Table 1

Cycling exposure measures calculated from the EPAQ2 questionnaire administered at the second health assessment

ExposureCalculation
Commuter cyclingRespondents were asked how frequently they normally travelled to work by car, public transport, bike or on foot (response options were ‘always’, ‘usually’, ‘occasionally’ or ‘never/rarely’). Responses were converted to fractions (always=1, usually=0.75, occasionally=0.25, never/rarely=0). Participants reported the distance between home and work and the average number of times per week they made this journey (multiplied by two to account for the return journey). When cycling was the only mode selected, the total weekly distance cycled was calculated by multiplying the distance from home to work by the number of journeys made. When cycling was selected alongside other modes, the distance cycled was weighted according to the frequency of cycling relative to the frequency of the other modes reported. For example, if a respondent selected ‘always’ for cycling and driving, it was assumed that cycling accounted for 10%, and driving for 90% of the distance travelled. The total number of journeys was then multiplied by the weighted distance travelled (miles/week)
Non-commuting utility cyclingRespondents were asked to recall the average number of journeys they made by bicycle to get about apart from going to work for each of the following distances: ‘less than 0.5 miles’, ‘0.5 mile–1.5 miles’, ‘1.5–2.5 miles’, ‘2.5–3.5 miles’, ‘3.5–5.5 miles’, and ‘more than 5.5 miles’. The total weekly distance travelled was computed by multiplying the reported number of trips by the midpoint value of each distance category (assumed to be 0.25 for <0.5 mile and 6 for >5.5 miles). These values were then summed to provide a measure of distance travelled (miles/week)
All utility cyclingDistance travelled for non-commuting utility cycling was added to the distance travelled for commuting cycling to derive a measure of total utility cycling (miles/week)
Recreational cyclingRespondents reported the average time spent ‘cycling for pleasure’ per session and the frequency of such sessions: ‘none’, ‘less than once a month’, ‘once a month’, ‘2–3 times a month’, ‘once a week’, ‘2–3 times a week’, ‘4–5 times a week’, or ‘everyday’. The average weekly cycling duration was computed by converting the frequency into a weekly numerical value (eg, 0.5/52 for ‘less than once a month’ and (2.5×12)/52 for ‘2–3 times per month’). The time spent cycling (min/week) was computed by multiplying the average session duration by the average weekly frequency
Total cyclingTo enable a measure of total cycling to be derived and to allow for comparisons with previous studies, the distance travelled for utility cycling was converted into an estimated duration. Based on self-report data from a recent study of UK adults, we assumed an average cycling speed of 10 miles/h.27 A measure of total time spent cycling (min/week) was derived by summing time spent in commuting, other utility and recreational cycling