In order to increase the reliability of our data analysis, each set of data was tested for normal distribution as well as homogeneity of variances in advance. Normal distribution was tested using a D’Agostino and Pearson omnibus normality test, homogeneity of variances using a Levene’s test in case of two groups and with a Bartlett’s test in case of more than two groups. In order to find significant changes in a variable across years, single linear regression analysis was used. To find significant differences between two groups, a Student’s t-test was used. Statistical analyses were performed using IBM SPSS Statistics (Version 19, IBM SPSS, Chicago, IL, USA) and GraphPad Prism (Version 5, GraphPad Software, La Jolla, CA, USA). Significance was accepted at P < 0.05 (two-sided for t-tests). Data in the text are given as mean ± standard deviation (SD).

In marathons, 0.2% of all finishers were African men and 0.1% African women during the 2000–2010 period (Table 2). In African finishers, 23.3% (overall finishers: 0.1%) were Ethiopian men, 38.4% (0.1%) Kenyan men, 16.9% (0.1%) Ethiopian women and 3.4% (0.01%) were Kenyan women (Figure 1A). In half-marathons, 0.1% of all finishers were African men and <0.1% were African women (Table 1). In African finishers, 27.1% (overall finishers: <0.1%) were Ethiopian men, 20.8% (<0.1%) were Kenyan men, 11.9% (<0.1%) were Ethiopian women and 11.9% (<0.1%) were Kenyan women (Figure 1B).

In half-marathons, the number of Non-African male finishers (14,753 ± 4,492) was higher than the number of Non-African female finishers (5,839 ± 2,122 finishers) (P < 0.05). For African men, the annual number of half-marathon finishers increased (P < 0.01) (Figure 2A). The number of Non-African male and female finishers increased over the last decade in half-marathons by 131.1% for men and by 199.6% for women (P < 0.01) (Figure 2B). In marathons, the number of Non-African male finishers (6,473 ± 2,966) was higher than the number of Non-African female finishers (1,362 ± 670.7) (P < 0.05). Only African men increased their annual number of marathon finishers (P = 0.01) (Figure 2C). Non-African men (P < 0.01) and women (P = 0.02) increased their annual number of marathon finishers by 160.9% for men and by 134.4% for women over the last decade (Figure 2D). In half-marathons, there was no difference between the number of African male (15.4 ± 7.7) and African female (6.1 ± 2.5) finishers (P > 0.05). Also for marathons, there was no difference between the number of African male (16.6 ± 9.1) and African female (4.9 ± 3.3) finishers (P > 0.05).

In half-marathons, the annual number of male Ethiopian finishers (5.8 ± 3) was higher (P < 0.05) than the annual number of female Ethiopian (2.5 ± 1.9) finishers (Figure 3A). In Kenyan athletes, there was no difference (P > 0.05) between the annual number of male (4.5 ± 2.2) and female (2.5 ± 1.1) finishers (Figure 3B). In Ethiopian as well as in Kenyan finishers, neither men nor women showed a significant change in the number of annual finishers across years (P > 0.05). In marathons, the mean number of male Ethiopians (5.5 ±1.8) was not different (P > 0.05) from the number of female Ethiopians (4 ± 1.8) (Figure 3C). In Kenyans, the number of male finishers (9.1 ± 5) was higher (P < 0.05) than the number of female finishers (0.8 ± 0.6) (Figure 3D). In both Ethiopians and Kenyans, neither men nor women showed a significant change in the number of finishers across the years (P > 0.05).

The change in participation for East African runners in specific races held in Switzerland is presented in Figure 4 for half-marathons and in Figure 5 for marathons. For men, a significant decrease in both Ethiopian and Kenyan marathoners in ‘Lausanne Marathon’ was found (Figure 5B) whereas for women, only the number of Ethiopian athletes decreased (Figure 5E). In contrast, the number of both male Ethiopian and male Kenyan runners increased in the same period in ‘Zurich Marathon’ (Figure 5C).

In half-marathons, the overall top three race times of African women (69.3 ± 1.2 min) were not faster than those of Non-African women (71 ± 1.7 min). In contrast, the top ten race times in African women (71 ± 1.4 min) were faster than those of Non-African women (72.7 ± 1.6 min). African men were faster than Non-African men regarding the top three (P = 0.04) and the top ten (P < 0.01) race times (Figure 6A).

