Skip Navigation LinksHome > March 2012 - Volume 44 - Issue 3 > Changes in Cardiorespiratory Fitness in 9- to 10.9-Year-Old...
Medicine & Science in Sports & Exercise:
doi: 10.1249/MSS.0b013e3182300267

Changes in Cardiorespiratory Fitness in 9- to 10.9-Year-Old Children: SportsLinx 1998–2010


Free Access
Article Outline
Collapse Box

Author Information

1Research into Exercise, Activity and Children’s Health Group, Liverpool John Moores University, Liverpool, UNITED KINGDOM; 2Faculty of Education, Community and Leisure, IM Marsh Campus, Liverpool John Moores University, Liverpool, UNITED KINGDOM; and 3Research Institute for Sport and Exercise Sciences, School of Sport and Exercise Sciences, Faculty of Science, Liverpool John Moores University, Liverpool, UNITED KINGDOM

Address for correspondence: Lynne M. Boddy, Ph.D., Tom Reilly Building, Liverpool John Moores University, Byrom Street, Liverpool L3 3AF, United Kingdom; E-mail:

Submitted for publication April 2011.

Accepted for publication July 2011.

Collapse Box


Purpose: The study’s purposes were to 1) assess changes in cardiorespiratory fitness (CRF) from 1998 to 2010, controlling for decimal age and body mass index (BMI), and 2) repeat the analysis in cohorts from 2005 to 2010, controlling for maturation, deprivation, and BMI.

Methods: A total of 27,942 (n = 14,247 boys) 9- to 10.9-yr-old participants from one UK city were included in this serial cross-sectional study from 1998–1999 to 2009–2010. An indices of multiple deprivation (IMD) score was assigned to each participant on the basis of home postcode. Stature, sitting stature, and body mass were estimated. BMI and somatic maturity were calculated. Performance on the 20-m multistage shuttle run test (20mSRT) was used to estimate CRF (total shuttles). One-way ANCOVAs were completed to assess temporal trends in CRF, separately by sex. Model 1 assessed changes from 1998 to 2010 and included decimal age and BMI as covariates. Model 2 assessed changes from 2005 to 2010 and included maturity, IMD, and BMI as covariates.

Results: Results indicate that 20mSRT performance has declined in UK schoolchildren. An annual decline of 1.34% and 2.29% was observed in boys and girls, respectively. In model 1, for boys, the baseline cohort performed better than all other groups with the exception of the 1999–2000 group. For girls, declines in 20mSRT performance were observed from 2003 onward. In model 2, for boys, the 2007–2008 and 2008–2009 cohorts completed fewer 20mSRT shuttles than all other groups. For girls, the 2007–2008 and 2008–2009 cohorts also performed worse than all other years.

Conclusions: The decline in CRF suggests children in the more recent SportsLinx cohorts may be at an increased risk of cardiometabolic illness in comparison with earlier cohorts. The promotion of vigorous physical activity is urged to promote CRF in children.

Cardiorespiratory fitness (CRF) is cardioprotective, with a substantial body of evidence suggesting that low fitness is associated with elevated clustered cardiometabolic risk in children and adolescents (3,10,11). The Chief Medical Officer in England recently highlighted the importance of fitness as a health determinant in children and advocated monitoring the fitness levels of children on a large scale (9). Evidence from the United Kingdom and internationally suggests that levels of CRF have declined in children in recent years (22,24,26), despite evidence suggesting that the prevalence of obesity and overweight has reached a plateau (4). Population-level studies assessing fitness in children are relatively rare in the United Kingdom, particularly studies that account for confounders within analyses. Studies that have controlled for body size and indices of multiple deprivation (IMD) describe declines in CRF independent of these confounders (4,24), suggesting that decrements are not entirely due to the widely documented increased prevalence of childhood obesity and overweight.

Accurately assessing CRF in the field is problematic. One commonly used method is the 20-m multistage shuttle run test (20mSRT), also known as the Pacer or bleep test. The appropriateness and reliability of the 20mSRT have been questioned in the past, with motivation, perceived worth, and competency being cited as variables that influence performance (19). However, a recent systematic review discussed the utility of the 20mSRT and concluded that the test is reliable and valid for use in pediatric studies (7). Furthermore, in large-scale field-based studies, the 20mSRT offers a cost-effective and low-resource alternative to the gold-standard laboratory-based assessments of peak oxygen uptake that are impractical on a large scale.

