Scholarly article on topic 'Incidence, risk, and protective factors of bicycle crashes: Findings from a prospective cohort study in New Zealand'

Incidence, risk, and protective factors of bicycle crashes: Findings from a prospective cohort study in New Zealand Academic research paper on "Health sciences"

CC BY
0
0
Share paper
Academic journal
Preventive Medicine
OECD Field of science
Keywords
{Bicycling / "Wounds and injuries" / Incidence / Risk / Epidemiology / "Cohort studies" / "Medical record linkage"}

Abstract of research paper on Health sciences, author of scientific article — Sandar Tin Tin, Alistair Woodward, Shanthi Ameratunga

Abstract Objective To estimate the incidence and risk of medically or police attended bicycle crashes in a prospective cohort study in New Zealand. Method The Taupo Bicycle Study involved 2590 adult cyclists recruited from the country's largest cycling event in 2006 and followed over a median period of 4.6years through linkage to four administrative databases. Incidence rates with Poisson distribution confidence intervals were computed and Cox regression modelling for repeated events was performed. Results The 66 on-road crashes and 10 collisions per 1000 person-years corresponded to 240 crashes and 38 collisions per million hours spent road cycling. The risk increased by 6% and 8% respectively for an extra cycling hour each week. There were 50 off-road crashes per 1000 person-years. Residing in urban areas and in Auckland (region with the lowest level of cycling), riding in a bunch, using a road bike and experiencing a previous crash predicted a higher risk. Habitual use of conspicuity aids appeared to lower the risk. Conclusion The risk is higher in urban areas and where cycling is less common, and increased by bunch riding and previous crashes. These findings alongside the possible protective effect of conspicuity aids suggest promising approaches to improving cycle safety.

Academic research paper on topic "Incidence, risk, and protective factors of bicycle crashes: Findings from a prospective cohort study in New Zealand"

.JL^äke.

ELSEVIER

Incidence, risk, and protective factors of bicycle crashes: Findings from a prospective cohort study in New Zealand^

Sandar Tin Tin *, Alistair Woodward, Shanthi Ameratunga

Section of Epidemiology Biostatistics, School of Population Health, University of Auckland, Auckland, New Zealand

ARTICLE INFO ABSTRACT

Objective. To estimate the incidence and risk of medically or police attended bicycle crashes in a prospective cohort study in New Zealand.

Method. The Taupo Bicycle Study involved 2590 adult cyclists recruited from the country's largest cycling event in 2006 and followed over a median period of 4.6 years through linkage to four administrative databases. Incidence rates with Poisson distribution confidence intervals were computed and Cox regression modelling for repeated events was performed.

Results. The 66 on-road crashes and 10 collisions per 1000 person-years corresponded to 240 crashes and 38 collisions per million hours spent road cycling. The risk increased by 6% and 8% respectively for an extra cycling hour each week. There were 50 off-road crashes per 1000 person-years. Residing in urban areas and in Auckland (region with the lowest level of cycling), riding in a bunch, using a road bike and experiencing a previous crash predicted a higher risk. Habitual use of conspicuity aids appeared to lower the risk.

Conclusion. The risk is higher in urban areas and where cycling is less common, and increased by bunch riding and previous crashes. These findings alongside the possible protective effect of conspicuity aids suggest promising approaches to improving cycle safety.

© 2013 The Authors. Published by Elsevier Inc. All rights reserved.

Contents lists available at SciVerse ScienceDirect

Preventive Medicine

journal homepage: www.elsevier.com/locate/ypmed

CrossMark

Available online 21 May 2013

Keywords: Bicycling

Wounds and injuries

Incidence

Epidemiology Cohort studies Medical record linkage

Introduction

Regular cycling provides significant health (Andersen et al., 2000; Bassett et al., 2008; Oja et al., 2011) and other benefits (Higgins, 2005; Litman, 2012). Despite this, cycling is not a popular mode of travel in New Zealand (Tin Tin et al., 2009) and accounts for only 2% of total travel time (Ministry of Transport, 2012a). While the bicycle is increasingly used for sport and recreation activity, just over one-fifth of adults reported engaging in either road cycling or mountain biking at least once over twelve months in the most recent national survey (Sport New Zealand, 2009).

