Intervention and longitudinal research

Relationship between physical activity, sport or fitness and cognitive testing or academic test results in children.

Terms

  • +: positive association
  • Ass: association
  • CAT3: Canadian Achievement Test 3
  • EEG: electroencephalogram
  • FCAT: Florida Comprehensive Achievement Test
  • IQ: intelligence quotient
  • PA: physical activity
  • PAQ-C: Physical Activity Questionnaire for Children
  • PE: physical education
  • SES: socio-economic status
  • SOFIT: System for Observing Fitness Instruction Time
  • WRAT3: Wide Range Achievement Test V3.

Ardoy et al., 2014, (Ardoy et al., 2014) University of Granada, Spain

Study design

Intervention study group-randomised controlled trial.

Sample

67 adolescents.

Methods

Control group, experimental group 1 and experimental group 2. CG received usual PE (two sessions/week), experimental group received four PE sessions/week and experimental group 2 received four PE sessions/week of high intensity.

Measures 

  • Cognitive performance (non-verbal and verbal ability, abstract reasoning, spatial ability, verbal reasoning and numerical ability) was assessed by the Spanish Overall and Factorial Intelligence Test.

Results

All the cognitive performance variables, except verbal reasoning, increased more in experimental group 2 than in control group (all P < 0.05). Average school grades (e.g., mathematics) increased more in experimental group 2 than in control group. Overall, experimental group 2 improved more than experimental group 1, without differences between experimental group 1 and the control group.

Association

+

Study limitations

The main limitation of this study was its small sample size and consequent small statistical power.

Chen et al., 2014, (Chen et al., 2014) Yangzhou University, Yangzhou, Jiangsu, People’s Republic of China

Study design

Intervention study, randomised control trial.

Sample

Thirty-four third-grade children and 53 fifth-grade pre-adolescents.

Methods

Pre-adolescent participants were randomly assigned into either an acute exercise group or a control group.

Measures 

  • Participants completed inhibition, working memory, and shifting-related executive function tasks prior to and following the treatment.

Results

Acute exercise benefited three primary aspects of executive function in general, regardless of the pre-adolescent age group, whereas the distinct components of executive function had (Gao et al., 2013) different developmental trajectories.

Association

+

Study limitations

The same instructor was utilised and the experimental process was not blinded, it is possible that the instructor consciously or unconsciously provided information (e.g., tone of voice, instructions, encouragement) differently to participants in each group, which is particularly difficult to control in group testing.

Gao et al., 2013, (Gao et al., 2013)  The University of Minnesota, USA

Study design

A repeated-measures crossover design was used.

Sample

208 Latino school children. Year 1, Grade 4 students.

Methods

Students were assigned to the intervention group and offered 30 minutes of exercise (DDR, aerobic dance) three times per week. Grade-3 and Grade-5 students made up the comparison group and were offered no structured exercise at school. In Year 2, the Grade-4 students were again assigned to the intervention, whereas Grade-5 and Grade-6 students were in the comparison group.

Measures 

  • The baseline measures included time to complete a one-mile run, BMI, and reading and math scores. Data were collected again nine months later. Overall, data were collected in 2009–2011 and analysed in 2012.

Results

The Dance Dance Revolution based exercise intervention improved children’s cardiorespiratory endurance and math scores over time.

Association

+

Study limitations

It was not possible to investigate the potential confounding effect that age and maturation and selection bias had on the outcome variables.

Ericsson & Karlsson, 2014, (Ericsson & Karlsson, 2014) Malmö University, Sweden

Study design

Longitudinal study design.

Intervention.

Sample

All pupils born 1990–1992.

Methods

An intervention group (n = 129) achieved daily PE (5 X 45 min/week) and if needed one extra lesson of adapted motor training. The control group (n = 91) had PE two lessons/week.

Measures 

  • Motor skills were evaluated by the Motor Skills Development as Ground for Learning observation checklist and school achievements by marks in Swedish, English, mathematics, and PE and proportion of pupils who qualified for upper secondary school.

Results

Daily PE and adapted motor skills training during the compulsory school years is a feasible way to improve not only motor skills but also school performance.

