I won’t post the whole study but I encourage you to read it through even though it is not strictly related to “the fastest race” but can give you an insight on what may be playing an important role in development of the next generation of fastest humans.
Worldwide variation in the performance of children and adolescents: An analysis of 109 studies of the 20-m shuttle run test in 37 countries
Abstract
This study is a meta-analysis of 109 reports of the performance of children and adolescents on the 20-m shuttle run test (20- mSRT). The studies were performed in 37 countries and included data on 418,026 children, tested between 1981 and 2003.
Results were expressed as running speed (km Á h71) at the final completed stage of the 20-mSRT. Raw data were combined with pseudodata using Monte Carlo simulation. The 20-mSRT performances were expressed as z-scores relative to all children of the same age and sex from all countries. An overall ‘‘performance index’’ was derived for each country as the average of the age- and sex-specific z-scores for all children from that country. Factorial analysis of variance was used to compare scores among countries and regions, and between boys and girls of the same age. There was wide and significant (P 5 0.0001) global variability in the performance of children. The best performing children were from the Northern European countries Estonia, Iceland, Lithuania, and Finland (0.6 – 0.9 standard deviations above the global average). The worst performing children were from Singapore, Brazil, USA, Italy, Portugal, and Greece (0.4 – 0.9 standard deviations below the global average). There is evidence that performance was negatively related to being overweight, as well as to a country’s average temperature.
…
Variation among countries
The performance indices (i.e. the average z-scores) for each of the 37 countries surveyed are shown in Table III. They ranged from þ 0.863 for Estonia to 70.867 for Singapore. There were significant differences between the performance indices of different countries (F ¼ 970, P 5 0.0001). The best performing nations were Northern European countries, notably Iceland and the Baltic states. The top seven nations were all from Northern and Central Europe. The worst performing nations were Southern European countries (Greece, Italy, Portugal, and Spain), Brazil, and some of the developed Pacific Rim nations (USA, Singapore, Hong Kong, and Australia). Of the bottom ten countries, seven came from Southern Europe and the Pacific Rim. When grouped into geographical regions (Northern, Central, Western, and Southern Europe; Africa; Pacific Rim; and Other Countries), there were significant differences in the unweighted mean performance index (P 5 0.0001). Northern Europe (mean performance index ¼ þ 0.60) outperformed all other regions, and all regions except the Pacific Rim (70.26) and Other Countries (70.19) outperformed Southern Europe (70.40).
In Europe, there was a clear north – south fitness gradient (Figure 3). The top five ranked countries were all from Northern Europe (Estonia, Iceland, Finland, Lithuania, and Ireland), while three Southern European countries (Greece, Portugal, and Italy) were among the six lowest ranked countries.
Discussion
Variation among countries
The variation in mean fitness between countries spanned over 1.7 standard deviations: 95% of Estonian children would perform better than the average Singaporean child. This span of performances equates, using a validated equation (Leger et al., 1988), to a difference of about 10 –´ 12 ml Á kg71 Á min71 (or about 25% of average V O2peak) between the most fit and least fit groups of children. To make these differences more concrete, it would mean that the average Estonian male adolescent would finish about 600 m ahead of the average young Singaporean in a 12-min run.
The superiority of Northern and Central European children, and the relatively poor performance of North American children, has been a recurrent theme in studies since the 1960s. In 1960, Knuttgen reported that 99% of Danish seventh- to twelfth grade girls and 96% of Danish boys performed better than American averages on the AAHPER 600-yard walk/run test. Shephard (1976) noted that Scandinavian children were much better on V O2peak tests than children from North America. In 1973, it was reported that 10-, 15-, and 17-year-old German boys and girls performed decidedly better on a PWC170 test (W Á kg71) than comparable Canadian children, and were of similar fitness to Czech children measured in other studies (Rutenfranz et al., 1973). Fredriksen et al. (1998) compared directly measured V O2peak values of European children, finding that children from Northern Europe were better than their peers from other parts of Europe. More recently, Koenig-McIntyre (1992) compared the one-mile run performances of 10-year-old Finnish, Norwegian, and Swedish children to those of US children. The passing percentages on criterion-referenced tests of the Northern European children (54 – 85%) easily exceeded those of US children (42 – 44%).
Methodological issues The issue of the representativeness of samples in cumulation studies of this sort has been discussed previously (Tomkinson et al., 2003a). In the studies used in this analysis, sampling procedures ranged from random national samples to local convenience samples, but no samples consisted entirely of disabled, diseased, or athletic groups. In meta-analyses drawing upon a large number of studies with a very large total sample size, the sheer mass of data points will tend to dampen irregularities arising from sampling inconsistencies. This is particularly true in cases where there have been a number of different studies from the same country. Results based on one or two small-sample studies are less reliable. In the present study, this was the case for Bolivia, Germany, Hong Kong, Ireland, Singapore, Suriname, and Turkey. By contrast, the results for Australia, Belgium, Estonia, France, Italy, Spain, and the UK were based on at least 5000 children and at least six separate studies each. They typically represented a wide geographical dispersion. For example, the Spanish studies covered children from Catalonia, the Canary Islands, the Basque Country, and Madrid; the Australian studies represented children from every state; and the British studies included children from each of the constituent nations. There were also very large samples (475,000) for Japan and Poland. We can therefore have considerable confidence in these results.
