Introduction

Not much research has been carried out on the effect of aging on running performance, defined as running time or running speed, on short distance road races. Most research is based on longer distance road races, like marathons [1–4]. However, running marathons is rather exceptional among elderly persons, while many potentially could run short distances. To be able to convincingly stimulate and instruct people, reliable evidence on potential performances and safety of running on age is highly needed.

So far, literature suggests that there is a significant age-related linear increase in running time starting at age 30–35 [3, 5–10], which accelerates beyond the age of 60–70 [1–4, 7, 11, 12]. Marathon running times of men are 8–18% faster compared to women [13, 14]. The decline in running performance with age seems greater for women than men due to biological influences [7, 9, 12, 13, 15].

In earlier work we showed that even (training for) long distance marches may decrease the rate of decline of VO2max with advancing age [16, 17]. Moreover, regular exercise is associated with a significant better physical function, improved cognition, a significant decrease in mortality and a better quality of life [2, 10, 12, 18–22]. However, with growing older, participation in exercise decreases. Preconceived negative ideas on exercise opportunities and performances at older ages are associated with very low levels of physical activity, which jeopardizes healthy aging [23]. However, over the last years, running competitions seem to be able to attract more elderly people and running performances have significantly improved, especially for elderly women [2, 7, 14].

To be able to improve evidence-based coaching in active aging, we conducted a study on the effects of aging, gender, and training on the running performances and drop-out rates in the most popular Dutch road run. This Nijmegen Seven Hills Run (see www.Zevenheuvelenloop.nl) is a 15-km run in a woody area that attracts an increasing number of elderly persons, leading to almost 25,000 participants in 2008, of whom 1,007 were older than 60 years.

Methods

Research population

The research population comprises all participants of the Seven Hills Run from 1995 to 2007, this means 194,560 persons. All participants who started in the competition were included. People registered for the run, but who did not participate, were excluded. Participants were grouped in 5-year age categories. According to the Dutch Athletic union, a master athlete is aged over 35 years (up till 2007 this age limit was 40 years). To have uniform categories over the period of 12 years, we made one youngest category for participants younger than 40 years (divided into men and women) and two oldest categories: women ≥60 years and men ≥65 years.

Data

Running records from the Nijmegen Seven Hills Run from 1995 to 2007 were used. The 15.0-km Nijmegen Seven Hills Run covers seven short slopes in a hilly countryside. The first part is mainly uphill, the second part varying, and the last part mainly downhill. The organization of the Seven Hills Run provided all data and coded names of participants. The records included date of birth, gender, athletic club membership, and running times after 5, 10, and 15 km. Running time was electronically measured by the Champion Chip® (ChampionChip Nederland, Nijmegen, The Netherlands), a chip attached to the shoe, which registers starting and finishing times. Training status was defined by membership of an athletic club. Competitive members were coded as trained, non-competitive, recreant members as less trained. Time was registered after 5, 10, and 15 km.

Statistical analysis

Statistical analyses were performed using SPSS (version 14.0). Extensive data cleaning was performed. Subjects with missing values in running time after 5 and/or 10 or 15 km were excluded for analysis. The relationship between running performance and age was examined by means of multiple regression analysis, adjusted for sex and training status (ANCOVA). χ 2-tests were performed to assess differences in participation numbers with passage of time in total, by age group, and between men and women. t tests were performed to assess a period effect in running performance between age groups and between men and women. Differences at the P < 0.05 level were considered statistically significant.