The average race times for African men were 130 ± 0 min for the top three and 130.2 ± 0.4 min for the top ten. Compared to top three (129.0 ± 1.0 min) and top ten (131.0 ± 1.9 min) race times in Non-African men there was no difference in performance (P > 0.05). The same was found for women (P > 0.05). The African top three women finished in 151.6 ± 2.5 min compared to Non-African top three women 150.6 ± 1.2 min. The mean race time for top ten African women was 154.6 ± 2.7 min compared to 152.8 ± 1.7 min for top ten Non-African women (Figure 6B). Comparing the race times of all African and Non-African finishers in half-marathons and marathons, African women (126 ± 52 min) were faster than Non-African women (145 ± 54 min) (P < 0.01). The average race time of all female finishers was 145 ± 54 min. Comparing African men (113 ± 49 min) and Non-African men (143 ± 62 min) there was also no difference in performance (P < 0.01). The mean race time of all male finishers was 143 ± 62 min. Comparing all finishers, men and women, African athletes (116 ± 50 min) were faster than Non-African athletes (144 ± 60 min) (P < 0.01). The mean finishing time of all finishers was 144 ± 60 min (Table 3).

Results are presented as mean ± SD.

In half-marathons, the top three African women (32.3 ± 1.5 years) and men (27.3 ± 5 years) were not younger than the top three Non-African women (33 ± 2 years; P = 0.67) and men (26.3 ± 4.2 years; P = 0.80). Also in men’s top ten there was no difference in age between African (26.1 ± 3.9 years) and Non-African (25.7 ± 4 years) athletes (P = 0.82). Comparing the top ten finishers for African (27 ± 6.4 years) and Non-African (32 ± 3.5 years) women, African women were younger (P = 0.04; Figure 7A) than Non-African women. In marathons, comparing the top three and top ten finishers, there was no difference in age between African and Non-African men, and between African and Non-African women (Figure 7B).

In contrast to our hypothesis, the number of Kenyan and Ethiopian finishers remained stable for both men and women in both half-marathons and marathons. The increase of African men in half-marathon and marathon was due to other African countries, which were not analysed in details. In contrast, the numbers of male and female finishers in Non-Africans increased in both marathons and half-marathons. This finding correlates with results presented by Jokl et al. 6 reporting an increased participation in general over the last 30 years in a large city marathon such as the ‘New York City Marathon’.

Regarding the studied running events, African mainly competed at larger marathon and half-marathon events (e.g. Greifenseelauf, Lausanne Marathon, Hallwilerseelauf, Genève Marathon for UNICEF and Zürich Marathon). In all other races, there were five or fewer Africans competing during this period. We assume for smaller events lower budgets and these events were less prestigious. Interestingly, the ‘Lucerne Marathon’, the ‘Maratona Ticino’ and the ‘Winterthur Marathon’ were all among the seven largest running events for half-marathons and marathons. However, fewer than four Africans were participating annually per race between 2000 and 2010. In personal conversation with the race directors from the ‘Lucerne Marathon’ and the ‘Winterthur Marathon’ we learned that prize money was less than 200 Swiss francs (i.e. about 220 US dollars) and no promotional fees were offered either for African or for Non-African athletes. They further reported a decline of the prize money over years leading to less attraction for elite African athletes. Therefore, they races become more attractive for regional marathoners and half-marathoners. So in common beliefs the opportunity to win the race increased in absence of the elite East African athletes. Other personal communications with the race directors of the ‘Greifenseelauf’, the ‘Hallwilerseelauf’, the ‘Zurich Marathon’ and the ‘Lausanne Marathon’ (i.e. races with a high number of African participants) revealed that special prize money and promotional fee (especially for board and lodge) for elite African athletes were offered. They specially invited elite athletes in general to make their competition more interesting. Perhaps from the ‘Maratona Ticino’ and the other races no data for prize money were available and beside the ‘Winterthur Marathon’ only from the ‘Lausanne Marathon’ an amount of the prize money was given 29 .

According to poverty and unemployment in East African countries, Onywera 27 and Onywera et al. 26 reported that the motivation to become a competitive athlete was mainly based on economic reasons for Kenyan elite athletes especially for 39% of all national athletes and 33% for international athletes. The international success of the East African running pioneers such as Mike Boit, Kipchoge Keino and others implicated a ‘westernation’ of the culture, especially in Kenya 27 . Therefore, running became more professional and more popular in Kenya. Affected by the great international success of the East African running pioneers, young athletes know about the possibility of becoming a successful international athlete. Motivation of competitive running therefore increased and young East-African athletes became support by managers of Western countries. So the East African athletes came closer to the Western civilisation and according to this theory of closeness the chance to make money with competitive running increased. Based on this sociocultural background and according to our different findings in participation of East African athletes in Swiss long-distance running events and personal conversation with the race directors the question about a correlation between prize money and participation went up. In addition, the respondent race directors from running events with a large number of Africans reported a stable budget for invited elite athletes. Therefore, the possibility to make money with competitive running in Switzerland is limited and the annual number of Kenyan and Ethiopian participants remains stable. For further studies it would be interesting to analyse potential correlations between prize money and participation of African elite athletes in larger countries (e.g. USA).

