Associations between Physical Activity, Fitness, and Academic Achievement Lydia Kwak, PhD, Stef P.J. Kremers, PhD, Patrick Bergman, PhD, Jonatan R. Ruiz, PhD, Nico S. Rizzo, PhD, and Michael Sjo¨stro¨m, MD, PhD Objectives To explore the associations between objectively assessed intensity levels of physical activity and academic achievement and test whether cardiovascular fitness mediates the association between physical activity and academic achievement. Study design Cross-sectional data were gathered in Swedish 9th-grade students (n = 232; mean age = 16 years; 52% girls). School grades, pubertal phase, skinfold thickness, cardiovascular fitness, and physical activity were measured objectively. Mother’s education, family structure, and parental monitoring were self-reported. Data were analyzed with linear regression analyses. Results After controlling for confounding factors, academic achievement was associated with vigorous physical activity in girls (b = .30, P < .01; explained variance of the model 26%), which remained after inclusion of fitness (b = .23, P < .05; explained variance 29%). The association was not mediated by fitness. In boys, academic achievement was associated with pubertal phase (b = .25, P < .05). After inclusion of fitness, it was only associated with fitness (b = .25, P < .05; explained variance of the model 30%). Conclusion In girls, academic achievement was associated with vigorous physical activity and not mediated by fitness, whereas in boys only fitness was associated with academic achievement. Further studies are necessary to investigate the potential longitudinal effect of vigorous physical activity on academic achievement, the role of fitness herein and the implications of these findings for schools. (J Pediatr 2009;155:914-8). he current debate on the requirement of physical education in schools1 and the substantial pressure on children, parents, and schools to maximize academic performance2 has resulted in a growing interest in a potential benefit of physical activity, namely its impact on educational outcomes.3 Cross-sectional3,4 and intervention studies5-7 have explored a possible relationship between physical activity and academic performance, however, with mixed results. A review on the association between school-based physical activities and academic outcomes revealed that most cross-sectional studies showed positive associations between academic performance and physical activity.8 It is difficult to draw an overall conclusion from previous studies, because many rely on self-reported data of academic achievement and physical activity.3,9 Additionally, variables such as socioeconomic status and mental health factors are not always accounted for.3 Neither is the influence of different intensity levels of physical activity, which according to Coe et al7 is of great importance, because they only observed a significant association between vigorous activity and academic achievement. Academic achievement has also been linked to cardiovascular fitness, both through its positive association3,4,10,11 and through the suggestion that fitness may mediate the association between physical activity and cognitive performance.12 This study explored the associations between objectively assessed light-, moderate-, and vigorous-intensity levels of physical activity and academic achievement, while accounting for several demographic, social, environmental, and biologic factors. Additionally, we tested whether cardiovascular fitness is a mediator in the association between physical activity and academic outcomes.
Methods The data source for this study is the Swedish part of the European Youth Heart Study, a school-based study designed to examine personal, environmental, and lifestyle influences on the risk factors for future cardiovascular diseases in a sample of children (9 to 10 years old) and adolescents (15 to 16 years old). The rationale, study design, sampling procedure, and participation rates have been described elsewhere.13 Briefly, groups of pupils, within each school (n = 42), were randomly selected proportional to the sizes of the respective schools. A total MET
From the Unit for Preventive Nutrition, Department of Biosciences and Nutrition, Karolinska Institute, Huddinge, Sweden (L.K., P.B., J.R., N.R., M.S.), and the Department of Health Promotion and Health Education, University of Maastricht, Maastricht, the Netherlands (S.K.) This study arises from the project, Instruments for Assessing Levels of Physical Activity and related Health Determinants-ALPHA, which received funding from the European Union, in the framework of the Public Health Program. Personal funding was received from the Swedish Council for Working Life and Social Research and the Spanish Ministry of Education (EX-2007-1124). The study sponsors had no involvement in the study design or the collection, analysis, and interpretation of data. The authors declare no potential conflicts of interest, real or perceived.
