Introduction
PART 1: Methods and Results of the Systematic Review
Methods
1. exp Obesity/ or exp Overweight/ |
2. (overweight or overweight or overweight).tw. |
3. obes*.tw. |
4. exp. Body Mass Index/ |
5. (body mass index or bmi).tw. |
6. exp. Adiposity/ |
7. adipos*.tw. |
8. exp. Child/ |
9. exp. Adolescent/ |
10. child*.tw. |
11. (adolesc* or youth or teen*).tw. |
12. young people.tw. |
13. (students or pupil*).tw. |
14. 8 or 9 or 10 or 11 or 12 or 13 |
15. exp. Education/ |
16. exp. Schools/ |
17. exp. Achievement/ |
18. ((school or academic* or education*) adj2 (attainment or performance or achievement* or outcome*)).tw. |
19. (math* or reading or writing or science).tw. |
20. 15 or 16 or 17 or 18 or 19 |
21. exp Prospective Studies/mt [Methods] |
22. exp Longitudinal Studies/mt [Methods] |
23. cohort.tw. |
24. longitudinal.tw. |
25. prospective.tw. |
26. 1 or 2 or 3 or 4 or 5 or 6 or 7 |
27. exp Cohort Studies/ |
28. 21 or 22 or 23 or 24 or 25 or 27 |
29. 14 and 20 and 26 and 28 |
30. limit 29 to English language |
Results
Reference (quality score) | Sample characteristics | Exposure: measure of body weight status | Outcome: measure of academic achievement | Main results | Confounders |
---|---|---|---|---|---|
Afzal et al. 2015 (58%) | Location: USA Cohort: NLSY
N = 2672 (cohort 1), 1991 (cohort 2) Age: 2–8 years (baseline), 8–16 years (follow-up) Sex: 50.2% (f, cohort 1), 48.8% (f, cohort 2) | BMIa
Persistent obesitye
Developed obesitye
Grew out of obesitye
| Maths, reading Peabody Individual Achievement Test (PIAT) | • n.s. association between change in OB and attainment in boys and girls | Sex, maternal education, maternal ethnicity, maternal obesity, poverty level, Home Observation Measurement of the Environment, child’s height |
Bisset et al. 2012 (85%) | Location: Canada Cohort: Quebec Longitudinal Study of Child Development
N = 1959 Age: 4–7 years (baseline), 8.2 years (follow-up) Sex: 49.7% (f) | BMIa
Overweightc
| Average of reading, math, writing Teacher report | • n.s. association between OW and academic achievement (crude and adjusted model) | Gender, breastfeeding duration, whether low birth weight, socio-familial adversity index, cognitive abilities aged 3y and 7y, internalising/externalising behaviour problems |
Black et al. 2015 (89%) | Location: Australia Cohort: Longitudinal Study of Australian Children
N = 7225 Age: 4–5 years (baseline), 12–13 years (follow-up) Sex: 49% (f) | BMIa
z score Overweightc
Obesityc
| Maths, literacy National Assessment Program—Literacy and Numeracy | • n.s association between OW and maths and literacy in boys and girls • OB among boys is associated with a 0.24 SD (SE (0.081) ↓ maths score and a 0.23 SD (SE 0.089) ↓ literacy score • Significant negative association between BMI z scores and maths (−0.068 SD (SE 0.024)) and literacy (−0.055SD (SE 0.023)) scores in boys • Significantly ↓ maths and literacy scores in girls with OB and for BMI z scores; n.s. association after controlling for cognitive abilities | Child’s age in months, age squared, region of residence, number of younger/older siblings, single-parent family, ethnicity, mother and father’s education level, household income quintiles and mother’s employment status, school type, teacher’s years of experience, low (<2500 g) birth weight, whether breast-fed at 6 months of age, mother’s smoking status while pregnant, maternal mother’s age at Birth, home environment index, cognitive ability |
Booth et al. 2014 (95%) | Location: UK Cohort: ALSPAC
N = 4260 Age: 11 years (baseline), 13 and 16 years (follow-up) Sex: 55% (f) | BMIa
z score Overweightd
Obesed
Developed obesityd
Persistent obesityd
Became healthy weightd
| English, maths, science Standardised National Exams (Key Stage 2, 3, 4) | • Significantly ↓ English, maths and Science grades in girls with OB at 13 and 16 years • ↓ English grades in persistent OW/OB girls and OW➔OB girls • n.s. association for boys with OW/OB and girls with OW • n.s. association for OW−/+, OB−/+, OW/OB +/− | Age, birth weight, gestation; age of mother at delivery, mother’s oily fish intake during pregnancy at 32 weeks gestation, maternal smoking in the first 3 onths of pregnancy; pubertal status, menarche status, maternal education, maternal occupational status, MVPA/week, depressive symptoms, full IQ, BMI z score at age 16 yrs |