During the studied period there were more African men than African women competing. The same can be reported for Non-African athletes. Participation differences between men and women are common findings in sports events. In 1988 at the Olympic Games in Seoul (KOR) only 26.1% of all participants were women 30 . Compared to the XXIX Olympic Games in Beijing (CHN) in 2008 women increased their percentage of participants up to 42.4% 30 . So, especially in earlier years there were restrictions on female athletes. In Kenya, women were banned from international competitions until the 1970s 27 . Regarding the percentage of female participants at the Olympic Games in Seoul, 5.5% of the Kenyan team were women. They increased their participation also up to 39.1% in the Olympic Games in Beijing (2008) 31 . Therefore, the female running culture has not such a long time history as that of men. The Kenyan men were well established in long-distance running since years. But today, a female running culture is growing in Kenya and it is also a great honour for female athletes to represent their country in international competitions 27 . Participation of Kenyan woman increased with a delay of several years compared to Kenyan men. Also in Ethiopia, there was a similar development. Comparing Olympic Games in 1980 (Ethiopia took not part in the Olympic Games in 1988 and 1984) and 2008, Ethiopian women increased their participation from 4.9% to 48.1% 31 . These reports may explain the difference between female and male participation trends of Kenyan athletes in marathons and Ethiopian athletes in half-marathons.

More interesting seems the fact of the difference in participation in male Ethiopian and Kenyan athletes in both half-marathons and marathons. Compared to their Ethiopian counterparts, more Kenyan men were participating in marathons and fewer in half-marathons. A possible explanation for this might be differences in anthropometry between Kenyan and Ethiopian runners. Ethiopians are more mesomorphic in somatotype, which includes more muscle mass especially expressed in a high thigh circumference 32 . Kenyans in contrast are more ectomorphic with slender legs 32 . Zillmann et al. 33 showed differences in anthropometric characteristics between recreational marathoners and recreational half-marathoners competing in Switzerland. They reported for half-marathoners a thicker thigh circumference compared to marathoners 33 . To the best of our knowledge anthropometric characteristics of Ethiopian runners have not been analysed yet. Based on reported differences in anthropometry between Ethiopians and Kenyans 32 we can only speculate that Ethiopian runners were more predestined for running half-marathons than marathons. Therefore, we had a higher number of Ethiopian participants in half-marathons than in marathons. Also in the IAAF top list from 2011 in marathons and half-marathons there was a significant difference between the number of Ethiopian and Kenyan runners for the top 20 race times 24 . In the top 20 marathoners there were 20 Kenyans 24 . In half-marathons there were 13 Kenyans and 9 Ethiopians ranked 24 . Future studies need to analyse anthropometric differences between Ethiopian and Kenyan long-distance runners.

In the two largest marathons held in Switzerland, the ‘Zurich Marathon’ and the ‘Lausanne Marathon’, a reverse trend was found regarding the participation of male East African marathoners. While the participation decreased in the ‘Lausanne Marathon’, it increased in the ‘Zurich Marathon’. Most probably the prize money was higher in ‘Zurich Marathon’ compared to ‘Lausanne Marathon’ during this period.

There was no difference in race times between elite African finishers and elite Non-African finishers in marathons. This finding does not support the hypothesis of a better running performance of elite African runners compared to Non-African runners and was not in line with the anecdotal well-known African dominance in the literature regarding long-distance running 8 9 20 21 22 23 . Comparing maximum oxygen uptake (VO₂max) between elite Kenyan and elite Non-African runners, disparate findings were reported 8 21 22 23 . Therefore, VO₂max as an important predictor variable of performance was critically discussed and in general there is no difference between Kenyan and Non-African runners 8 . Also, no differences were found in proportion of muscle fibers and muscle oxidative enzymes 8 21 . Running economy, defined as the steady-state of submaximal oxygen uptake at a given running velocity, was further critically discussed as an important factor of performance 8 . Saltin et al. 34 reported a better running economy for Kenyans while Bosch et al. 35 reported no difference in the oxygen uptake at given performance levels. In summary, no significant physiological differences between elite African runners and elite Non-African runners could have been found yet and this goes in line with our unique results in marathons.