0022-3476/$ - see front matter. Copyright Ó 2009 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2009.06.019
Vol. 155, No. 6 December 2009 of 1137 pupils (49%) agreed to participate and provided written informed consent by 1 of the parents or legal guardian; additionally all pupils gave verbal consent. The sampling frame was representative of the source population.13 In Sweden, no written marks are given until the last 2 years of compulsory education, therefore only data from the 9th-graders (15 to 16 year olds) are included (61.4%) in this study; 353 students from 14 schools. Complete data on physical activity were available on 232 children (65.7%). A required sample size of 118 children was determined to be large enough to detect a medium effect size of 0.15 with a power of 80% and an alpha of 5%. Students from the complete data set do not differ with regard to sex, sum of skinfolds, or mother’s education from the sample from which they ¨ rebro City Counoriginate. The local ethical committees (O cil no. 690/98, Huddinge University Hospital no. 474/98) approved the study. Academic achievement was assessed through the schools; individual written marks were reported for 17 school subjects (Swedish, English, Biology, Chemistry, Physics, Mathematics, Social Sciences, History, Geography, Religion, Physical Education, Health and 5 additional subjects of preference). For each subject there are 4 categories of credit points (0, 10, 15, and 20), a student can receive a maximum of 340 credit points. The sum of grades for the 17 subjects was used as a continuous variable. The mother’s education was used as a proxy for socioeconomic status14 and was parent-reported. Two dichotomous items (‘‘Is the mother the only parent in the household’’ and ‘‘Is the father the only parent in the household’’ [0 = no, 1 = yes]) were used to assess family structure. Parental monitoring was assessed with 2 items on a 3-point scale (1 = definitely yes, 2 = maybe, 3 = definitely no): ‘‘I always have to tell my parents where I am when I go out?’’ and ‘‘If I am out, I always have to be back by a certain time.’’ The following biologic variables were measured: sex (1 = female, 0 = male), age (in years), pubertal phase, and sum of skinfolds. Pubertal phase was assessed according to Tanner and Whitehouse,15 breast development in girls and genital development in boys was used for pubertal classification. The skinfold thickness of the biceps, triceps, subscapular, suprailiac, and triceps surae were measured in duplicate with the Harpender skinfold caliper on the left side of the body according to the criteria of Lohman et al.16 The measurements were repeated if the difference between the duplicates exceeded 2 mm, accordingly the 2 closest measurements were averaged. The 5 skinfolds were summed. Physical activity was measured with an accelerometer (model WAM 7164; Manufacturing Technology Inc., Shalimar, Florida), worn on the right hip with an elastic waistband. Participants wore the accelerometer during waking hours for 4 consecutive days (2 weekdays and at least 1 weekend day), excluding periods of bathing or other water activities. The monitor was set to accumulate the number of counts in 1-minute epochs. Data were analyzed with a Microsoft Access–based macro. The average minutes per day spent in light (<3 metabolic equivalents
[METs]), moderate (3-6 METs) and vigorous ($6 METs) intensity levels counts were converted to MET values using the cut-off limits published by Trost et al.17 MET is a physiological concept expressing the energy cost of physical activities as multiplies of resting metabolic rate, 1 MET is considered as the resting metabolic rate obtained during quiet sitting.18 Validation studies with accelerometers have shown that the accelerometer is a valid measure of children’s physical activity intensity levels.19 Participants were excluded from the analysis if the accelerometer was not worn for at least 3 days, with a minimum of 10 hours registration per day. Cardiovascular fitness was measured with a bicycle ergometer (Monark 829E; Ergomedic, Vansbro, Sweden), which was preprogrammed to increase the workload by 50 watts for boys and 40 watts for girls, every third minute until exhaustion. Heart rate was measured with a Polar heart rate monitor around the chest. Exhaustion was classified by a subjective assessment by the observer that the pupil could not continue, even after encouragement. To calculate the maximal power output (in Watts [W]) the following formula was used, in which W1 (W) corresponds to workload at the last fully completed stage, W2 (W) to workload at the final incomplete stage and T (s) to time at the last incomplete stage: W1 + (W2-W1)*T/180. Cardiovascular fitness was expressed as the maximal output (W). Differences in baseline characteristics between boys and girls were explored, using Student t tests or c2 tests. To test the associations between the intensity levels of physical activity and academic achievement linear regression analyses were used. Academic achievement was regressed in 2 steps: in step 1 the potential distal variables (mother’s education, family structure, parental monitoring, sex, pubertal phase, and sum of skinfolds) were entered, and in step 2 the proximal variables (physical activity intensity levels variables). To test the hypothesis that fitness mediates the relationship between the proximal variables and academic achievement a stepwise approach was followed on the basis of recommendations by Baron and Kenny.20 For fitness to be a mediator, several conditions have to be met. First, the proximal variables must significantly correlate with academic achievement (B-D pathway) (Figure; available at www.jpeds.com), and second, the proximal variables must be correlated with fitness (B-C pathway). In the third step, it must be shown that fitness is associated with academic achievement after controlling for the proximal variables. To establish whether fitness completely mediates the relationship between the proximal variables and academic achievement, the effect of the proximal determinants on academic achievement controlling for fitness should be zero. If the effect of introducing fitness reduces, then fitness is a partial mediator. Because both physical activity and fitness have been reported to be strongly related to sex, all analyses were stratified for sex. Moreover, potential interactions between the other variables were explored, no significant interactions were found. Statistics were analyzed with SPSS for Windows (version 15.0; 915