Association

+

Study limitations

Lack of separate baseline motor skills data for the intervention and the control group.

Hillman et al., 2014, (Hillman et al., 2014) University of Illinois at Urbana-Champaign, Urbana-Champaign, Illinois, USA

Study design

Intervention study.

Randomised control trial.

Sample

221 children (7–9 years).

Methods

Randomly assigned to a 9-month afte-rschool PA program or a wait-list control.

Measures 

  • In addition to changes in fitness (maximal oxygen consumption), electrical activity in the brain (P3-ERP) and behavioural measures (accuracy, reaction time) of executive control were collected by using tasks that modulated attentional inhibition and cognitive flexibility.

Results

Fitness improved more among intervention participants from pre-test to post-test compared with the wait-list control. The intervention enhanced cognitive performance and brain function during tasks requiring greater executive control. These findings demonstrate a causal effect of a PA program on executive control, and provide support for PA for improving childhood cognition and brain health.

Association

+

Study limitations

The use of a wait-list control renders it difficult to attribute the observed group differences entirely to the PA participation because other aspects of the program such as the educational component, social interaction with peers and intervention staff, and refining motor skills may have contributed to the results.

Macdonald, Abbott, lisahunter, Hay, & McCuaig, 2014, (Macdonald et al., 2014) University of Queensland, Australia

Study design

Intervention study.

Sample

Twelve Year 5 students, their classroom teachers, and the school principal’s perspectives are shared in this paper.

They were key informants from 107 students and five teachers who participated in the intervention.

Methods

Qualitative focus.

Measures 

  • Researcher field observations, along with a diary kept by the dedicated AKAM teacher, were used to interrogate the complexity and pragmatics of both delivering the intervention and succeeding in the intervention.

Results

Data suggested that the intervention group benefited from and welcomed the additional daily physical activity when it offered high time-on-task, fun, and reflected students’ interests. The intervention design with a dedicated physical activity leader and professional development support seemingly promoted teachers’ confidence and enthusiasm.

Association

+

Niemann et al., 2013, (Niemann et al., 2013) Jacobs University Bremen, Germany

Study design

Intervention.

Sample

Primary school students (n = 42, mean age = 9.69, SD = .44; experimental group (EG), n = 27; control group (CG), n = 15).

Methods

The students were randomly assigned to an experimental (EG) and a control group (CG).

Measures 

  • Physical activity
  • Saliva sampling and analysis
  • Cognitive testing.

Results

The results indicate that intensive physical activity only attenuates the reactivity of the hypothalamic-pituitary-gonadal axis in habitually low active preadolescents, but had a beneficial effect on cognitive performance for all participants independent of their physical activity level and testosterone.

Association

+

Study limitations

The size of the saliva sample time.

Points was relatively small (one pre-exercise, one post-exercise).

Riley, Lubans, Morgan, & Young, 2014, (Riley et al., 2014) University of Newcastle, Australia

Study design

Intervention design, randomised control trial.

Sample

Two classes from a single school (n = 54).

Methods

Classes were randomised to receive either the six-week EASY Minds intervention (n = 27) or follow their usual school programme (n = 27). The intervention involved the embedding of PA across the pre-existing mathematics program for 3 × 60 min sessions per week. Changes in PA were measured using accelerometers and ‘on task’ behaviour was measured using momentary time sampling observation.

Measures 

  • EASY Minds program.

Results

The EASY Minds program demonstrated that integrating movement across the primary mathematics syllabus is feasible and efficacious in enhancing school based-PA and improving on-task behaviour in mathematics lessons.

Association

+

Study limitations

The program was delivered by the researcher, a HPE trained specialist, with extensive experience in the primary classroom.

Shachaf, Katz, & Shoval, 2013, (Shachaf et al., 2013) Givat Washington Academic College of Education

Study design

Intervention design.

Sample

491 high school students in Grades 10, 11 and 12.

Methods

Participants divided into three comparison groups.

Measures 

  • Academic achievement test.