Nevertheless, regarding the IAAF top list of marathons and half-marathons and corresponding to common beliefs, East African runners had better running performances than Non-African runners 24 . Why do marathon events in Switzerland differ to marathons held in other countries? A plausible explanation might be the thesis that not the top of the top East African athletes were competing in Swiss long-distance running events because they were less prestigious and there was less prize money to win. For example in the Lausanne marathon there were about 2,150 U.S. Dollar (i.e. 2,000 Swiss francs) for winning the marathon 29 . Comparing to the ‘New York City Marathon’ as the most prestigious marathon event worldwide the winner got about 100’000 U.S. Dollars 36 . To confirm this thesis about ‘second class’ elite East African runners in Swiss marathons further studies should analyze East African performance trends in larger countries. Especially performance findings of East African athletes from the World Marathon Majors (i.e. a collective of the worlds’ largest marathon events) should be compared to our data of smaller country with smaller races.

Interestingly, elite African athletes were faster than elite Non-African athletes in half-marathons. This finding reflects the general opinion, anecdotal findings and international top lists of the IAAF 24 and goes not in line with our results found for marathons. Based on scientific literature we cannot explain this fact with physiological differences between Africans and Non-Africans. An explanation might be different sociocultural backgrounds and differences in running experience. A recent study compared a variety of predictor variables for performance between half-marathons and marathons 33 . The authors reported that marathoners had a larger number of completed long-distance competitions in their running history than half-marathoners. So in Switzerland, marathoners start their running career with half-marathons to gain experience 33 . Relatively to the world records 2011 in half-marathons and marathons 37 Non-African half-marathoners differ to Non-African marathoners in race times. The top Non-African race time in men’s half-marathons was 7.3% away of the world record. In marathons, however, there was only a gap of 3.4%. In Swiss long-distance events Non-African marathoners were better than half-marathoners in the worldwide ranking. More newcomers and recreational runners were competing in Swiss half-marathons might be a possible explanation for this. Non-African recreational runners were motivated for running by health benefits 2 4 . It is possible to run a favorable performance beside a workaday life because not as many training kilometers and training units are needed for half-marathons compared to marathons 33 which is essential in Switzerland. Therefore, with a lot of newcomers in Non-Africans the possible top race time have not been reached yet. In contrary, top African runners were professional runners with a large running history since childhood and they were also driven by economical reasons to compete in Swiss half-marathons 25 26 27 . So, they run better performances compared to Non-African athletes in half-marathons.

In contrast to our second hypothesis, elite African finishers were not younger than elite Non-African finishers neither in half-marathons nor in marathons. First, the majority of African runners in Switzerland between 2000 and 2010 were elite East African runners (76.7%) from Ethiopia (39.5%) and Kenya (37.21%). Performance is limited by age-related physiological changes. The maximal oxygen uptake (VO₂max), as one of the most important factors, declines with age 3 10 14 16 18 . In addition to VO₂max, the maximum heart rate (HR_max), the lactate threshold and the running economy change with age. HR_max significantly decreases with age 10 15 16 , the lactate threshold, as the exercise intensity (%VO₂max), increases significantly with age 10 17 19 . Top performances in long-distance running events will be achieved at younger ages (<40 years) 1 3 6 for both Africans and Non-Africans. Therefore, we can explain the same age of top performances comparing African and Non-African athletes in marathons and half-marathons. Especially in half-marathons, there was no age difference but a significant difference in performance between Africans and Non-Africans. This could be explained by an age-dependent loss in performance as well. The correlation between age and performance is relatively and not absolutely. So, within each group peak performance will be achieved at the same age regardless of a difference in performance between these two groups. The difference in performance would be explained by the fact that Non-African athletes have not reached their possible peak performance yet and not by age-differences.

A strength of the present study is the high number of runners analyzed in marathons and half-marathons from one country for an entire decade. To date, a demographical analysis about participation in marathons and half-marathons in such a way is unique. However, this study is limited in that determining factors for performance such as training, anthropometry, physiology, nutrition and motivation were not included. A limitation is also that data about prize money only was given in personal conversations with the race directors. Future studies should investigate participation, age and performance trends in marathoners and half-marathoners in larger countries such as the United States of America in order to strengthen demographic research. In addition, psychological and sociocultural backgrounds of the motivation of East African runners in relation to participation trends in half marathons and marathons should be analyzed in-depth.