Many fitness testing programs operate within schools, with children grouped into school years on the basis of date of birth. A 1-year cohort (September 1 and August 31) of children can differ in chronological age by nearly 12 months. This will have implications for performance and the interpretation of fitness test scores. Furthermore, within a large cohort, there will inevitably be some variation in physical maturation between participants independent of chronological age. In addition to maturation, deprivation is an established confounding variable when assessing fitness (15) and body composition (23) in children. For example, evidence from the AVENA study describes a positive association between socioeconomic status and muscular strength in boys and CRF, muscular strength, and agility in girls (15). When attempting to monitor population-level changes in CRF, it is important to take age and/or physical maturation and socioeconomic status into account as well as other more routine confounders such as body mass index (BMI). To date, no population studies have assessed temporal trends in fitness over a long period, controlling for other relevant confounders, such as maturation. The Liverpool SportsLinx project (25) annually assesses components of fitness in 9- to 10.9-yr-old (school year 5) schoolchildren across the city. The project typically includes ∼80% of 9- to 10.9-yr-old children within the Liverpool Local Education Authority, and the methods have remained consistent from 1998 to present. In 2004, SportsLinx began collecting home postcode data, allowing an estimate of deprivation to be calculated for each participant. In 2005, sitting stature was introduced into the battery of field-based tests to allow an estimate of somatic maturity to be calculated, using regression equations (18). Because of the lack of population-level studies investigating changes in CRF and, in particular, the absence of any studies that account for maturation within analyses, the purposes of this serial cross-sectional study were the following:

1. To assess changes in CRF from 1998 to 2010 using SportsLinx data, controlling for decimal age and BMI.

2. To repeat the analysis in cohorts from 2005 to 2010, controlling for maturation, indices of deprivation, and BMI.

Back to Top | Article Outline


After gaining local research ethics committee approvals, written informed parental consent, participant assent, and medical screening, 27,942 participants (n = 14,247 boys) were included in this serial cross-sectional study. All primary schools across the Liverpool Local Education Authority were invited to take part in the SportsLinx program each academic year from 1998 to 2010. The SportsLinx field-based fitness testing procedures have been described previously (24,25), and the measures included in this study have acceptable test–retest reliability (5). Briefly, children attend a local authority sports hall with their school class to complete a battery of fitness tests adapted from the Eurofit battery (1). Two senior fitness officers lead the testing sessions throughout the 1998–2010 periods, assisted by undergraduate students.

Back to Top | Article Outline

An IMD score was assigned to each participant on the basis of home postcode using the National Statistics Postcode Directory database (8) and the GeoConvert application (17). IMD scores provide an estimate of area deprivation and comprise seven domains of deprivation. IMD scores are a commonly used method of estimating deprivation in similar studies within the United Kingdom (12). Stature and sitting stature were estimated to the nearest 0.1 cm (Seca, Bodycare, Birmingham, United Kingdom), and body mass was estimated to the nearest 0.1 kg (Seca, Bodycare) by two trained fitness officers using standard techniques (16). BMI was calculated (body mass (kg)/stature squared (m2)). Leg length was calculated by subtracting sitting stature from whole-body stature. Somatic maturity was estimated using decimal age, stature, sitting stature, leg length, and body mass using regression equations. This method has been used in similar-aged children in the past (13) and demonstrates acceptable agreement with skeletal age (18). Somatic maturity was expressed as time to peak height velocity (TPHV). The number of completed shuttles on the 20mSRT (the Queen’s University of Belfast protocol [6]) was used as an estimate of CRF.

Back to Top | Article Outline
Statistical analyses

CRF data were not normally distributed and were log transformed. CRF data were back transformed for presentation purposes. One-way ANCOVAs, with Bonferroni post hoc analyses, were completed to assess differences in CRF between the year cohorts. Analyses were completed using two models: one including all cohorts from 1998 to 2010 and the second including cohorts from 2005 to 2010. For the first model, decimal age (in the absence of TPHV data) and BMI were included as covariates. For the second model, TPHV, BMI, and IMD were included as covariates. Decimal age was removed as a covariate in model 2 to avoid collinearity because decimal age is included in the calculation for TPHV. Annual decline in number of completed shuttles was estimated by plotting the adjusted 20mSRT values from 1998 to 2010 and recording the slope of the regression line. Percentage decline in performance was calculated by fitting a regression of time versus 20mSRT (log transformed); the slope was then antilogged and subtracted from 1 to give the percent annual decline. All statistical tests were completed separately by sex, and an α value of P ≤ 0.05 was used to detect statistical significance. Analyses were conducted using SPSS v 17.0 (SPSS, Inc., Chicago, IL).