Results

Results of the study indicate the existence of a positive relationship between academic achievement and participation in competitive sport for female high school students, while a negative relationship was found between academic achievement and participation in competitive sport for male students. Female athletes who participated in competitive sport attained a higher level of academic achievement in numeracy and verbal reasoning than females who did not participate in competitive sport. On the other hand, the results indicate that male high school students who did not participate in competitive sport attained a higher level of academic achievement in numeracy and verbal reasoning than male athletes who participated in competitive sport.

Association

+ for females
– for males

Study limitations

Did not include the addition of an intervention in order to ascertain whether the relationship between academic achievement and sport at different levels of intensity changes as a result of the intervention.

Tine & Butler, 2012, (Tine & Butler, 2012) Dartmouth College, Hanover, NH, USA

 

Study design

Intervention.

Randomised experimental design.

Sample

Participants (n=164) were sixth and seventh grade students at a public middle school in New England (age range: 10 years, 4 months–13 years, 6 months).

Methods

Stratified sampling was used to randomly assign students to the experimental (exercise) or control (movie) condition. Participants in the experimental condition (n=86; 45 female and 41 male) included 44 lower-income participants and 42 higher-income participants. The control condition (n= 78; 40 female and 38 male) was comprised of 36 lower-income participants and 42 higher-income participants.

Measures 

  • Physical characteristics
  • Heart rate
  • Test of attention.

Results

Improved selective attention.

Association

+

Study limitations

Did not measure/ address how long the selective attention benefits from an acute bout of exercise last for higher- and lower-income children.

Koivusilta, Nupponen, & Rimpela, 2012, University of Turku, Finland (Koivusilta et al., 2012)

Study design

Longitudinal study design.

Sample

The Adolescent Health and Lifestyle Surveys (AHLS), collected biennially in 1981–89 (baseline) and representing 14- and 16-year-old Finns were individually linked with national registries of the highest educational level and SEP.

Methods

Multinomial logistic regression analysis was used to study the associations between the outcomes (highest attained educational level, SEP) and PA (sports clubs, spontaneous, intensity).

Measures 

  • Highest educational level
  • Type of physical activity
  • Childhood socio-economic background
  • School performance.

Results

Participation in physical activity in adolescence and particularly its high intensity, predicts higher educational levels and SEP in early middle age. School performance to some degree mediates the impact of PA. PA behaviours in adolescence—or possibilities to participate in PA—are a potential mechanism in generating better health of higher socio-economic and educational groups in adult age.

Association

+

Study limitations

The boys’ lower participation rates in the surveys led to a slightly higher proportion of women in the follow-up as compared with the entire Finnish population of the same age.

Stevens, To, Stevenson, & Lochbaum, 2008 (Stevens et al., 2008) Texas Tech University, USA

Study design

Longitudinal study.

Sample

Elementary School students.

Methods

Structural equation modelling.

Measures 

  • Socio-economic status
  • Physical activity
  • Physical education
  • Prior mathematics achievement
  • Prior reading achievement.

Results

Physical activity was significantly and positively related to both mathematics and reading achievement in boys and girls. Physical education participation was not significantly related to achievement. Socio-economic status accounted for approximately 26% of the physical activity. Future longitudinal research is discussed that incorporates more comprehensive physical activity and achievement variables.

Association

+

Study limitations

Assessments did not consider the intensity or frequency of each child’s participation in physical education.

Wittberg, Northrup, & Cottrell, 2012, (Wittberg et al., 2012) Mid-Ohio Valley Health Department, Parkersburg, USA

Study design

Longitudinal study.

Sample

Three cohorts of students (n = 50.1% male) enrolled in a West Virginia public school system.

Students.

Methods

n =1,725 received baseline fitness and academic assessments as fifth graders and at a 2-year follow-up assessment.

Measures 

FitnessGram to assess fitness in aerobic capacity and WESTEST, a criterion-based assessment, for academic performance.

Results

Students’ aerobic capacity is associated with greater academic achievement as defined by standardised test scores. This advantage appears to be maintained over time.

Association

+

Study limitations

FitnessGram tests were administered by different physical education teachers who, despite training and review, may have varied administration techniques.