For many people, safety concerns are a major barrier to riding a bicycle (Kingham et al., 2009; Mackie, 2009; Winters et al., 2011) and it is true that cyclists bear a higher risk than most other types of road users if time-based exposure is considered (Tin Tin et al., 2010; Wardlaw, 2002). For each million hours spent cycling on New Zealand roads, 29 deaths or injuries resulted from collisions with a motor vehicle

☆ This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

* Corresponding author at: Section of Epidemiology and Biostatistics, School of Population Health, University of Auckland, Private Bag 92019, Auckland 1142, New Zealand.

E-mail addresses: s.tintin@auckland.ac.nz (S. Tin Tin), a.woodward@auckland.ac.nz (A. Woodward), s.ameratunga@auckland.ac.nz (S. Ameratunga).

(cf. 10 car driver deaths/injuries, 7 car passenger deaths/injuries and 5 pedestrian deaths/injuries) (Ministry of Transport, 2012b) and 31 injuries resulted in death or hospital inpatient treatment (cf. 2 driver injuries, 3 car passenger injuries and 2 pedestrian injuries) (Tin Tin et al., 2010). Furthermore, almost as many bicycle crashes occurred off-road (Munster et al., 2001).

Current statistics and epidemiological research in New Zealand and elsewhere (Amoros et al., 2011; Beck et al., 2007; Boufous et al., 2012; Buehler and Pucher, 2012; Garrard et al., 2010; Ministry of Transport, 2012b; Tin Tin et al., 2010) typically refer to a single official data source, either police reports or hospital records, which are known to under-count bicycle crashes (Elvik and Mysen, 1999; Langley et al., 2003; Tercero and Andersson, 2004). Other studies have relied on cross-sectional survey data (Aultman-Hall and Kaltenecker, 1999; Heesch et al., 2011 ; Moritz, 1997) thereby failing to account for reverse causation and potential biases (af Wâhlberg et al., 2010; Jenkins et al., 2002; Tivesten et al., 2012). Few prospective studies have reported the incidence and correlates of bicycle crash injuries (de Geus et al., 2012; Hoffman et al., 2010) but the findings could have been biased by differential loss to follow-up (Greenland, 1977).

This paper investigated the incidence of attended bicycle crashes and associated factors in a cohort of cyclists followed over a median period of 4.6 years. Attended bicycle crashes include those resulting in an admission to hospital, notification to the police or the Coroner (Medical Examiner), or a claim lodged with the Accident Compensation Corporation

0091-7435/$ - see front matter © 2013 The Authors. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/! 0.1016/j.ypmed.2013.05.001

(ACC), the government-funded universal no-fault injury compensation scheme.

Methods

Design, setting and participants

The Taupo Bicycle Study is a prospective cohort study with the sampling frame comprising cyclists, aged 16 years and over, who enrolled online in the Lake Taupo Cycle Challenge, New Zealand's largest mass cycling event held each November. Participants have varying degrees of cycling experience ranging from competitive sports cyclists to relative novices of all ages.

Recruitment was undertaken at the time of the 2006 event. Detailed methodology was described elsewhere (Thornley et al., 2008). In brief, email invitations, containing a hyperlink to the study information page, were sent to 5653 contestants who provided their email addresses at registration for the event. Those who agreed to participate in the study were taken to the next page containing a web questionnaire and asked about demographic characteristics, general cycling activity and crash experience in the past twelve months, and habitual risk/protective behaviours with options ranging from never to always. Copies of the questionnaire can be obtained from the authors. The questionnaire was completed and submitted by 2438 cyclists (43.1% response rate). Another 190 cyclists were recruited from the 2008 event by including a short description about the study in the event newsletter. Ethical approval was obtained from the University of Auckland Human Participants' Ethics Committee.