THE JOURNAL OF PEDIATRICS
Vol. 155, No. 6
Table I. Descriptive statistics Girls Variable Age (y) Mother’s education Nine-year compulsory education University degree Family structure Lives with both parents Other arrangement Pubertal phase 2 3 4 5 Sum of skinfolds (mm) Physical activity (min/day) Light Moderate Vigorous Fitness (W) Max HR (beats/min) Sum of 17 subjects
0 6 53 50 105
0 5 44.2 41.7 60.9 (19.2)
120 120 120 116 116 120
440 (61.5) 57 (21.9) 12 (11.0) 170 (29.9) 194 (8.1) 246 (52.2)
2 3 18 80 108
1.8* 2.7 16.1 71.4 40.0 (16.0)*
284-594 15-109 0-69 107-256 170-217 60-330
112 112 112 109 109 112
432 (87) 65 (32)† 16 (12)z 242 (42)* 195 (8.3) 215 (58)*
217-620 18-215 1-53 151-350 167-213 60-330
W, Watt; Max HR, maximum heart rate. †P < .05, for differences between boys and girls. zP < .01, for differences between boys and girls. *P < .001, for differences between boys and girls.
SPSS, Inc, Chicago, Illinois) and a P value < .05 denoted statistical significance.
that fitness was not significantly associated with academic achievement, suggesting that in girls the influence of vigorous activity on academic achievement is not mediated by fitness.
Results Discussion The descriptive statistics of the study sample are shown in Table I. The linear regression analyses in girls with academic achievement as dependent variable (Table II) showed that the distal variables explained 16% of the total variance in academic achievement, with only a significant association with age. The inclusion of the proximal variables increased the explained variance to 26%. Only vigorous physical activity was significantly associated with academic achievement, which remained after inclusion of fitness, and increased the explained variance to 29%. In boys, the linear regression analyses with academic achievement as dependent variable showed that in the first 2 steps academic achievement was only significantly associated with pubertal phase. The distal variables explained 24% of the variance in academic achievement, whereas the inclusion of the proximal variables increased the variance to 26%. In the final step of the analyses, only fitness was significantly associated with academic achievement. With the inclusion of fitness, the explained variance increased to 30%. It was only possible to test for a mediation effect of fitness in girls, because in boys there was no significant association between any of the proximal variables and academic achievement. In girls, however, academic achievement was significantly associated with vigorous physical activity. Moreover, vigorous physical activity was significantly associated with fitness (Table III). Introducing fitness to the model showed 916
Vigorous physical activity was the only intensity level that significantly correlated with academic achievement, solely in girls. This suggests that in girls, there is an association between more time spent in vigorous physical activity and higher grades. Our findings are in line with the suggestion of Coe et al7 that a ‘‘threshold’’ level of physical activity intensity is necessary to produce beneficial effects and that vigorous physical activity intensity levels meet this threshold. Our study expands on the research by Coe et al7 by use of objectively assessed physical activity and by controlling for socioeconomic status. Moreover, it provides a potential explanation for the absence of a relationship between physical activity and academic achievement observed in other studies,21 in which no distinction is made between intensity levels. Both physiological and psychological mechanisms have been proposed to be involved in the association between physical activity and academic achievement. Physiological mechanisms include a stimulation of brain growth factors, such as an increase in brain-derived neurotrophic factor22 and in blood flow to the cortex of the brain.23 A factor in the psychological mechanism that might play an important role is self-esteem, through its positive association with physical activity.24 Carlson et al21 imply that factors related to psychological growth, might modify or mediate the relationship Kwak et al
Table II. Linear regression analysis with academic achievement as dependent variable and distal factors (step 1), proximal factors (step 2) and fitness (step 3) as independent factors in girls (n = 120) and boys (n = 112)
Table III. Linear regression analysis with fitness as dependent variable and distal factors (step 1), proximal factors (step 2) as independent factors in girls (n = 120) Fitness
Academic achievement Step 1 Girls Distal factors Mother’s education Family structure Parental monitoring Age Pubertal phase Sum of skinfolds Proximal factors Light activity Moderate activity Vigorous activity Fitness Boys Distal factors Mother’s education Family structure Parental monitoring Age Pubertal phase Sum of skinfolds Proximal factors Light activity Moderate activity Vigorous activity Fitness
.13 .19 .11
.14 .22 .09
.22 .19 .13
.18 .27† .19
.15 .25† .21
.09 .19 .20
.08 .03 .09 .01 .17 .14
.09 .06 .05 .08 .15 .11
.14 .09 .36*
*P < .01.