All children were tested between 1981 and 2003. However, there is now convincing evidence that in this period performance on tests of cardiovascular fitness has declined globally (Tomkinson et al., 2003a), at a rate of about 0.4% per year. This would mean that a direct comparison of scores would tend to reflect relatively poorly on those countries where the bulk of the measurements had been taken very recently (for example, Brazil, where the mean year of testing was 2001), and artefactually enhance the performance indices of those countries where the bulk of the testing had taken place earlier (for example, Canada, with a mean year of testing of 1985). When the mean performance index for each country was regressed against the mean year of testing, the index declined by an average of 0.032 units with every year of testing. Correcting for this effect by increasing or decreasing the performance index appropriately yielded a time-corrected perfor- mance index. The ranking of countries was not greatly changed, and there was a strong correlation between the uncorrected and corrected rankings (r ¼ 0.92, P 5 0.0001). The main beneficiaries were Australia (28th to 19th place), Djibouti (20th to 12th), Japan (12th to 6th), and Poland (27th to 17th), whereas Benin (14th to 21st), Canada (10th to 18th), Mauritius (15th to 23rd), and the Netherlands (25th to 33rd) all fell in the rankings.
Motivation is likely to be an important factor in test performance, both at the individual level and at a social or cultural level. The degree of ‘‘humiliation tolerance’’, for example, will affect a child’s drive to do their best. However, none of the studies in this review quantified motivation.
Possible socio-economic correlates
There are a number of broad socio-economic factors that could, in principle, impact on children’s fitness. These include the affluence of the country, its distribution of wealth, the ‘‘critical mass’’ of young people (i.e. the percentage of children and adolescents in a society), and the importance of sport in the national psyche. In addition, one would anticipate associations between aerobic test performance and both the children’s physical activity and the incidence of paediatric overweight.
Physical activity and overweight
Several recent studies have compared physical activity among children from different countries using objective measurements. Vincent, Pangrazi, Raustorp, Tomson and Cuddihy (2003) found that Swedish children took more daily steps than Australian children, who in turn were more active than American children. Riddoch et al. (2004) found the following order in the activeness of children from different countries using accelerometry: Norway, Estonia, Portugal, Denmark. Livingstone (2001) found a north – south gradient in the obesity levels of European children, with children from the south (Italy, Spain and Greece – but also Hungary) being fatter than children from Northern Europe. More recently, Lobstein and Frelut (2003) confirmed this trend, with high levels of child overweight in Spain, Italy, and Greece compared with Denmark, Germany, the Czech Republic, and Slovakia. If we consider that overweight and lack of physical activity are contributing factors to low aerobic fitness, then all of these findings are broadly consistent with the fitness differentials reported here.
Affluence
There are plausible mechanistic arguments that a society’s affluence might affect the fitness of its children. Wealthy countries can provide children with the time and equipment necessary for recreation, including school-based sport lessons.
Distribution of wealth
These children also have the social capital and education to understand the importance and value of vigorous physical activity, and health care systems to protect them from debilitating childhood diseases. On the other hand, children from affluent societies may also have the time and resources for greater sedentary activity (watching television, playing video games). They are less likely to use active transport and less likely to perform vigorous household or commercial work.
In this study, there was no clear relationship between the performance index and national wealth. Per capita GDP, expressed in $PPP or parity purchasing power, was available for 36 of the 37 countries in this meta-analysis. There was no relationship between GDP and the performance index (r ¼ 0.02, P ¼ 0.89). High fitness levels were found in children from the relatively poor transitional economies of Eastern Europe (e.g. Estonia, with a per capita GDP of $PPP 11,000). Conversely, very wealthy countries could show either very high fitness (Iceland, $PPP 30,200) or very low fitness (USA, $PPP 36,300). Relative affluence, at least in so far as it is quantified by GDP, was unrelated to children’s fitness.
A case can also be made a priori that countries where the distribution of wealth is less equal will show lower fitness. Relatively disadvantaged families in these countries (typically, the under-employed) may not have adequate resources or the social capital to encourage active recreation in their children, while the relatively advantaged (the over-employed) may lack time or succumb to the effortless lifestyle of abundance. Furthermore, there is a great deal of evidence that income inequality is correlated with lower life expectancy, increased risk of cardiovascular and other diseases, impairment of children’s growth, and social disintegration (Wilkinson, 2000).
In strongly hierarchical societies, subordinate members experience high anxiety, as an evolutionary legacy of the need to be prepared for fight or flight when threatened by superordinates. This is manifested in higher cortisol concentrations and subsequent impairment of growth and immune function (Wilkinson, 2000). The most common economic measure of the distribution of income within a country is the Gini coefficient. It ranges from 0 (‘‘perfect equality’’) to 1 (‘‘maximum inequality’’). Gini coefficients for the distribution of household income were available for 30 of the 37 countries in this meta-analysis.