Back to Top | Article Outline


Table 1 displays the unadjusted baseline characteristics (plus SD) from 1998 to 2010.

Table 1
Table 1
Image Tools

After adjusting for decimal age and BMI, ANCOVA found many significant differences between cohorts. Table 2 displays adjusted mean 20mSRT performance, percentage change from baseline (calculated from log-transformed data), and number of participants within each cohort, plus the equivalent distance in meters and end running speed (calculated from adjusted mean shuttles) from 1998–1999 to 2009–2010. A decline in 20mSRT performance over time was observed. The mean annual decline in performance was 1.34% or 0.58 shuttles (11.6 m) and 2.29% or 0.71 shuttles (14.2 m) for boys and girls, respectively.

Table 2
Table 2
Image Tools

For boys, the baseline cohort performed better than all other groups with the exception of the 1999–2000 to 2002–2003 groups (P < 0.001). Apart from the baseline group, the 1999–2000 cohort outperformed all the other cohorts (2000–2001, P = 0.001; 2001–2002, P = 0.031; 2003–2005, P = 0.043; all other cohorts, P ≤ 0.001). The 2003–2004 cohort performed worse than all groups with the exception of the 2006–2007 to 2008–2009 cohorts (2004–2005, P = 0.006; 2005–2006, P = 0.023; 2009–2010, P = 0.01; all other cohorts, P ≤ 0.001). The 2007–2008 group performed worse than all other cohorts with the exception of the 2003–2004 and 2008–2004 and 2008–2009 groups (all P < 0.001). The 2008–2009 group performed worse than all groups with the exception of the 2003–2004, 2006–2007, and 2007–2008 cohorts (2009–2010, P = 0.002; all other cohorts, P < 0.001). The most recent cohort performed better than the 2003–2004 (P = 0.01), 2007–2008 (P < 0.001), and 2008–2009 (P = 0.002) groups but poorer than the 1998–1999 (P < 0.001) and 1999–2000 (P < 0.001) cohorts.

For girls, significant differences in 20mSRT performance from baseline were observed from 2003 onward (all P < 0.001). In addition, all cohorts from 2003 to 2004 onward performed worse than all 1998–2003 cohorts (2008–2009, P = 0.009; all other groups, P ≤ 0.001). The 2007–2008 cohort performed worse than all other cohorts with the exception of 2003–2004, 2006–2007 and 2008–2009 (2005–2006, P = 0.02; all other groups, P < 0.001). The 2008–2009 cohort performed significantly worse (2003–2004, P = 0.009; 2005–2006, P = 0.004; all other groups, P ≤ 0.001) than all years with the exception of 2007–2008. The most recent cohort performed better than the 2007–2008 and 2008–2009 groups.

After controlling for TPHV, IMD, and BMI, ANCOVAs detected additional statistically significant differences between the 2005–2010 groups. Table 3 displays the adjusted means, percentage change from baseline, number of participants within each cohort, distance and end running speeds, and differences between cohorts. For boys, the 2007–2008 and 2008–2009 cohorts completed significantly fewer 20mSRT shuttles than all other groups (all P < 0.001). For girls, the 2007–2008 and 2008–2009 cohorts performed worse than all other years (all P < 0.001).