All participants were resurveyed in 2009 using a web questionnaire. The questionnaire asked about changes in cycling activity and risk/protective behaviours, as well as crash experience in the past twelve months, and was completed by 1537 cyclists (58.5% response rate).

injury outcome data

Injury outcome data were collected through record linkage to four administrative databases, covering the period from the date of recruitment to 30 June 2011. All participants consented to link their data to the following databases.

Insurance claims

In New Zealand, ACC provides personal injury cover for all residents and temporary visitors to New Zealand no matter who is at fault. The claims database is a major source of information on relatively minor injuries with over 80% of the claims related to primary care (e.g., GPs, emergency room treatment) only (Accident Compensation Corporation, 2012). Approval for record linkage was obtained from the ACC Research Ethics Committee.

Hospital discharge and mortality data

The hospital discharge data contains information about inpatients and day patients discharged from all public hospitals and over 90% of private hospitals in New Zealand. The mortality data contains information about all deaths registered in New Zealand. Diagnoses in each hospital visit and underlying causes of death are coded under ICD-10-AM. Bicycle crashes were identified using the E codes V10-V19; those that occurred on public roads were identified using the E codes V10-V18.3-9, V19.4-6, V19.9; and those that involved a collision with a motor vehicle were identified using the E codes V12-V14, V19.0-2 and V19.4-6. Readmissions were identified as described previously (Davie et al., 2011) and excluded.

Police reports

In New Zealand, it is mandatory that any fatal or injury crash involving a collision with a motor vehicle on a public road be reported to the police. This database therefore contains information on all police-reported bicycle collisions.

Match rate and data quality

There was a 99.0% match rate by the National Health Index number. The completeness of the linked data, based on the capture-recapture models, was 73.7% for all crashes, 74.5% for on-road crashes and 83.3% for collisions (Tin Tin et al., 2013). In comparison with self-reported data collected in 2009, the linked data had 63.1% sensitivity, 93.5% specificity and 59.0% positive predictive value for all crashes and 40.0% sensitivity, 99.9% specificity and 91.7% positive predictive value for collisions.

Analyses

The study sample was restricted to the 2590 participants who were resident in New Zealand at recruitment. All baseline data were complete for the 2435 participants (94.0%). Missing values were computed using multiple imputation with 25 complete datasets created by the Markov chain Monte Carlo method (Schafer, 1997), incorporating all baseline covariates and injury outcomes.

Bicycle crashes extracted through record linkage were categorised into on-road crashes (crashes that occurred on public roads) and others, as factors predicting these crashes may differ. Crashes involving a collision with a motor vehicle were also identified. As more than a single crash may be experienced during follow-up, incidence rates of repeated events were calculated using the person-years approach. Exposure-based incidence rates were also estimated for on-road crashes and collisions, using the average time spent road cycling at baseline. Confidence intervals were based on the Poisson distribution. The participants were censored on 30 June 2011 or date of death.

Cox proportional hazards regression modelling for repeated events was performed using a counting process approach and factors influencing the likelihood of experiencing crash episodes were identified. Hazard ratios (HRs) were first adjusted for cycling exposure and then adjusted for all covariates. SAS (release 9.2, SAS Institute Inc., Cary, North Carolina) was used for all analyses.

Probabilistic bias analyses (Lash et al., 2009) assessed the potential impact of outcome misclassification bias on association estimates, assuming that the sensitivity and specificity of the linked data ranged from 0.65 to 0.75 and from 0.94 to 0.99 respectively for on-road and other crashes and from 0.40 to 0.85 and from 0.98 to 1.00 respectively for collisions.

The impact of changes in exposures on association estimates was assessed by incorporating repeated measurements (at baseline and in 2009) of covari-ates in the Cox models. This analysis was restricted to 1526 cyclists who were resident in New Zealand and completed the second questionnaire.