Capogrossi et al. 2013 (74%) | Location: USA Cohort: ECLS-K
N = 21,260 Age: 1st grade (baseline), 8th grade (follow-up) Sex: 49% (f) | BMIa
z score | Maths, Reading ECLS-K test based on Woodcock-McGrew-Werder Mini-Battery of Achievement | • Significantly negative association between reading scores in boys in 5th grade and BMI z scores • n.s. association in girls • n.s. association in boys for reading in 8th grade and maths in 5th grade • Positive association between BMI z scores and reading in boys 8th grade | Baseline BMI z score, birth weight, age, baseline attainment score, previous attainment score, urbanicity, whether parents are married, number of siblings, household income, how involved the parent is in the child’s school, how often the parent helps the child with homework, how often the child has changed schools, whether the child has a consistent bedtime, the number of students enrolled in the school, whether the school is Title 1, number of years the teacher has taught and whether the teacher has a Master’s degree |
Carter et al. 2010 (84%) | Location: Canada Cohort: National Longitudinal Survey of Children and Youth
N = 2582 Age: 2–5 years (baseline), 8–11 years (follow-up) Sex: 48.0% (f) | BMIb
Persistent obesityc
Developed obesityc
Grew out of obesityc
| Maths Mathematics Computation Test of CAT/2, shortened version (IRT) | • Significantly ↑ maths scores in O+/− students than O- • n.s. association between O+ and O−/+ and math attainment | Age, gender, ethnicity, sleep, physical activity, chronic condition status, birth weight, household income, family structure; maternal education, working status, age at birth of child, smoking status, degree of positive parenting |
Chen et al. 2012 (68%) | Location: Taiwan Cohort: primary study
N = 409 Age: 6 years (baseline), 12 years (follow-up) Sex: 48.4% (f) | BMIa
Persistent obesityc
Developed obesityc
| Average score of language, maths, science, social studies School records | • n.s. negative association O+ and O−/+ and average attainment | School absenteeism, IQ (Raven’s Colored Progressive Matrices), gender, parental education, number of siblings, family structure |
Cueto 2005 (84%) | Location: Peru Cohort: primary study
N = 438 Age: 12 years (baseline), 14y (follow-up) Sex: 51% (f) | BMIa
| Maths, reading Local attainment test | • n.s. negative association between BMI and maths and reading | Student dropped out, baseline achievement, grade in school |
Crosnoe & Muller 2004 (79%) | Location: USA Cohort: Add Health
N = 11,658 Age: 15 years (baseline), 16 years (follow-up) Sex: 51% (f) | BMIb
Overweighte
| Grade point average of maths, science, English, and social studies Self-reported A-F | • Significantly ↓ average attainment in OW students compared to healthy weight peers | Gender, age, ethnicity, parental education, family structure, prior attainment, athletic activities, educational aspiration, school attendance, homework efforts, participation in non-athletic activities, romantic activities, school SES, minority representation, school mean BMI |
Datar et al. 2004 (84%) | Location: USA Cohort: ECLS-K
N = 11,192 Age: Kindergarten (baseline), 1st grade (follow-up) Sex: 49.93(f) | BMIa
Obesitye
| Math, reading ECLS-K test based on Woodcock-McGrew-Werder Mini-Battery of Achievement | • n.s. negative association between OB and maths and reading attainment in boys and girls | Hours/day watches television or videos, days/week child exercises for ≥20 min, number of activities that parent participates in with child at least once a week, birth weight, baseline test score, ethnicity, mother’s education, family income, urbanicity |
Datar & Sturm 2006 (53%) | Location: USA Cohort: ECLS-K
N = 7090 Age: kindergarten (baseline), 3rd grade (follow-up) Sex:51 (f) | BMIa
Persistent obesitye
Developed obesitye