Table 3
Table 3
Image Tools
Back to Top | Article Outline


There is a lack of population-level studies that assess serial cross-sectional changes in CRF in children within the United Kingdom. Furthermore, no population-level studies have controlled for maturation within analyses. Consequently, the purposes of this study were to assess changes in CRF from 1998 to 2010 using SportsLinx data, controlling for decimal age and BMI, and to repeat the analysis in cohorts from 2005 to 2010, controlling for maturation, indices of deprivation, and BMI. Findings from the 1998–2010 analysis describe a significant decline in CRF over time, with all boys’ cohorts from 2000 to 2010 and girls’ cohorts from 2003 to 2010 performing significantly worse than baseline. For boys, the largest decrement in performance was observed in 2003–2004 and 2007–2008, with a 5.6% difference (calculated from log-transformed data) between these cohorts and baseline, which equated to ∼3 shuttles or 60 m. For girls, the poorest performance was observed in 2008–2009, a mean decline of 8.1% from baseline, which equated to a reduction of ∼3 shuttles or 60 m. The adjusted baseline mean value for boys (47.1 shuttles) equates to the 63rd percentile for the 2009–2010 cohort, and vice versa, the 2009–2010 adjusted mean (42.1 shuttles) equates to the 42nd percentile in 1998–1999. For girls, the baseline adjusted mean value (34.1 shuttles) equates to the 72nd percentile for the 2009–2010 cohort; alternatively, the most recent cohort’s adjusted mean value (29.1 shuttles) equates to the 36th percentile at baseline. The overall declines reported are likely to be clinically meaningful, with previous research highlighting a linear relationship between fitness and cardiometabolic risk and increases in risk across each declining quartile of fitness (2,11). The mean declines in fitness of 1.34% for boys and 2.29% for girls observed in the study are greater than those reported internationally (26), a finding that mirrors previous research in the United Kingdom (22). In the absence of longitudinal studies, it is difficult to explain this finding. Because the analyses accounted for BMI, the declines are not likely to be due to changes in body fatness but may be due to insufficient vigorous physical activity. Furthermore, the declines were calculated using the number of completed shuttles rather than end running speed. Our findings suggest that end running speed did not change substantially over time with the exception of a decline of 0.5 km·h−1 between 1998–1999 (11.0 km·h−1) and all other years (10.5 km·h−1) for boys and a 0.5-km·h−1 decline from 10.0 km·h−1 to 9.5 km·h−1 from 2003 to 2004 onward in girls. Therefore, children from the latter cohorts attained similar running speeds as the earlier groups but could not complete as many laps at the given speeds. Further, longitudinal research is required to assess changes in fitness and offer plausible explanations for the observed declines.

For the second analysis, differences that were not apparent in the initial analyses were detected between the 2005–2010 cohorts. The 2007–2008 and 2008–2009 cohorts performed significantly worse than all other years. Findings suggest that the 2009–2010 groups showed an improvement in CRF in comparison with the 2007–2009 cohorts. It remains to be seen whether the improvement in performance in the most recent cohort is a sign of a plateau or even an improvement in CRF, and more years’ data are required to investigate trends. To our knowledge, this is the first serial cross-sectional population-level study to assess changes in CRF in a large sample of schoolchildren after controlling for somatic maturity as well as other confounders.

There are several limitations within this study. Primarily, the 20mSRT was used to estimate CRF. The utility of the 20mSRT as a reliable and valid method of assessing CRF has been questioned in the past. However, recent studies and systematic reviews have concluded that the 20mSRT is valid and reliable for use with children (5,7,22), particularly on a large-scale where laboratory assessments of peak oxygen uptake are impractical. A second limitation of the study is the testing environment; variables such as time of day were not controlled and may have affected 20mSRT performance. Another limitation is the unequal distribution of participants across the cohort years. The varied numbers reflect resource limitations, lack of school uptake for the testing program due to school inspections, and facility constraints; there is also a lack of information on those children and schools that declined the invitation to participate; therefore, sampling bias cannot be estimated. Despite the low sample size within some of the cohorts (n = ∼490 for the smallest cohorts), the “small” cohorts compare favorably with other studies reporting secular trends in CRF (22). Furthermore, the total sample size is substantial and represents a population-level study involving ∼80% of 9- to 10.9-yr-old children for the majority of the cohorts. An additional limitation is that all participants were from the Liverpool Local Education Authority catchment area, which is made up of several areas of high deprivation and lacks any significant areas of affluence or rural communities. Therefore, results may only be applicable to similar urban populations. Finally, TPHV was used to estimate maturation from 2005 to 2010. The inclusion of maturation within model 2 resulted in the ANCOVA detecting additional statistically significant differences between groups, but the overall results were not practically different from model 1. The age group of participants included in this study was predominantly prepubertal, and therefore, the influence of maturation may be less substantial than if the participants were older. Furthermore, the assessment of somatic maturation represents a proxy measure that is less invasive and resource intensive than other methods such as the Tanner scale but also potentially less accurate. Despite being a proxy for maturation, this approach demonstrates acceptable agreement with skeletal age (18) and has been used in studies with similar-aged children in the past (13,14).