Results

The participants' baseline characteristics are presented in Table 1. During a median follow-up of 4.6 years, six deaths occurred, of which one was due to a bicycle-car collision and five others were due to cancer. A total of 855 participants experienced 1336 bicycle crashes, of whom 32.4% experienced more than a single crash (Table 2). This corresponds to 116 crashes per 1000 person-years (95% CI: 109.93, 122.47) or 391 crashes per million hours spent cycling per year (95% CI: 370.38,412.62).

On-road bicycle crashes

There were 66 crashes per 1000 person-years or 240 crashes per million hours spent road cycling per year (Table 3). The adjusted HR for one hour increase in average time spent cycling each week was 1.06 (95% CI: 1.05-1.08). Age over 35 years, residing in urban areas or in the Auckland region, riding in a bunch, using a road bike and history of a crash at baseline predicted a higher risk whereas being overweight or obese, cycling off-road and using lights in the dark lowered the risk. Bicycle commuting, however, did not increase the risk.

Bicycle crashes involving a collision with a motor vehicle

There were 10 collisions per 1000 person-years or 38 collisions per million hours spent road cycling per year (Table 4). The adjusted HR for one hour increase in average time spent cycling each week was 1.08 (95% CI: 1.05-1.12). Due to a very small number of events, "overweight" and "obese" categories were combined and helmet use was excluded in the multivariate models. Residing in urban areas, riding a road bike and having a crash history were associated with an increased risk.

Table 1 (continued)

Table 1

Baseline characteristics of the participants in the Taupo Bicycle Study, New Zealand, 2006-2011.

Age (years)

16-35 579 22.4

36-50 1351 52.2

51 + 660 25.5

Gender

Male 1874 72.4

Female 715 27.6

Ethnicity

Maori 104 4.0

Non-Maori 2486 96.0

Level of education

High/secondary school or less 535 20.7

Polytechnic 654 25.3

University 1395 53.9

Missing 6 0.2

Body Mass Index

<25 1361 52.5

25-30 1000 38.6

30 + 214 8.3

Missing 15 0.6

NZDep2006 scoresa

1-3 1292 49.9

4-7 919 35.5

8-10 343 13.2

Missing 36 1.4

Urbanicity of residence

Main urban area 2013 77.7

Others 541 20.9

Missing 36 1.4

Region of residence

Auckland 919 35.5

Wellington 534 20.6

Others 1101 42.5

Missing 36 1.4

Cycling characteristics

Years of cycling

<1 204 7.9

1-4 1307 50.5

5+ 1068 41.2

Missing 11 0.4

Ever ride off-road

Yes 992 38.3

No 1584 61.2

Missing 14 0.5

Ever ride in the dark

Yes 1731 66.8

No 853 32.9

Missing 6 0.2

Ever ride in a bunch

Yes 1838 71.0

No 732 28.3

Missing 20 0.8

Cycle to work at least once a weekb

Yes 771 32.8

No 1541 65.6

Missing 36 1.5

Type of bicycle most commonly used

Road 2244 86.6

Mountain 195 7.5

Others 139 5.4

Missing 12 0.5

Crash in the past 12 months

Yes 801 30.9

No 1784 68.9

Missing 5 0.2

Always wear helmet

Yes 2555 98.6

No 25 1.0

Missing 10 0.4

Wear fluorescent colours

Always 759 29.3

Sometimes 1310 50.6

Never 500 19.3

Missing 21 0.8

Always use front and back lights in the darkc

Yes 1432 82.7

No 298 17.2

Missing 1 0.1

Use reflective materials in the darkc

Always 850 49.1

Sometimes 489 28.2

Never 386 22.3

Missing 6 0.3

Ever listen to music while riding

Yes 423 16.3

No 2153 83.1

Missing 14 0.5

a 2006 New Zealand Deprivation Index with decile ten the most deprived neighbourhood and decile one the least.

b Restricted to 2438 participants who reported travelling to work at least once a week.

c Restricted to 1731 participants who reported cycling in the dark.

