coached exercise, parental encouragement—support routines and skills that indirectly boost academic engagement.

Together, these mechanisms explain why both activity and fitness independently relate to working memory. The weak partial mediation indicates that fitness carries part of the physiological story but that behavioral, psychosocial and possibly direct neural effects of activity remain important. That explains why the direct PA → WM path remains sizable even after accounting for PF.

Parental education, SES and the social shaping of activity and achievement

Disentangling SES from parental education mattered in this model. Parental education—especially maternal—showed direct associations with both activity and academic outcomes, while the broader SES composite did not show direct links in this sample. Interpreting that pattern requires care.

Parental education and household practices

• Education encapsulates attitudes, knowledge and daily practices that shape a child’s routines. Parents with higher education frequently place greater emphasis on organized activities, structured schedules and health behaviors such as regular sport participation. Mothers, often more directly involved in daily logistics for younger adolescents, may act as gatekeepers to extracurricular sports and active commuting.
• Parental education also correlates with active academic support, homework monitoring and enriched home learning environments that directly benefit math skills.

Why SES did not emerge as a direct predictor here

• The SES index captured assets, household resources and income, but in this urban Turkish sample the range of SES variability may have been limited. The study sampled students across different neighborhoods within a single metropolitan area and public-school system; as a result, extreme economic disparities might be underrepresented.
• SES often operates through mediators—school quality, parental involvement, neighborhood safety, access to facilities. If those mediators are unevenly distributed or not directly measured, SES may not show up as a direct effect in a single‑wave model.

Policy and program implications follow from this nuance: improving parental awareness and hands-on involvement (especially maternal engagement) can raise activity levels; targeting structural barriers captured by SES—safe play spaces, subsidized sports programs—requires community-level interventions.

Practical classroom and school strategies grounded in the findings

The model’s implications are actionable. Schools can deploy strategies that increase both activity and fitness in ways likely to support working memory and mathematics achievement.

Short activity breaks and distributed movement

• Short, frequent activity breaks during lessons preserve cerebral blood flow and sustain attention. Randomized crossover trials in adolescent populations have shown that brief activity intervals during prolonged sitting improve working memory performance without disrupting classes.
• Practical rollout: 5–10 minute movement bursts (e.g., dynamic stretching, agility ladders, or brisk stair climbs) between lessons or before math classes. These breaks demand minimal space and staffing and can be integrated into the school day.

Fitness-focused physical education

• PE programs that mix aerobic conditioning (to raise cardiorespiratory fitness) with coordination and agility drills can target both physiological and psychomotor pathways.
• High-quality PE with repeated exposure improves VO2max and motor skills—both of which relate to cognitive benefits. Curricula should include circuit training, group games requiring rapid decision‑making, and skill sequences that promote visuomotor integration.

Targeted sport and afterschool programming

• Organized sports often combine aerobic load and coordination demands. Evidence from youth football and other team sports links regular participation and higher fitness to improvements in processing speed, inhibitory control and working memory.
• To ensure equitable access, schools and municipalities should subsidize participation fees, provide equipment and support transport, especially in neighborhoods where parental work schedules and resources limit extracurricular involvement.

Parental engagement and communication

• School outreach that educates parents about the cognitive benefits of activity can encourage family routines that support exercise, active play and sports participation.
• Messaging should highlight simple, feasible actions: active commuting where safe, family-based recreational activities, and scheduling time for unstructured play that develops motor coordination.

Integration with mathematics instruction

• Because working memory resources are intimately tied to math skill, combining movement with math practice offers an efficient dual intervention. Examples include math games that require physical responses or kinesthetic problem-solving tasks.
• Teachers can scaffold difficulty to offload working memory—use external representations, stepwise prompts and worked examples—while fitness programs strengthen the underlying capacity to handle cognitive load.

Real-world examples of movement-to-learning programs

Several initiatives illustrate how schools and communities can implement the study’s implications.

1. Activity breaks during lectures: In an urban secondary school pilot, teachers introduced three 5‑minute activity bursts during long academic blocks. Students performed dynamic balance and coordination tasks. Over a term, teachers reported improved on-task behavior, and short-term working memory scores rose relative to a control cohort.
2. PE redesign with coordination emphasis: A district converted standard PE lessons into mixed-format sessions combining shuttle runs, balance circuits and ball-handling drills. After one academic year, students demonstrated measurable gains in shuttle-run laps, standing long jump and improvements on visuospatial working-memory tasks.
3. Family sport vouchers and transport: A municipality provided subsidized club membership and weekend transport for students from lower-income neighborhoods. Enrollment in organized sports increased, and participating students recorded higher cardiorespiratory fitness and better attendance, suggesting reduced barriers translate into sustained activity.

These examples combine low-cost, scalable changes—schedule tweaks, curricular redesign, targeted subsidies—that align with the mechanisms identified by the study.

Limitations that matter for interpretation

The study’s design and context constrain causal claims and generalizability.

Cross-sectional design

• Associations observed in a single point in time do not establish causation or directionality. Higher working memory might lead students to engage in more complex sports or follow structured activity plans; reciprocally, activity may enhance working memory. Longitudinal or experimental designs are required to parse direction and magnitude of causal effects.

Self-report for physical activity

• IPAQ‑SF provides a practical snapshot but is prone to recall bias and social desirability, especially in adolescents. Device-based monitoring (accelerometers) would yield more precise measurement of intensity, duration and patterns of movement.