| Maths, reading ECLS-K test based on Woodcock-McGrew-Werder Mini-Battery of Achievement | • Significantly ↓ maths and reading scores in O−/+ girls compared to O- in girls • n.s. association for O−/+ boys and O+ boys and girls | Gender, age in months in spring of third grade, race/ethnicity, birth weight, annual family income, mother’s education, single-parent household, maternal depression scores, parent–child interaction, hours of television watching, parent-reported physical activity, amount of physical education, school characteristics such as enrolment, percent minority, and urbanicity |
Gable et al. 2008 (56%) | Location: Cohort: ECLS-K
N = 8000 Age: Kindergarten (baseline), 1st and 3rd grade (follow-up) Sex:52.0% | BMIa
Persistent obesitye
Developed obesitye
| Reading, Maths ECLS-K test based on Woodcock-McGrew-Werder Mini-Battery of Achievement | • Significantly ↓ maths and reading scores in O−/+ compared to never-obese children • n.s. association between O+ and maths and reading | Ethnicity and SES |
Gable et al. 2012 (72%) | Location: USA Cohort: ECLS-K
N = 6250 Age: Kindergarten (baseline), 3rd and 5th grade (follow-up) Sex: 52.0% (f) | BMIa
Persistent obesitye
Developed obesitye
| Maths ECLS-K test based on Woodcock-McGrew-Werder Mini-Battery of Achievement | • Significantly ↓ maths scores in O+ and O−/+ boys and girls compared to O- | Child age at study enrolment, ethnicity, maternal education, maternal employment status, parental psychological well-being, parent expectations of child educational achievement, household income, family structure, child’s interpersonal skills, internalising behaviour |
Kenney et al. 2015 (88%) | Location: USA Cohort: ECLS-K
N = 3362 Age: 5th grade (baseline), 8th grade (follow-up) Sex: 49.4% (f) | BMIa
BMI z scorese
Overweighte
Obesitye
| Maths, reading ECLS-K test based on Woodcock-McGrew-Werder Mini-Battery of Achievement Teacher reported: Academic Rating Scale: Poor (1) – Outstanding (5) | • n.s. negative association between change BMI z scores and standardised maths and reading scores in boys and girls • One-unit ↑ in BMI z score significantly associated with a 0.12 SD ↓ in teacher ratings of girls’ reading ability (95%CI:- 0.23,-0.01) • One-unit ↑ in BMI z score significantly associated with a 0.30 SD ↓ in teacher ratings of boy’s maths ability (95%CI:-0.43,-0.17) | Race/ethnicity, SES (annual family income and highest parental education), physical activity, television watching, maternal depression, overall child health, family structure, parent–child interaction, parental disciplinary behaviours |
Kranjac 2015 (89%) | Location: USA Cohort: ECLS-K
N = 5072 Age: 5.7 years (baseline), 14.1 years (follow-up) Sex: 44.6% (f) | BMIa
Overweightc
Obesityc
Developed overweighte
Developed obesitye
| Maths ECLS-K test based on Woodcock-McGrew-Werder Mini-Battery of Achievement | • Significantly ↓ maths score (5.77 points, SE 0.07) in OW • Significantly ↓ maths score (7.97 points, SE0.69) in OB • Effect of obesity on maths achievement is stronger as children age • Maths trajectories of children with OW with high levels of self-efficacy ↑by 3.62 points more than children with OW with low levels of self-efficacy (p < 0.005) • n.s mediating effect of self-efficacy in children with OB | Gender, ethnicity, parental education, self-efficacy |
Li & O’Connelly 2012 (67%) | Location: USA Cohort: ECLS-K
N = 6178 Age: 5.7 years (baseline), 11.2 years (follow-up) Sex: 50.0% (f) | BMIa
Persistent obesitye
| Maths, reading ECLS-K test based on Woodcock-McGrew-Werder Mini-Battery of Achievement | • n.s. negative association between OB and maths/reading | Gender, ethnicity, SES, parental education, mother’s education, single-parent household, primary language at home, school type, school location, percentage minority |
Lu et al. 2014 (68%) | Location: Taiwan Cohort: Taiwan Education Panel Survey
N = 8690 Age: 7th grade (baseline), 9th grade (follow-up) Sex: 50.7 (f) | BMI Overweightg
| Average attainment Comprehensive cognitive ability (CCA) scores | • Significantly ↓ average scores in boys (−1.04, SE 0.57) and girls (−1.66, SE 0.78) with OW | CCA score in the 7th grade, gender, own expected education level and ability education level, family income, parents’ education level, marital status, school location, school type |
Manes 2015 (84%) | Location: USA Cohort: NICHHD Study of Early Child Care and Youth Development
N = 915 Age: 9 years (baseline), 15 years (follow-up) Sex:50.8% (f) | BMIa
