Analysis of global race records reveals how gender, age, nationality and distance combine to shape endurance running, from 5km speed to marathon dominance.
Study: Associations between nationality, gender, age and running performance in endurance runners: An empirical analysis of global long-distance race data from 1999 to 2024. Image credit: kovop / Shutterstock
In a recent study published in the journal Scientific Reports, a group of researchers investigated how nationality, gender, and age were associated with running performance in 5km, 10km, half-marathon, and marathon events around the world from 1999 to 2024.
background
Why do runners from Kenya and Ethiopia dominate marathon competitions when there are more participants from Japan and the United States? Millions of people enjoy endurance running, but their performance varies depending on factors such as age, gender, training culture, and geographic location.
A review of existing literature reveals age-related differences in peak athletic performance and shows that East African runners outperform other competitors in long-distance races in ways that may reflect unique physiological, environmental, and cultural advantages. However, there are limited studies that have compared these patterns simultaneously across multiple race distances. Therefore, further research is needed to better understand how demographic factors interact across endurance events.
About research
The researchers analyzed publicly available data from official records maintained by World Athletics (formerly the International Association of Athletics Federations) and the Road Race Statistics Association.
This dataset consists of eligible records for 5km, 10km, half marathon, and marathon races reported between 1999 and 2024, with 5km data supplemented from ARRS as World Athletics records were only available from 2018 to 2024.
After filtering and removing incomplete data, a final dataset containing 152,943 runner records consisting of runners (91,182 men and 61,761 women) from 203 countries, regions, or national teams was established.
Performance cutoff points were set to maintain consistency when comparing between races. Runners who did not complete the race, were disqualified from completing the race, did not start, or had incomplete records were excluded from the analysis. Neutral teams (refugee teams and neutral athlete teams) were also excluded to avoid statistical bias.
The researchers conducted mixed effects modeling and multiple regression analysis to examine the relationship between athletes’ nationality, age, gender, and running performance. A three-way analysis of variance (ANOVA) was used to examine the interaction of nationality and gender on running performance while adjusting for year of competition.
Researchers used t-tests to compare men and women and Pearson correlation coefficients to assess the relationship between age and race outcomes. Statistical significance was set at a probability value of less than 0.05. Because this is an observational analysis of public racing records, the results identify associations rather than causation.
Percentage of nationalities of runners by race distance and gender. Countries with a percentage of less than 1% are included in “Other”.
Research results
Results showed consistent differences in endurance running performance by gender, nationality, age, and long-distance race experience. Men consistently outperformed women in all four racial categories.
The average running speed for all races was faster for men than for women. Only a distance of 5 km had an average total speed greater than 20 km/h. Male runners were significantly faster than female runners at all distance levels. However, this performance difference was particularly pronounced at half-marathon and marathon distances.
Participation patterns also varied by country and distance. In the 5km competition, runners from the United States and Japan formed the largest participating group, followed by Kenya, France, and Australia. France and Kenya had large numbers of participants in the 10km race, while Japan and Kenya had notable participation in the half marathon and marathon.
Kenyan and Ethiopian runners consistently rank highly in all endurance events, confirming East Africa’s continued influence in distance running.
Bahraini athletes also had exceptional average race times in some events, but findings based on nationality should be interpreted with caution as some countries may include naturalized athletes.
As the race distance increased, the age pattern also changed. In the 5km race, the average age of the men and women was almost exactly the same at 25 years. However, women are now older than men in 10km, half marathon, and marathon races. Marathon runners were the oldest group in this study, with most between 25 and 35 years old. We also observed that younger athletes were more common in shorter races, while older runners were more common in longer endurance events.
Running performance varies greatly depending on nationality. Runners from East Africa, particularly Kenya, Ethiopia, Uganda, and Eritrea, have consistently recorded good results, especially in long-distance events, which may reflect adaptation to high altitudes, strong running cultures, genetics, and long-term training systems. In contrast, Japan’s success was tied to organized corporate running teams and the country’s ingrained “ekiden” culture.
The COVID-19 pandemic caused a significant drop in race participation across all distances from 2020 to 2022, resulting in races being canceled and public health restrictions introduced in various countries. However, participation has since recovered, especially in long-distance events.
The study found a strong association between age and race performance in half marathons and marathons (particularly for women). In shorter events such as the 5km race, there was a weak correlation between age and race performance. Interestingly, female runners in 5 km events showed a negative relationship between age and race time, suggesting that slightly older women in younger age groups often achieved faster times.
conclusion
This study shows that endurance running performance is shaped by a complex interaction of nationality, gender, age, and race distance. East African athletes, particularly those from Kenya and Ethiopia, remained particularly prominent in the long-distance events, with Japan and the United States contributing to large numbers of participants throughout the race. Men consistently posted faster race times, but women participating in longer events were generally older.
Research data shows that performance patterns vary by race distance and demographic background. These results are important to athletes, coaches, and sports organizations because they help develop training programs, identify talent, and understand the evolving global endurance running culture.
Reference magazines:
- Liao, G., Wang, J. (2026). Associations between nationality, gender, age and running performance in endurance runners: An empirical analysis of global long-distance race data from 1999 to 2024. Scientific Reports. Doi: https://doi.org/10.1038/s41598-026-46032-9