CRF is a key determinant of health, and low CRF is independently associated with increased clustered cardiometabolic risk in children (3,21). This is the first study to highlight changes in fitness at the population level after controlling for relevant confounders. The substantial declines in CRF observed in this study suggest that children from more recent cohorts may be at an increased risk of cardiovascular disease in comparison with the earliest cohorts that participated in the SportsLinx program. The upturn in performance observed in 2009–2010 is a promising finding and may reflect the numerous physical activity and health promotion interventions being delivered locally and nationally. Further years’ data are required to detect any changes in fitness at a population level. Investment in physical activity interventions that target improvements in CRF as well as improving energy expenditure is encouraged. Because CRF is a product of vigorous physical activity (20), improvements in CRF may be achieved by placing more emphasis on the intensity of physical activity within existing and planned physical activity policies and programs.

The SportsLinx project is funded by the Liverpool City Council and Liverpool John Moores University.

The authors can disclose that there are no conflicts of interest relating to this study.

All authors had full access to the data and can take responsibility for the accuracy and integrity of the data and analysis. G.S. designed the SportsLinx project. L.M.B., S.J.F., and G.S. drafted and edited the article. L.M.B. analyzed the data, and G.A. gave expert input on analysis. Liz Lamb managed the SportsLinx program, and Glen Groves acted as senior fitness officer. Allan Hackett led the nutrition part of the SportsLinx study, and Sue Taylor worked as a researcher on the project up to 2004. Liverpool John Moores Sports Development students assisted fitness testing. G.S. is the guarantor of the article.

The authors thank Liz Lamb, Principal Health and Physical Activity Officer, and Glen Groves, senior Fitness Officer, Liverpool City Council; Dr. Allan Hackett, reader in community nutrition at Liverpool John Moores University; Dr. Sue Taylor, senior lecturer at Glyndwŵr University; and the schools, children, and parents involved in the study.

The results of this study do not constitute endorsement by the American College of Sports Medicine.

Back to Top | Article Outline


1. Adam C, Klissouras V, Ravazzolo M, Renson R, Tuxworth W. EUROFIT: European Test of Physical Fitness. Rome (Italy): Council of Europe, Committee for the Development of Sport; 1998. p. 72.

2. Andersen LB, Hasselstrom H, Gronfeldt V, Hansen SE, Froberg K. The relationship between physical fitness and clustered risk, and tracking of clustered risk from adolescence to young adulthood: eight years follow-up in the Danish Youth and Sport Study. Int J Behav Nutr Phys Act. 2004; 1: 6.

3. Andersen LB, Sardinha L, Froberg K, Riddoch C, Page A, Anderssen S. Fitness, fatness and clustering of cardiovascular risk factors in children from Denmark, Estonia and Portugal: the European Youth Heart Study. Int J Pediatr Obes. 2008; 3: 58–66.

4. Boddy LM, Hackett AF, Stratton G. Changes in fitness, body mass index and obesity in 9–10 year olds. J Hum Nutr Diet. 2010; 23: 254–9.

5. Boddy LM, Stratton G, Hackett AF. The test/re-test reliability of a field-based fitness test battery in 9–10 year old schoolchildren. In: Baquet G, Berthoin S, editors. Children and Exercise XXV: The Proceedings of the 25th Pediatric Work Physiology Meeting. London (UK): Department of Health; 2010. p. 213–6.

6. Boreham C, Savage JM, Primrose D, Cran G, Strain J. Coronary risk factors in schoolchildren. Arch Dis Child. 1993; 68: 182–6.

7. Castro-Pinero J, Artero EG, Espana-Romero V, et al.. Criterion-related validity of field-based fitness tests in youth: a systematic review. Br J Sports Med. 2010; 44: 934–43.