Off-road bicycle crashes

There were 50 crashes per 1000 person-years (Table 5). The risk was lower in university graduates, overweight or obese cyclists and less experienced cyclists but higher in those who cycled in the dark or in a bunch and those who had a crash history.

The effect estimates mentioned above were similar to those obtained from complete case analyses. Potential misclassification of crash outcomes during the linkage process may underestimate the actual incidence rate and may bias the hazard ratios to the null (Appendix A). Likewise, potential misclassification of exposures (due to changes over time) may underestimate the risk estimates in most cases (Appendix B).

Discussion

Main findings

In this study, cyclists experienced 116 crashes attended medically or by police per 1000 person-years, of which 66 occurred on the road and 10 involved a collision with a motor vehicle. There were 240 on-road crashes and 38 collisions per million hours spent road cycling and the risk increased by 6% and 8% respectively for one hour increase in cycling each week. After adjusting for all covariates, participants' age, body mass index, urbanity, region of residence, cycling off road, in the dark or in a bunch, type of bicycle used and prior crash history predicted the crash risk with variations in effect estimates by crash type.

Table 2

Number of participants who experienced bicycle crashes in the Taupo Bicycle Study, New Zealand, 2006-2011.

No. of All crashes On-road Off-road Collisions with

crashes crashes crashes a motor vehicle

N % N % N % N %

1 578 22.3 421 16.3 328 12.7 91 3.5

2 161 6.2 90 3.5 66 2.6 10 0.4

3 66 2.5 26 1.0 19 0.7 3 0.1

4 28 1.1 10 0.4 9 0.3

5 14 0.5 3 0.1 4 0.2

6 4 0.2 2 0.1 0 0.0

7 1 0.0 0 0.0 0 0.0

8 2 0.1 0 0.0 1 0.0

9 1 0.0 1 0.0

Total 855 33.0 553 21.4 427 16.5 104 4.0

Incidence of on-road bicycle crashes in the Taupo Bicycle Study, New Zealand, 2006-2011.

No Rate per 1000 person-years (95% CI) Rate per million hour cycled-years (95% CI) Crude hazard ratio ( 95% CI) Adjusted hazard ratioa 95% CI) Adjusted hazard ratiob (95% CI)

Total 755 65.60 (61.00, 70.45) 240.47 223.62, 258.26)

Age (years) 16-35 36-50 51 + 158 385 212 61.51 64.06 72.32 (52.29, 71.88) (57.82, 70.80) (62.92, 82.74) 217.95 239.44 262.78 185.29, 254.71) 216.11,264.59) (228.60, 300.64) 1.00 1.04 ( 0.87,1.25) 1.18 ( 0.96,1.45) 1.00 1.07 ( 0.89, 1.29) 1.16 (0.94, 1.43) 1.00 1.20 (0.98, 1.47) 1.37 (1.08, 1.74)

Gender Male Female 574 181 68.65 57.48 (63.15, 74.51) (49.41, 66.49) 245.08 226.95 (225.44, 265.98) 195.09, 262.52) 1.20 ( 1.01,1.41) 1.00 1.14 ( 0.97, 1.36) 1.00 1.06 (0.89, 1.27) 1.00

Ethnicity Maori Non-Maori 23 732 48.93 66.31 (31.02, 73.43) (61.59, 71.29) 192.82 242.36 122.23, 289.33) 225.11,260.57) 0.74 ( 0.49,1.12) 1.00 0.77 0.51, 1.18) 1.00 0.94 (0.62, 1.44) 1.00