| Maths, reading Woodcock-Johnson Psycho- Educational Battery – Revised (WJ-R) | • Significant association between ↑BMI and ↓ reading and maths attainment • n.s. association between BMI and reading and maths attainment after controlling for cognitive abilities (executive function, concentration) | SES, gender, executive function, concentration, internalising behaviour |
Mo-Suwan et al. 1999 (89%) | Location: Hat Yai municipality, southern Thailand Cohort: primary study
N = 2252 Age: 3rd–6th grade (baseline), 7th–9th grade (follow-up) Sex: 53.4% (f) | BMIa
Persistent overweightf
Developed overweightf
Grew out of overweightf
| Grade Point Average in maths and Thai Language Teacher reported | • Significantly ↓ average scores in OW−/+ compared to OW- • n.s. association between OW+ and O+/− and attainment | Age, gender, school, grade |
Murasko et al. 2015 (78%) | Location: USA Cohort: ECLS-B and ECLS-K
N = 9950 (ECLS-B);
N = 18,820 (ECLS-K) Age: B = 16.83 months (baseline), 57.75 months (follow-up); ECLS-K = 5.7 years (baseline), 14.1 years (follow-up) Sex: B = 48.8% (f); K = 48.7% (f) | BMIe
Overweighte
Obesitye
| Maths, reading ECLS-B test items taken from: PreLAS 2000; Peabody picture vocabulary; Preschool Comprehensive Test of Phonological and Print Processing; Test of Early Mathematical Ability-3 ECLS-K test based on Woodcock-McGrew-Werder Mini-Battery of Achievement | ECLS-B: • n.s. association between maths and reading and boys and girls with OW and OB ECLS-K: • Significantly ↓ maths scores in girls (−0.74, SE 0.17); n.s. for girls with OW, • n.s. association between reading and girls with OW and OB • n.s. association between maths and reading in boys with OW and OB | Age, gender, ethnicity, birthweight, household size, presence of mother (resident), maternal age, resident father, paternal age, parents educational level, household income |
Palermo & Dowd 2012 (89%) | Location: USA Cohort: Child Development supplement of Panel Study of Income Dynamics
N = 2820 Age: 8.6 (baseline); 11.69 (1st follow-up) and 13.98 (2nd follow-up) Sex: 50% (f) | BMIa
Overweighte
Obesitye
| Reading Broad reading score of the Woodcock-Johnson Revised Tests of Achievement (WJ-R) | • n.s. association between reading in boys and girls with OW and OB | Race/ethnicity, gender, age, parental education, household income |
Roberts & Hao 2013 (78%) | Location: USA Cohort: Teen Health 2000 study
N = 3134 Age: 11–17 years (baseline), 12–18 years (follow-up) Sex: not reported | BMIa
Overweighte
Obesitye
| Average attainment (good vs poor) Teacher reported | • n.s reduced odds of poor school performance in OW and OB boys and girls | Age, gender, family income, prior academic performance. |
Ruijsbroek et al. 2015 (79%) | Location: The Netherlands Cohort: PIAMA
N = 1531 Age: 3–5 years (baseline), 6–8 years and11 years (follow-up) Sex: 51% (f) | BMIb
Overweightc
Persistent overweightc
Developed overweightc
Grew out of overweightc
| Average of Spelling, maths, study skills and world studies Cito test z scores | • Cito test scores were significantly ↓ children with OW (−0.04 z score (95% CI −0.07; 0.00) per year with overweight) •n.s. association between attainment and OW+, OW−/+ and OW+/− | SES (maternal education level), sex |
Sabia & Rees 2015 (63%) | Location: USA Cohort: Add Health
N = 11,822 Age: 7th grade (baseline), end of high school (follow-up) Sex: 52.2% (f) | BMIb at baseline BMIa at follow-up Overweighte
Obesitye
| Cumulative high school grade point average (GPA) School records | • Significantly ↓ GPA in girls with OW (−0.123 points, SE 0.03) and OB (−0.289 points, SE 0.04) • Significantly ↓ GPA in boys with OB (−0.071, SE 0.04), n.s. association in boys with OW | Parental education, household income, parental marital status, child’s cognitive ability, race, religiosity, age, number of biological siblings, birth order, percentile height-for-age, pubertal development, disability status, and attractiveness of the child’s personality and grooming, self-esteem, depression, |
Suchert et al. 2016 (83%) | Location: Germany Cohort: primary cohort
N = 1011 Age: 14.1 years (baseline), 15.0 years (follow up) Sex: not reported | BMIa
Overweightf/
Obesityf