There was again no significant relationship (r ¼70.26, P ¼ 0.16) between the Gini coefficient relationship between the percentage of the population aged 0 – 14 years and the performance index (r ¼ 0.10, P ¼ 0.56).
There are thus few obvious socio-economic correlates of fitness at the national level. It is possible that cultural factors, which are harder to quantify, may be important. The Northern European (Baltic and Scandinavian) and, to a lesser extent, the Central European states stood out as having excellent levels of children’s fitness. This may partly be due to their long tradition of institutionalized and organized participation in physical activity, starting with the gymnastics movements in Central and Northern Europe in the
early 1800s: Jahn’s Turnvereine in Germany, Sokol gymnastics in Central Europe, and Ling gymnastics in Sweden. These led to massive youth movements throughout the twentieth century.
Climate
There was a significant negative relationship (r ¼ 70.44, P ¼ 0.007) between the performance index and the average annual temperature of the capital city of the country, so that children from colder countries performed better. This could be related to the effect of temperature on test performance: children would be expected to perform worse under very hot and perhaps humid conditions. However, it might also be expected that children would perform poorly under very cold conditions. Alternatively, the differences in performance could reflect different cultural attitudes towards physical activity in hot and cold countries, with children from very hot countries tending to avoid exertion in the heat. The most likely explanation, however, is that this relationship is an artefact created by the outstanding performance of children from Northern Europe. When these countries are excluded from the analysis, there is no longer a significant relationship between temperature and performance (r ¼ 0.18, P ¼ 0.34). Differences between the sexes Boys easily outperformed girls, the differences being significant (P 0.0001) in every age group. Because differences between the sexes are consistent across a wide range of countries with different social, political, and economic systems, they are probably biological rather than social in origin. Previous studies that failed to find differences between pre-pubertal boys and girls probably suffered from small sample sizes.
Conclusion
The data provided in Table II represent world standard performance on the 20-mSRT for the mean and the performance index. Children in countries with low Gini coefficients tended to have high levels of fitness, while those in countries with relatively high Gini coefficients could have either high or low fitness. The country with the fittest children (Estonia) had a relatively high Gini coefficient of 0.37. As with absolute differences in wealth among countries, variability in wealth within countries may work either to enhance or reduce fitness. Cultural commitment to sport: Olympic success Children’s fitness is likely to be affected not only by economic but also by cultural factors, such as the importance of sport and physical activity in the nation’s psyche. The national cultural commitment to sport is difficult to quantify, but a reasonable proxy might be a country’s performance in the Olympic Games. We used an ‘‘Olympic index’’ based on the number of medals won in post-war summer and winter Olympics, corrected for participation rates and for the number of medals available at each Olympics. This number was then divided by the country’s population.
There was a moderate but significant relationship between the performance index and the Olympic index (r ¼ 0.32, P ¼ 0.028 one-tailed). In particular, Finland, Denmark, Estonia, and Iceland had both a high performance index and a high Olympic index, whereas Singapore, Hong Kong, and Brazil scored poorly on both indices.
This correlation may be interpreted in several ways. In could mean that there is a ‘‘trickle-down’’ effect, whereby countries that invest a large amount of money in elite athletic performance reap benefits at the grass roots level. Equally, it may be interpreted as evidence for a ‘‘trickle-up’’ effect, whereby countries with high grass roots fitness produce athletes who can participate in elite competition. It is also possible that a third factor – for example, the extent to which a government will invest in sports infrastructure – affects each index independently. It should be noted, however, that Olympic medals may not always be a good indicator of sporting success or national interest in sport. Very small countries may be disadvantaged, and the inclusion of Winter Olympic Games tends to exclude hot countries.
Critical mass of children There is also a rationale for suggesting that aerobic test performance could be affected by the number of children in a society. A large percentage of children may constitute a ‘‘critical mass’’, which makes it politically and logistically easier to encourage youth sport. In our data set, however, there was no year of testing (1996). They may therefore serve as a yardstick to compare performances among countries and across ages and sexes.
This comprehensive review of worldwide variation in children’s aerobic test performance showed wide variation in fitness levels. Baltic and Scandinavian countries fared best; Mediterranean and developed Pacific Rim countries had the least fit children. The Northern European countries provide a model of how affluent nations can sustain high apparent fitness in their children, and further research should explore the reasons for this. Neither affluence nor the distribution of wealth appears to be important in determining children’s performance. Recent studies (Olds & Dollman, 2004; Tomkinson, Olds, & Gulbin, 2003b) have suggested that about half the decline in aerobic test performance in Australia over the last decade is associated with increases in fatness, although the causal arrow may point in either direction. The agreement in between-country variability in the incidence of overweight and test performance supports these findings. Cultural and even climatic factors may also play a role.
The data on which these conclusions are based are lacunary. There are at present very few data on developing countries, especially in South America and Asia, and data on certain age-groups in developed countries are lacking. For some countries, the sample sizes were very small, while the other studies did not use random samples. A coordinated global approach is needed to systematically monitor changes in children’s fitness. Mechanisms should be put in place to facilitate data-pooling, such as an Internet-based data repository. There is also a critical need for standardization in the way the 20-mSRT is administered and the results analysed and reported.