Hollar, Massiah et. al. 2010, (Hollar et al.) University of Miami

Study design

Intervention involving four study schools and one control school.

Sample

This study analysed data from a sub-population incorporating children who qualified for free or subsidised lunches (n=1197). 

Methods

Two-year intervention of dietary and physical activity intervention. Standardised academic test scores examined at the end of each year. Analysis adjusted for school clustering of behaviour and demographics.

Measures 

  • FCAT scores (standardised testing)
  • Demographics
  • BMI.

Results

Overall, children attending intervention schools had significantly higher maths scores in both study years. 

Association

+

Study limitations

Only one control school. As analysis only incorporated  lower SES, results may have limited generalisability.

Hillman, Pontifex, et.al., 2009, (Hillman et al., 2009) University of Illinois

Study design

Intervention incorporating moderate treadmill walking.

Sample

20 pre-adolescents (mean age 9.5, SD 0.5yrs) from Illinois.

Methods

Guardian completed health and demographic questionnaire. Children visited laboratory on two separate days (mean 10, SD 9 days apart) involving either resting session then 20 minute PA session (or order vice versa 50% of children). Tests administered after either rest or PA session.

Measures 

  • EEG at 64 sites
  • Modified flanker test (to assess inhibitory control) 
  • Academic level via WRAT3
  • Cardio-respiratory fitness using indirect calorimetry.

Results

Significantly better performance at reading comprehension after PA session compared with rest. No effect for arithmetic or spelling. Significant improvement in response accuracy and larger P3 amplitude (cognitive control) following PA session only.

Association

+ (reading, cognitive control) 
0 (arithmetic and spelling)

Study limitations

Small sample size. Testing order did not alter during the study and may have affected results (reading, spelling, then arithmetic).

Budde, Voelcker-Rehage, et. al. 2008, (Budde et al., 2008) Humboldt University

Study design

115 children attending an elite performance school in Berlin (mean age 15, SD 0.9 years).

Sample

Intervention with random assignment to coordinative exercise or sport lesson intervention with pre- and post- concentration and attention testing.

Methods

Children randomly assigned to experimental (coordinative exercise) or control group (normal sport lesson). Pre-test before session and post-test after either coordinative exercise or normal sport lesson.

Measures 

  • D2-test (test of concentration and attention).

Results

D2-test results were significantly higher post-exercise intervention (both coordinative and normal sports lesson). Interaction between group by performance thus subsequent.

ANOVA indicated that coordinative exercise led to significantly higher improvement in concentration and attention.

Association

+

Study limitations

No inactive control group. D2-test learning may have occurred thus resulting in higher scores post-intervention.

Davis, Tomporowski, et al.,  2007, (Davis et al., 2007) Medical College of Georgia

Study design

Intervention where children were randomly assigned to low-dose, high-dose exercise program or control condition.

Sample

94 sedentary overweight children aged 7-11 years from Augusta, Georgia.

Methods

Standardised cognitive assessment test was administered before and after intervention. 

Measures 

  • Cognitive Assessment Systems (standardised test for cognitive processes).

Results

Planning scores for high-dose group significantly greater than control.  No difference between low-dose and control.

Association

+

Study limitations

Data from overweight sedentary children only thus may have limited generalisability. Children not blinded to their assignment group.

Ahamed, Macdonald et. al. 2007, (Ahamed et al., 2007)

Study design

16-month cluster randomised controlled trial. Intervention involving Action School BC with pre- and post- academic performance testing. 

Sample

Data from eight schools including 214 children from grades 4 and 5.

Methods

Half schools participated in higher PE each week and thus less academic activity. Control schools maintained usual activity. 

Measures 

  • CAT3 (academic achievement)
  • Teacher logs of PE time
  • PAQ-C to measure child report of PA.
     

Results

Although children spent less time in academic activity in the higher PE schools, this had no significant impact on standardised test scores.

Association

+ (improved learning per unit of time)

Study limitations

Children at the higher scoring schools may have been higher performers. School SES not assessed. PA self-report used.