8. National Statistics Postcode Directory [Internet]. [cited 2010 Nov 22]. Available from:

9. Department of Health. Annual Report of the Chief Medical Officer. Crown Copyright; 2009. p. 27.

10. Eisenmann JC, Katzmarzyk PT, Perusse L, Tremblay A, Despres JP, Bouchard C. Aerobic fitness, body mass index, and CVD risk factors among adolescents: the Québec family study. Int J Obes (Lond). 2005; 29: 1077–83.

11. Ekelund U, Anderssen S, Froberg K, Sardinha L, Andersen LB, Brage S. Independent associations of physical activity and cardiorespiratory fitness with metabolic risk factors in children: the European Youth Heart Study. Diabetologia. 2007; 50: 1832–40.

12. Fairclough SJ, Boddy LM, Hackett AF, Stratton G. Associations between children’s socioeconomic status, weight status, and sex, with screen-based sedentary behaviours and sport participation. Int J Pediatr Obes. 2009; 4: 299–305.

13. Fairclough SJ, Boddy LM, Ridgers ND, Stratton G, Cumming SP. Biological maturity and primary school children’s physical activity: influence of different physical activity assessment instruments. Eur J Sports Sci. 2011; 11 (4): 241–8.

14. Hopkins ND, Stratton G, Tinken TM, et al.. Relationships between measures of fitness, physical activity, body composition and vascular function in children. Atherosclerosis. 2009; 204: 244–9.

15. Jimenez-Pavon D, Ortega FB, Ruiz JR, et al.. Influence of socioeconomic factors on fitness and fatness in Spanish adolescents: the AVENA study. Int J Pediatr Obes. 2010; 5: 467–73.

16. Lohman T, Roche AF, Martorell R. Anthropometric Standardization Reference Manual. Champaign (IL): Human Kinetics; 1988. p. 3–13.

17. Mimas. Welcome to GeoConvert [Internet]. [cited 2010 Nov 1]. Available from:

18. Mirwald R, Baxter-Jones A, Bailey D, Beunen G. An assessment of maturity from anthropometric measurements. Med Sci Sports Exerc. 2002; 34 (4): 689–94.

19. Naughton GA, Carlson JS, Greene DA. A challenge to fitness testing in primary schools. J Sci Med Sport. 2006; 9: 40–5.

20. Ortega FB, Ruiz JR, Castillo MJ, Sjostrom M. Physical fitness in childhood and adolescence: a powerful marker of health. Int J Obes (Lond). 2008; 32: 1–11.

21. Ruiz J, Ortega F, Rizzo N, et al.. High cardiovascular fitness is associated with low metabolic risk score in children: the European Youth Heart Study. Pediatr Res. 2007; 61: 350–5.

22. Sandercock G, Voss C, McConnell D, Rayner P. Ten year secular declines in the cardiorespiratory fitness of affluent English children are largely independent of changes in body mass index. Arch Dis Child. 2010; 95: 46–7.

23. Shrewsbury V, Wardle J. Socioeconomic status and adiposity in childhood: a systematic review of cross-sectional studies 1990–2005. Obesity (Silver Spring). 2008; 16: 275–84.

24. Stratton G, Canoy D, Boddy LM, Taylor SR, Hackett AF, Buchan IE. Cardiorespiratory fitness and body mass index of 9–11-year-old English children: a serial cross-sectional study from 1998 to 2004. Int J Obes (Lond). 2007; 31: 1172–8.

25. Taylor S, Hackett A, Stratton G, Lamb L. SportsLinx: improving the health and fitness of Liverpool’s youth. Educ Health. 2004; 22: 3–7.

26. Tomkinson G, Leger L, Olds T, Cazorla G. Secular trends in the performance of children and adolescents (1980–2000): an analysis of 55 studies of the 20m shuttle run test in 11 countries. Sports Med. 2003; 33: 285–300.

Cited By:

This article has been cited 1 time(s).

European Journal of Pediatrics
Cardiorespiratory fitness predicts clustered cardiometabolic risk in 10-11.9-year-olds
Houston, EL; Baker, JS; Buchan, DS; Stratton, G; Fairclough, SJ; Foweather, L; Gobbi, R; Graves, LEF; Hopkins, N; Boddy, LM
European Journal of Pediatrics, 172(7): 913-918.
Back to Top | Article Outline


©2012The American College of Sports Medicine


Article Tools



Search for Similar Articles
You may search for similar articles that contain these same keywords or you may modify the keyword list to augment your search.

Connect With Us