Results

Characteristics of the research population from 1995 to 2007

The youngest participant of the Nijmegen Seven Hills Run was 10 years, the oldest 80 years; 2.8% of the participants was ≥60 years, this means 5,507 people, of which 77 people were ≥75 years. The mean age of the participants increased significantly from 39.0 years (SD 9.2) in 1995 to 41.3 (SD 10.8) in 2007 (P < 0.001). Significantly more men (77.7%; P < 0.001) than women and more untrained (89.7%, P < 0.001) than trained people participated

Determinants of running time

With aging mean running time over 15 km increased. The aging effect on 15-km running time could be estimated as \( {\hbox{Time}}\left( {\min } \right) = 70.9 + 0.15 * {\hbox{age}}\left( {\hbox{years}} \right) \). On average, aging 1 year increased running time 0.2%, or 9 s (P < 0.001). Mean running times were influenced by sex, training, and the year in which the run was ran. Mean running times were 13% faster for men (P < 0.001; Table 1). Trained participants on average had a 15.7% better performance compared to less trained participants (P < 0.001; Figs. 1 and 2). The estimated mean running time corrected by age, sex, and training for 15 km is: \( 67.75 + 9,20 * {\hbox{sex}} - 12,34 * {\hbox{training}}\,{\hbox{status}} + 0,20 * {\hbox{age}} \), 29.3% (in which sex is 0 for men and 1 for women, training status is 0 for trained and 1 for untrained persons). Total variance in mean running time after 15 km was explained mostly by gender (14.8%) and less by age and training status (3.3% and 11.8%, respectively, all P < 0.001).

Table 1 Mean running time after 15 km by age and gender over the total of 12 runs
Full size table
Fig. 1
figure 1

Mean running time over 15 km by age and training status for men

Full size image
Fig. 2
figure 2

Mean running time over 15 km by age and training status for women

Full size image

Participation rate over the last 12 years

Participation in the Nijmegen Seven Hills Run significantly increased over the period 1995–2007. In 1995, there were 6,756 participants and in 2007 this number was 23,000 (P < 0.001). The largest relative increases in participation numbers for both sexes were in the older age groups (≥55 years), for example from 84 men aged 60 years and older in 1995 to 796 men in 2007. The number of participants has significantly grown from 1995 to 2007 with 1,101 to 6,125 for women, compared to an increase of 5,655 to 17,233 men (P < 0.01). The drop-out number was small. Between 1995 and 2007, 413 participants dropped-out (0.2% of total number of participants). No difference in relative number of drop-out was seen between young (<60 years) and old, nor between trained and untrained participants. Overall the number of drop-outs of participants over 60 years of age ranged from 0–0.7% per year of the run. Relatively more women (0.3%) than men (0.2%) dropped-out (P < .05). Cause of drop-out is not known.

Running time over the last 12 years

From 1995 to 2007, the mean running time after 15 km increased significantly with 9.9%. In every age group, mean running time significantly increased, except for women ≥60 years. On average, running time deterioration was 12–28 s from year to year for men and 17–34 s a year for women, ranging over age categories. Overall, most of the trained participants had constant mean running times between 2 years. Untrained participants had more variable performance results during the years.

Running performance during the run

Men had significant faster mean running velocities than women in all parts of the run, but this difference decreased with aging. Men <40 years had mean running speeds 1.15 times faster than those of age-matched women (P< 0.01). Men ≥60 years had mean running speeds only 1.10 times faster than those age-matched women (P < 0.01). When the two groups of participants <60 years and ≥60 years were compared, the percentage decline in running velocity with aging was significantly larger for men than women (3.4% versus 9.3%, P < 0.001) and significantly larger for trained than untrained participants (P < 0.001). However, trained participants of all ages had significant higher mean running speeds compared to age-matched untrained participants (P < 0.001).

Discussion

This study presents data showing that 15-km running performance is influenced by age, gender, training state, and the year of the run, of which gender and training had the largest influence. Age effects on running speeds were rather small and the drop-out number among older persons is very low. During the last 12 years, the number of elderly participants increased most quickly, and their good running performances prove that elderly persons can very well train for, run, and finish in a 15-km run. As a result of this increasing participation of elderly people mean running time increased, especially for untrained participants.