Distal factors Mother’s education Family structure Parental monitoring Age Pubertal phase Sum of skinfolds Proximal factors Light activity Moderate activity Vigorous activity
*P < .01. †P < .05.
between physical activity and academic achievement differently in boys and girls. This could explain the lack of a significant association between vigorous physical activity and academic achievement in boys in the present study. However, the latter could also be a result of the low variance in vigorous physical activity observed in the studied boys. The lack of a significant association between cardiovascular fitness and academic achievement in girls corresponds with findings of Dwyer et al,25 who were unable to relate their scholastic ratings with cardiorespiratory endurance, when measured with a cycle ergometer. However, many other studies have observed positive associations between cardiovascular fitness and academic achievement,3,4,26 with some studies even showing a stronger relationship in girls than boys.3,27 According to Etnier et al,12 fitness may indeed mediate the relationship between physical activity and cognitive performance. However, because fitness is such an overall measure
of the physiological changes that occur in response to physical activity it might not be a sensitive enough measure.12 Another explanation for the absence of a significant association between cardiovascular fitness and academic achievement in the present study could be the low average fitness level observed in girls.28 This study has several strengths. First, academic achievement, physical activity, cardiovascular fitness, sum of skinfolds and pubertal phase were all measured objectively and were not based on self-report as in many previous studies. Second, the study sample seems to be a representative sample regarding academic achievement, the higher sum of school grades observed in this study girls in comparison to boys corresponds with national Swedish data.29 A third strength is the incorporation of several potential confounders, such as social economic status, family structure, and parental monitoring, which have not all been included in previous research on this topic. Results need to be interpreted with caution, because the study has several limitations. The first limitation pertains to the cross-sectional design of the study, which limits the possibility to draw conclusions regarding the causality of any of the observed relationships. No conclusion can be drawn from this study whether vigorous physical activity or fitness leads to improvements in academic achievement, or whether poor academic achievement results in less time spent in vigorous physical activity, that is, students who perform poorly in academic subjects may be restricted by their parents with regard to how much time they are allowed to devote to sports, dance classes, and other activities. A second limitation is the use of accelerometers, even though seen as a ‘‘golden standard’’; they are limited in capturing any activities with little displacement of the body, such as cycling and snowboarding. This study provides important practical implications for future longitudinal studies, which explore the potential effects of physical activity on academic achievement. On the basis of our cross-sectional study, we recommend that future studies should include different intensity levels of physical
Associations between Physical Activity, Fitness, and Academic Achievement
THE JOURNAL OF PEDIATRICS
activity and take sex into consideration. Studies should investigate the clinical relevance and implications of these findings for schools and evaluate possible interventions for increasing academic achievement through physical activity. n Submitted for publication Mar 12, 2009; last revision received May 5, 2009; accepted Jun 10, 2009. Reprint requests: Lydia Kwak, PhD, Department of Biosciences and Nutrition, Unit for Preventive Nutrition, NOVUM, Huddinge, Sweden, SE-14157. E-mail: [email protected]
References 1. Puehse U, Gerber M. International Comparison of Physical Education Concepts, problems, prospects. Aachen: Meyer and Meyer; 2005. 2. Ahamed Y, MacDonald H, Reed K, Naylor PJ, Liu-Ambrose T, Mckay H. School-based physical activity does not compromise children’s academic performance. Med Sci Sports Exerc 2007;39:371-6. 3. Sigfusdottir ID, Kristjansson AL, Allegrante JP. Health behaviour and academic achievement in Icelandic school children. Health Educ Res 2007;22:70-80. 4. Kim HY, Frongillo EA, Han SS, Oh SY, Kim WK, Jang YA, et al. Academic performance of Korean children is associated with dietary behaviours and physical status. Asia Pac J Clin Nutr 2003;12. 186-92. 5. Shephard RJ, Lavallee H. Academic skills and required physical education: the Trois Rivieres experience. CAHPER. J Res Suppl 1994;1:1-12. 6. Sallis JF, McKenzie TL, Kolody B, Lewis M, Marshall S, Rosengard P. Effects of health-related physical education on academic achievement: project SPARK. Res Q Exerc Sport 1999;70:127-34. 7. Coe DP, Pivarnik JM, Womack CJ, Reeves MJ, Malina RM. Effects of physical education and activity levels on academic achievement in children. Med Sci Sports Exerc 2006;38:1515-9. 8. Trudeau F, Shephard RJ. Physical education, school physical activity, school sports and academic achievement. Int J Behav Nutr Phys Act 2008;5:10. 9. Field T, Diego M, Sanders CE. Exercise is positively related to adolescents’ relationships and academics. Adolescence 2001;36:105-10. 10. McAuley E, Courneya KS, Rudolph DL, Lox CL. Enhancing exercise adherence in middle-aged males and females. Prev Med 1994;23: 498-506. 11. Tremblay MS, Inman JW, Willms JD. The relationship between physical activity, self-esteem, and academic achievement in 12-year-old children. Pediatr Exerc Sci 2000;12:312-23. 12. Etnier JL, Nowell PM, Landers DM, Sibley BA. A meta-regression to examine the relationship between aerobic fitness and cognitive performance. Brain Res Rev 2006;52:119-30.