Education High/secondary school or less Polytechnic University Missing 161 180 413 1 67.23 61.58 67.00 36.26 (57.24, 78.45) (52.91, 71.26) (60.69, 73.78) (0.92, 202.01) 226.28 226.16 253.96 166.67 192.68, 264.06) 194.33, 261.73) 230.05, 279.67) 4.22, 928.61) 1.00 ( 0.83,1.21) 0.92 ( 0.77,1.10) 1.00 0.93 ( 0.77,1.11) 0.89 ( 0.74,1.06) 1.00 1.04 (0.86, 1.25) 0.96 (0.80, 1.15) 1.00

Body Mass Index < 25 25-30 30 + Missing 444 268 42 1 73.14 60.29 45.10 15.90 (66.50, 80.27) (53.29, 67.96) (32.50, 60.96) (0.40, 88.57) 251.97 238.81 176.42 58.14 229.07, 276.53) 211.08, 269.18) 127.15, 238.47) 1.47, 323.93) 1.00 0.82 ( 0.71, 0.96) 0.62 0.45, 0.85) 1.00 0.89 (0.76,1.04) 0.66 (0.48,0.91) 1.00 0.89 (0.76, 1.04) 0.69 (0.49, 0.95)

NZDep2006 scoresc 1-3 4-7 8-10 Missing 382 287 83 3 66.17 70.74 54.16 20.36 (59.70, 73.15) (62.79, 79.41) (43.14, 67.14) (4.20, 59.50) 243.62 262.43 194.13 59.43 219.80, 269.32) 232.95, 294.62) 154.63, 240.66) 12.26,173.69) 1.00 1.06 ( 0.91,1.24) 0.82 ( 0.65,1.04) 1.00 1.07 (0.92, 1.25) 0.80 (0.63, 1.02) 1.00 1.19 (1.02, 1.40) 0.93 (0.73, 1.19)

Urbanicity of residence Main urban area Others Missing 644 108 3 72.18 44.25 20.36 (66.71, 77.98) (36.30, 53.43) (4.20, 59.50) 262.30 170.36 59.43 242.43, 283.36) 139.75, 205.68) 12.26,173.69) 1.00 0.60 0.49, 0.74) 1.00 0.62 (0.50, 0.76) 1.00 0.70 (0.56, 0.87)

Region of residence Auckland Wellington Others Missing 337 143 272 3 83.26 59.72 55.28 20.36 (74.61,92.64) (50.33, 70.34) (48.91,62.26) (4.20, 59.50) 293.29 238.05 203.08 59.43 262.80, 326.33) 200.63, 280.42) 179.66, 228.70) 12.26,173.69) 1.00 0.72 0.59, 0.87) 0.67 0.57, 0.78) 1.00 0.75 (0.62, 0.92) 0.67 (0.57, 0.78) 1.00 0.83 (0.68,1.01) 0.80 (0.67, 0.95)

Years of cycling <1 1-4 5+ Missing 42 388 322 3 45.77 66.59 68.30 59.33 (32.99, 61.87) (60.12, 73.55) (61.04, 76.18) (12.24,173.39) 225.86 256.53 224.69 369.69 162.78, 305.30) 231.64,283.37) 200.81, 250.62) 76.24,1080.40) 0.67 0.49, 0.93) 0.98 (0.84,1.14) 1.00 0.79 (0.57,1.10) 1.06 (0.91,1.23) 1.00 0.95 (0.68,1.32) 1.08 (0.92,1.27) 1.00

Ever ride off-road Yes No Missing 239 513 3 54.36 72.76 48.12 (47.69, 61.71) (66.59, 79.33) (9.92, 140.63) 222.76 250.24 180.07 195.41, 252.86) 229.05, 272.86) 37.14, 526.25) 0.75 0.64, 0.87) 1.00 0.78 (0.67, 0.91) 1.00 0.74 (0.63, 0.87) 1.00

Ever ride in the dark Yes No Missing 563 192 0 72.79 51.23 0.00 (66.90, 79.06) (44.24, 59.02) 244.93 230.23 0.00 225.11,266.02) 198.82, 265.20) 1.42 ( 1.20,1.67) 1.00 1.27 (1.07,1.50) 1.00 1.06 (0.88,1.28) 1.00

