The survey was sent to all those who had been registered with parkrun since 2004 (whether their participation had lapsed or not) and those who had never done a parkrun (around 43% of those registered), which may account for the relatively low response rate of 100 866 survey returns (around 4.4% of registrants and 7.7% of participants). The following responders were removed: 37 039 who consented to view the survey but did not answer any questions; 1786 who had registered with parkrun but had not yet participated; 1349 who did not consent; 681 who self-identified exclusively as volunteers; and 12 who provided invalid or malicious responses. This left 59 999 responses, of which approximately 75% were matched to parkrun data, resulting in 45 662 participants with matched mean 5 km times from the parkrun database.
Demographics
Figure 1 shows the demographics of participants ranked by average running time (see Supplementary Table S1 for details). In comparison with the full parkrun population, the sample had a similar proportion of females (51.7% for the sample versus 51.3% for the population), a similar ethnic and employment background, and was older (48.0 years for the sample versus 40.5 years for the population).13 The latter was primarily because the survey was restricted to those aged ≥16 years.
Figure 1. Characteristics of survey participants ranked by average running time: a) count; b) proportion male and female; c) age; d) Index of Multiple Deprivation quartile (Q1 is most deprived); e) activity level at registration in bouts of 30 min or more in previous 4 weeks. Comparison with walkers at P <0.001 with effect sizes: *small, **moderate, ***large. Red and pink bars represent walkers. FR = front runners. MR = median runners. RW = runners/walkers. SR = slower runners. W = walkers.
Responders were normally distributed about a median of 27.5 to 30 min but with a tail of slower runners, runners/walkers, and walkers (Figure 1a). Of the sample, 51.5% were female, ranging from 4.2% for front runners to 80.3% for walkers (Figure 1b). The median age increased from 37.8 years for front runners to 56.9 years for walkers (Figure 1c). There were fewest participants from IMD Q1 (most deprived areas) and most from IMD Q4 (least deprived areas), with walkers more likely to be from deprived communities (Figure 1d). Around one-third of slower runners, runners/walkers, or walkers were inactive or did about one bout of activity per week at registration (Figure 1e).
Those faster than median runners showed significant demographic differences from walkers with large effect sizes. Slower runners and runners/walkers were statistically similar to walkers and were more likely to be female, older, from deprived communities, and less active at registration.
Health conditions
Figure 2 gives the characteristics of survey participants with health conditions ranked by average running time. Figure 2a shows that the proportion limited by at least one health condition lasting ≥12 months rose from 3% for front runners to 25% for slower runners, 28% for runners/walkers, and 45% for walkers. The overall proportion for the full sample was 9.3% (see Supplementary Table S1 for details). Slower runners, runners/walkers, and walkers had a median of two health conditions compared with a median of one health condition for the full sample. Slower runners, runners/walkers, and walkers collectively represented 4.3% of the sample and reported 19.8% of health conditions. The most reported conditions are shown in Figure 2b (see Supplementary Table S1 for details). For the full sample, the top five conditions were depression, arthritis, anxiety, asthma, and hypertension; slower runners, runners/walkers, and walkers also reported fibromyalgia, obesity, and chronic pain.
Figure 2. Characteristics of survey participants ranked by average running time: a) proportion limited by a health condition for ≥12 months; and b) proportion with each health condition (only top 10 conditions shown). Note: participants could have more than one health condition. Comparison with walkers using χ2 test at P<0.001 with effect sizes: *small, **moderate, ***large. FR = front runners. MR = median runners. RW = runners/walkers. SR = slower runners. W = walkers.
Motives for first participating and impact following participation
Supplementary Figure S1 illustrates responders’ motives for first participating in parkrun paired, where possible, with impact measures (see Supplementary Table S2 for details). The graphs are ranked in order of most to least selected motive for the full sample.
The three most selected motives were ‘to contribute to my fitness’ (57.0%), ‘to improve my physical health’ (37.2%), and ‘to gain a sense of personal achievement’ (27.2%); these had large proportions of people reporting improvements of 90.1%, 85.4%, and 91.4%, respectively.
Fewer slower runners, runner/walkers, and walkers selected ‘to contribute to my fitness’, while more selected ‘to improve my physical health’.
‘To manage my weight’ was selected by 19.6% of the sample and was more likely to be selected by slower runners (33.8%), runners/walkers (33.0%), and walkers (32.7%), with improvement for approximately 55% of runners with times slower than the median.
‘To improve or manage my health condition, disability, or illness’ was selected by 17.4% of those with a health condition and was more likely to be selected by walkers (31.5%). A total of 66.8% of all responders reported improvements to ‘your ability to manage your health condition, disability, or illness’, with no statistical differences between walkers and other participants.
Few selected as a motive ‘to improve my mental health’ (12.7%), ‘to feel part of a community’ (11.3%), or ‘to improve my happiness’ (6.5%). However, large proportions of responders reported improvements in these areas: 69.5%, 71.1%, and 79.6%, respectively. There were few statistical differences between walkers and other responders.
Few responders selected ‘to spend time outdoors’ (10.0%) or ‘to be active in a safe environment’ (3.9%), although the former was statistically more likely to be selected by walkers and the latter by runners slower than the median. ‘The amount of time you spend outdoors’ was improved for 74.8%, while ‘your ability to be active in a safe environment’ was improved for 60.0% of participants. There were higher values for walkers at 81.8% and 71.3%, respectively.
More than 20% of slower runners, runners/walkers, and walkers were more likely to select ‘my friends, family, or colleagues encouraged me to’ and, while more walkers selected ‘a health professional advised me to’, this was only 1.8% compared with 0.3% for the full sample. (It should be noted that the survey was carried out as parkrun practice was being set up.) Finally, 51.9% of the full sample improved ‘your overall lifestyle choices (for example, diet and smoking)’, with little difference between walkers and other responders.