Developed overweight Persistent overweight Grew out of overweight | Average grade of maths and German Self- reported grades (1–6, lower indicates better) | • Significantly ↓ attainment in OW−/+ (−0.18 grades, 95%CI −0.35; −0.01) • n.s. association between academic attainment and OW and OB • n.s. association between academic attainment and OW+, OW+/− | Sex, age, type of school students attend |
Telford et al. 2012 (71%) | Location: Australia Cohort: primary cohort
N = 757 Age: 8.5 years (baseline), 10.5 years (follow-up) Sex: 47.0% (f) | Body Fat (DEXA) | Maths, reading, writing Local government education authority | • n.s. association between %BF and attainment | SES |
Veldwijk et al. 2012 (83%) | Location: The Netherlands Cohort: PIAMA
N = 1543 Age: 8 years (baseline), 12 years (follow-up) Sex: 51% (f) | BMIa at 8y BMIb at 12y Overweightc
Persistent overweightc
Developed overweightc
| Average of spelling, maths, study skills and world studies Cito test z scores | • n.s. negative association OW, OW+ and OW−/+ and average attainment | Gender, maternal smoking, maternal age at birth, breastfeeding duration, birth weight, parental education, lifestyle factors (physical activity, screen time, breakfast skipping), child’s psychological health, being bullied, school absenteeism due to illness |
von Hinke Kessler Scholder et al. 2012 (82%) | Location: UK Cohort: ALSPAC
N = 3001 Age: 9 and 11 years (baseline), 11 and 14 years (follow-up) Sex: 51% (f) | Fat mass (DEXA) at age 9 and 11 adjusted for height, height 2, gender, age | Average of English, maths, science National exams (Key Stage 2 and 3) | • Significantly negative association between fat mass at 11 years and average scores at age 14 years • n.s association between BF at 9 years and average scores at age 11 years | Birth weight, number of siblings, age, family income, mother’s education, whether mother smoked or drank alcohol during pregnancy, mother’s mental health, maternal age at birth, length of breast feeding, mother’s parents education, raised by natural father, family’s social class, parental employment status, parental involvement in child development, area deprivation |
Wendt 2009 (95%) | Location: U.S.A Cohort: ECLS-K
N = 12,719 Age: Kindergarten (baseline), 3rd grade (follow-up) Sex: 49% (f) | BMIa
Persistent obesitye
Developed obesitye
| Reading, maths ECLS-K test based on Woodcock-McGrew-Werder Mini-Battery of Achievement | • Significantly ↓ maths scores in OW, O+ and O−/+ boys and girls • n.s association between O+ and O−/+ and reading in boys and girls | Birth weight, number of school changes, frequency student reads by him/herself /week, learning problems, bedtime, TV viewing, hours of non-parental care/ week, ethnicity, gender, grade level, parent’s involvement with student’s school activities, student lives with both biological parents, number of places a student lives during the last year of interview time, parents’ expectation for student’s schooling, number of siblings, SES, household has computer for student’s use, teacher enjoys teaching, teacher’s years of teaching, teacher’s degree, teacher is White, private school, % of student in school tested at or above grade level on national standardised, school experiences problems of teacher’s turn-over rates, School’s size, % of minority students, school location, security problems |
Zavodny et al. 2013 (78%) | Location: USA Cohort: ECLS-K
N = 18,820 Age: 1st grade (baseline), 8th grade (follow-up) Sex: 49% (f) | BMIa
Overweighte
Obesitye
| Language, reading, maths, science ECLS-K test based on Woodcock-McGrew-Werder Mini-Battery of Achievement | • Significantly ↓ maths and reading/language scores in OB students compared to healthy weight peers • n.s association between OB and science • n.s association between OW and maths, reading and science | Sex, race/ethnicity, birth weight, foreign-born status, hours of television watched/week; number of siblings, SES, school region, urban/ suburban/rural, public/private non-religious/ Catholic/ other religious, percent minority students, percent of students receiving free lunch; teachers’ age, years of teaching, teachers’ education, indicator variable for the teacher and child being the same race/ethnicity |