Sample and context

• The sample represents urban public-school ninth-graders in one metropolitan region. Rural populations, private or alternative schooling systems, and international samples may show different SES distributions, parental roles and access to activity resources. The SES index captured household resources but not neighborhood-level variables like park access or perceived safety.

Unmeasured confounders

• Sleep quality, nutrition, mental health, classroom instruction quality and genetic predispositions were not included but could moderate or confound observed links. Future models should incorporate broader lifestyle and biological covariates.

Formative modeling of fitness

• Modeling PF as a formative construct preserved the unique contributions of endurance, strength and agility. However, formative indicators behave differently than reflective measures; interpretation focuses on how each component contributes conceptually rather than interchangeable reliability. This choice aligns with theory but carries statistical nuances.

Directions for future research

Three research priorities would sharpen policy-relevant conclusions.

1. Longitudinal and intervention trials

• Randomized school‑based trials that manipulate activity dose, type and frequency, and track cognitive and academic outcomes over months to years will clarify causal pathways. Interventions that combine activity with cognitive training could test for additive or synergistic effects.

2. Objective activity measurement and physiological mediators

• Wearable accelerometers paired with cardiorespiratory testing, neuroimaging markers (e.g., hippocampal volume, white-matter microstructure), and blood biomarkers (BDNF) will specify the extent to which fitness mediates cognitive changes.

3. Equity-focused implementation research

• Studies that test scalable strategies to expand access—community sports subsidies, transport solutions, safe-play urban design—will determine which structural interventions reduce SES-related barriers to activity and translate into cognitive and educational benefits.

What this means for educators and policymakers

Evidence points to actionable priorities: preserve and enhance physical education with attention to aerobic and motor-skill components, embed brief movement breaks across the school day, and remove access barriers for activity opportunities outside school. Given the strong predictive role of working memory for mathematics, these movement strategies can be framed as investments in both health and core academic skills.

Parental engagement matters. Programs that inform and enable families—especially mothers who often shape daily activities—will amplify school-based efforts. Where socioeconomic constraints limit access, policy must target structural supports: subsidized clubs, afterschool transport and safe public spaces.

FAQ

Q: Will increasing a teen’s physical activity automatically raise their math grades? A: Physical activity associates with better working memory and, through it, stronger mathematics performance. The study found robust links, but causality cannot be proven from cross-sectional data. Well‑designed longitudinal or randomized studies show that structured increases in activity and fitness can improve cognitive function; translating that into grade changes depends on multiple factors including classroom instruction and study habits.

Q: How much and what type of activity matters most for working memory? A: The study used self-reported overall PA and fitness tests capturing aerobic capacity, strength and agility. Evidence suggests both aerobic activity (to boost cardiorespiratory fitness) and coordination-rich exercises (to develop psychomotor integration) support working memory. Short, regular sessions and consistent participation over time matter more than isolated events.

Q: Does family income matter more than parental education? A: Parental education—particularly maternal education—showed specific associations with activity and math in this sample. The composite SES index did not predict activity or math directly here, likely because SES effects often operate indirectly and because sample variability in SES was limited. Income and material resources matter when they constrain access to safe spaces, organized sports and quality instruction; they influence outcomes mainly through these mediating channels.

Q: Should schools cut academic time in favor of more PE? A: Short, integrated activity breaks and a well-structured PE program do not require sacrificing core instruction. Evidence indicates that adding brief movement breaks can enhance attention and working memory without reducing instructional time. Replacing low-value activities (lengthy transitions, passive assemblies) with movement that primes cognition yields better returns than wholesale cuts to academics.

Q: Can parents help at home? A: Yes. Encouraging active commuting where feasible, scheduling regular family physical activities, enrolling children in organized sports, and modeling active behavior increases the likelihood adolescents maintain consistent PA. Informing parents about the cognitive benefits of activity can motivate practical changes in family routines.

Q: What are the next steps for researchers? A: Priorities include longitudinal and randomized intervention studies with objective PA measurement, physiological markers of brain health, and designs that stratify by SES and parental-education levels. Implementation research should test scalable, equity-oriented programs that remove access barriers.

Q: Are there low-cost ways to implement changes in low-resource schools? A: Many effective strategies require little equipment: activity breaks (dynamic movements, mobility drills), playground-based skill circuits, peer-led activity sessions, and community partnerships for shared-field access. Creative scheduling and teacher training amplify impact with minimal cost.

Q: How generalizable are these findings beyond the study’s setting? A: The mechanisms—the role of working memory in math and the cognitive benefits of activity—generalize broadly. However, the specific null effect of SES and the magnitude of parental-education associations may vary by community context, school systems and cultural norms. Local assessments and pilot programs will help tailor strategies.

Q: Does improving fitness always improve cognition? A: Fitness improvements often coincide with cognitive benefits, especially in domains like working memory and attention. That said, fitness is one pathway among several. The study found partial mediation: fitness explains part of the effect of activity on cognition, but direct behavioral and psychosocial influences of physical activity also matter.

Q: If resources are limited, where should schools focus first? A: Prioritize consistent, structured daily movement—short active breaks and quality PE classes that balance aerobic and coordination training. Target outreach to families to encourage regular afterschool activity. Where possible, ensure equitable access to extracurricular sport through subsidies or shared community programs.

Physical activity and physical fitness intersect with cognitive systems central to mathematics. This study clarifies that relationship in adolescence: working memory sits at the juncture, parental education matters, and fitness carries part—but not all—of the physiological explanation. Programs that expand activity opportunities, strengthen fitness components and engage families will therefore address both health and learning goals.