Coe, Pivarnik, Womack, et. al. 2006, (Coe et al., 2006) Michigan State University

Study design

Intervention study where children were randomly assigned to PE during first or second semester.

Sample

214 grade 6 children attending one Michigan Public School.

Methods

Children were randomly assigned to PE during either first or second semester. When not doing PE, children participated in an exploratory task such as art or computer. 

Measures 

  • Grades (maths, sciences, English, world studies) 
  • Standardised test scores
  • PA assessed using SOFIT and 3 Day Physical Activity Recall.

Results

Although children spent less time in academic activity while enrolled in PE, this had no significant impact on standardised combined test scores. High vigorous activity out of school was significantly associated with higher combined test scores.

Association

+ (improved learning per unit of time)
+ for vigorous activity 

Study limitations

Only one school. No control group. SES not assessed.

Mahar, Murphy et al., 2006, (Mahar, 2006) East Carolina University

Study design

Classroom-based intervention incorporating ‘Energizers’- 10 minutes’ classroom-based PA each day.

Sample

243 kindergarten through to Year 4 children in 15 classes at one school in North Carolina.

Methods

Pre- and post-test of observed on-task behaviour of 3rd and 4th grade students only.

Measures 

  • Pedometers
  • Observation of on-task behaviour for each child every 10 seconds. 

Results

Children in Energizer groups took significantly more steps post intervention compared to control group. Children in the Energizer group also scored better in on-task behaviours post intervention.

Association

+

Study limitations

Pedometers only measured steps not PA intensity. Test performance may have been influenced by other factors (than PA).  

Sallis and McKenzie  1999, (Sallis JF et al., 1999) San Diego State University

Study design

Intervention study  with two experimental groups and one control  group.

Sample

Southern California single school district, seven schools.

Methods

Schools were randomly assigned to PE taught either by specialists, trained teacher or control (class teacher). 

Measures 

  • Achievement test
  • Direct observation of time spent in PE classes.

Results

Children in Specialist and Trained Teacher schools spent significantly less time in non-PE academic and significantly more time doing PE than control schools without impacting on standardised academic achievement test scores.

Association

+ (improved learning per unit of time)

Study limitations

Sample from affluent school district. Measure of PE class time only (no measure sport or time or PA time).

Shephard and Lavallee, 1994 (Shephard RJ et al., 1994) University of Toronto

Study design

Intervention study.

Sample

546 primary school children from an urban and rural school.

Methods

Study group received one additional hour per day of PE, taught by a specialist PE teacher. Controls received 13-14% more academic time than the experimental group. 

Measures 

  • Unweighted average of classroom marks for: French (first language), maths, English, science and mean of all five assessments.

Results

No significant difference in academic achievement detected in first year of study. However, the next year grades 2, 3, 5 and 6 study group students significantly outperformed control group students in academic achievement. Girls gained a larger academic advantage than boys in the enhanced physical education class. 

Association

+ (one year later)

Study limitations

No information regarding the two-year post intervention period prior to follow up. Intervention held at same school, contamination of study and/or control groups may have occurred.

Dwyer, Coonan et al., 1979 (Dwyer T et al., 1979; Dwyer T, Coonon W.E., Leitch D.R., Hetzel B.S., & Baghurst P.A., 1983), University of Tasmania

Study design

Intervention study with two experimental and one control group.

Sample

519 grade 5 (10-year-olds) from seven self-selected schools in Adelaide. Three classes were selected from each school. 

Methods

The three classes randomly allocated to one of three groups:  fitness, skill or control. Intervention took place over 14 weeks. Trained and blinded personnel performing physical measurements and marking tests.

Measures

  • Height and weight
  • Skin-fold thickness
  • Endurance fitness
  • Two measures of academic success (arithmetic and reading tests).

Results

Despite reduction in academic learning time for the fitness and skills groups (210 mins per week, 14% of total teaching time) no significant differences in arithmetic performance or reading skills gains evident. At two-year follow-up, intervention schools had an advantage in teacher ratings of classroom behaviour.

Association

+ (improved learning per unit of time)

Study implications

Short period of observation.