With aging, mean running time showed only limited decline of 0.2% per year. The decline in performance per year is smaller than described for marathon performances (10.5–14.8% per decade) [2]. Running performance is influenced by several physiological variables, which can explain the decline in running performance with age. Most important are maximum aerobic capacity (VO2max), muscle function, strength, and lactate threshold [8, 13, 15, 17]. Aging is associated with a decline in muscle mass, which becomes progressive beyond the age of 50 [2, 4, 11, 13, 24, 25]. The lactate threshold is reached at a faster rate, through which elderly people have a lower tolerance for strenuous physical activities [5, 7, 13, 26–28]. With growing old, cardiovascular parameters decline also, which causes decline in VO2max [2, 7, 8, 13–15, 20, 29, 30]. Training can increase VO2max and the lactate threshold, even for elderly people [5, 7, 10, 13, 16–18, 20, 21, 25, 29, 30]. Possibly, elderly participants may also show less motivation to reach their fastest running performance [5].

Velocity in men on average was 13.0% faster than in women, comparable to studies on marathons [3, 13, 14]. Difference in running performance between sexes is probably caused by a lower VO2max [7, 14, 16, 28] and less muscle mass for women compared to men [12, 14]. Running performance differed 15.7% on average between training groups. Training effects on running speed were larger for men than for women, and aging had a larger negative effect on running speed for trained than untrained participants. This is consistent with the faster decline of VO2max with aging in trained compared to sedentary elderly people, but running performance remains better in elderly trained than elderly untrained participants as has been shown for VO2max [3, 12, 20, 29]. The increase in participation of elderly and women, was also described by Jokl et al. [2]. However, results on women and older people are still based on small groups. The participation number of the Seven Hills Run was restricted to a certain maximum for reasons of logistics. For the past years, the limit was increased with 1,000 individuals every year. During these years, the Seven Hills Run has changed from a small quite competitive run into the largest Dutch road run for a more varied (untrained) public.

Over the last decade, running performance decreased significantly for all participants with 9.9%. Other studies showed better or constant running performance with passage of time [2, 7, 14]. Their conclusions were that improvements were possible by better tracks and more competition. For the studied years, the track of the Seven Hills Run was the same, but participation numbers increased, most in the older age groups. Probably the change in competitiveness, caused by less professionally trained participants, in combination with the increasing number of elderly, is the best explanation for higher average finish time.

Participants ≥60 years had significant slower running times compared to younger participants, especially in the last part of the run, men even more than women. Other studies found a higher decline for women compared to men, but their results were possibly confounded by selection bias [7, 9, 13, 21]. In our study, effect modification by age is possible, because in the male group, compared to women, more participants ≥65 years were present.

Analysis of master running performance in athletic records has not frequently been carried out in the past. However, our data comprise a lot of information on performance, health, and aging. An advantage of using these records is that they are based on very large samples. Other studies examined maximum performance in master athletes and less trained older persons, which makes it very likely that these results are not representative for the average older person.

A limitation of this study is the lack of information concerning confounders, such as medical diagnoses, laboratory, and physiological parameters. With no upper limit for age, a considerable range of variability is allowed, within which the desired homogeneity of the exposure of the run may not be achieved. Because of this, exact age was used for most analyses. In addition, variable weather conditions could have influenced running performance, although these conditions are relatively stable.

It is unclear whether competitive membership of an athletic club is an accurate measurement instrument for training status, because other participants also prepared themselves for the run and no data were available for the duration and intensity of trainings within a club.

It is well evidenced that exercise has multiple positive effects on health and substantially contributes to healthy and successful aging. This study showed that age only has a minor effect on running times and that elderly people can very well run a 15-km road race. Running performance is better in members of athletes clubs, also in the elderly participants. Overall, our data are promising and can stimulate and inform professionals in guiding elderly in running. Our results underline that running holds great opportunities to improve general health, because it is a simple, cheap, safe, and healthy means of active aging. Moreover, running and other exercises are equally available to all aging societies, which clarifies why the World Health Organization emphasizes this exercise as an important opportunity to reach active aging [31].