Vol. 155, No. 6 13. Hurtig Wennlo¨f AH, Yngve A, Sjo¨stro¨m M. Sampling procedure, participation rates and representativeness in the Swedish part of the European Youth Heart Study (EYHS). Public Health Nutr 2003;6:291-9. 14. Kaplan GA, Keil JE. Socioeconomic factors and cardiovascular disease: a review of the literature. Circulation 1993;88:1973-98. 15. Tanner JM, Whitehouse RH. Clinical longitudinal standards for height, weight, height velocity, weight velocity, and stages of puberty. Arch Dis Child 1976;51:170-9. 16. Lohman T, Roche K, Martorell R, editors. Anthropometric standardization reference manual. Champaign, IL: Human Kinetics; 1991. p. 55-70. 17. Trost SG, Pate RR, Sallis JF, Freedson PS, Taylor WC, Dowda M, et al. Age and gender differences in objectively measured physical activity in youth. Med Sci Sports Exerc 2002;34:350-5. 18. Ainsworth BE, Haskell WL, Leon AS, Jacobs DR Jr., Montoye HJ, Sallis JF, et al. Compendium of physical activities: classification of energy costs of human activities. Med Sci Sports Exerc 1993;25:71-80. 19. Trost SG, Ward DS, Moorehead SM, Watson PD, Riner W, Burke JR. Validity of the computer science and application (CSA) activity monitor in children. Med Sci Sports Exerc 1998;30:629-33. 20. Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol 1986;51:1173-82. 21. Carlson SA, Fulton JE, Lee SM, Maynard LM, Brown DR, Kohl HW, et al. Physical education and academic achievement in elementary school: data from the early childhood longitudinal study. Am J Public Health 2008;98:721-7. 22. Cotman CW, Berchtold NC. Exercise: a behavioral intervention to enhance brain health and plasticity. Trends Neurosci 2002;25:295-301. 23. Herholz K, Buskies W, Rist M, Pawlik G, Hollmann W, Heiss WD. Regional cerebral blood flow in man at rest and during exercise. J Neurol 1987;234:9-13. 24. Tremblay MS, Inman JW, Willms JD. The relationship between physical activity, self-esteem, and academic achievement in 12-year-old children. Pediatr Exerc Sci 2001;12:312-23. 25. Dwyer T, Sallis JF, Blizzard L, Lazarus R, Dean K. Relation of academic performance to physical activity and fitness in children. Pediatr Exerc Sci 2001;13:225-38. 26. Chomitz VR, Slining MM, McGowan RJ, Mitchell SE, Dawson GF, Hacker KA. Is there a relationship between physical fitness and academic achievement? Positive results from public school children in the northeastern United States. J Sch Health 2009;79:30-7. 27. Grissom JB. Physical fitness and academic achievement. J Exerc Physiol 2005;8:11-25. 28. Ortega FB, Ruiz JR, Mesa JL, Gutierrez A, Sjo¨stro¨m M. Cardiovascular fitness in adolescents: the influence of sexual maturation status-the AVENA and EYHS studies. Am J Hum Biol 2007;19:801-8. 29. Betyg och studieresultat i gymnasieskolan. Available from: http://www. skolverket.se. Accessed May 4, 2009.
Kwak et al
Figure. Conceptual research model.
Associations between Physical Activity, Fitness, and Academic Achievement