Ever ride in a bunch Yes No Missing 620 130 5 76.03 39.81 55.60 (70.17, 82.26) (33.26, 47.27) (18.05,129.76) 254.62 190.79 214.57 234.97, 275.48) 159.41, 226.55) 69.67, 500.74) 1.90 ( 1.57, 2.31) 1.00 1.70 (1.41,2.07) 1.00 1.49 (1.21,1.82) 1.00

Cycle to work at least once a week Yes 270 No 475 Missing 10 74.46 61.74 52.45 (65.84, 83.89) (56.32, 67.56) (25.15, 96.45) 231.69 247.06 193.58 204.87, 261.03) 225.34,270.31) 92.83, 356.00) 1.19 ( 1.02,1.39) 1.00 1.07 (0.92,1.25) 1.00 1.06 (0.88,1.29) 1.00

Type of bicycle most commonly used Road 707 Mountain 29 Others 17 Missing 2 70.84 33.17 28.30 36.26 (65.72, 76.26) (22.22, 47.64) (16.49,45.31) (4.39,130.97) 249.30 217.07 107.70 163.40 231.26, 268.37) 145.37,311.74) 62.74,172.44) 19.79, 590.26) 1.00 0.47 0.32, 0.68) 0.40 0.25, 0.65) 1.00 0.56 (0.39, 0.83) 0.38 (0.23, 0.62) 1.00 0.72 (0.49,1.06) 0.42 (0.26, 0.69)

Crash in the past 12 months Yes No Missing 299 451 5 84.08 (74.82, 94.17) 56.87 (51.74, 62.37) 217.54 (70.63, 507.67) 283.89 217.00 618.57 252.62,317.96) 197.43, 237.98) 200.85,1443.53) 1.48 ( 1.28,1.72) 1.00 1.42 (1.22,1.64) 1.00 1.34 (1.15,1.56) 1.00

Rate per 1000 person-years (95% CI)

Rate per million hour cycled-years (95% CI)

Crude hazard ratio (95% CI)

Adjusted hazard ratioa (95% CI)

Adjusted hazard ratiob (95% CI)

747 65.81 (61.17,70.70) 8 70.86 (30.59,139.61) 0 0.00

202 59.29 (51.40, 68.06)

404 69.82 (63.18,76.97)

148 66.66 (56.35,78.31)

1 10.36 (0.26, 57.72)

Always wear helmet Yes No

Missing

Wear fluorescent colours Always Sometimes Never Missing

Always use front and back lights in the darkd Yes 454 71.07 (64.68, 77.92)

No 109 81.24 (66.71, 98.00)

Missing 0 0.00

Use reflective materials in the darkd Always 265

Sometimes 160

Never 137

Missing 1

Ever listen to music while riding

Yes 128

No 621

Missing 6

70.95 (62.66, 80.03) 72.08 (61.35,84.16) 78.10 (65.57, 92.32) 18.82 (0.48,104.83)

68.16 (56.86, 81.04) 64.91 (59.90, 70.22) 93.23 (34.21, 202.93)

240.68 (223.73, 258.58) 298.21 (128.75, 587.60) 0.00

215.81 (187.07, 247.71) 252.67 (228.63, 278.55) 255.13 (215.68, 299.71) 40.71 (1.03,226.84)

231.95 (211.10, 254.30) 319.84 (262.62, 385.82) 0.00

230.75 (203.80, 260.27) 240.61 (204.77, 280.92) 286.30 (240.37, 338.45) 72.63 (1.84,404.65)

232.28 (193.79, 276.19) 241.07 (222.48, 260.80) 476.89 (175.01, 1037.98)

0.97 (0.48,1.98) 1.00

0.89 (0.72,1.10) 1.04 (0.86,1.25) 1.00

0.88 (0.71,1.08) 1.00

0.90 (0.73,1.11) 0.91 (0.73,1.15) 1.00

1.06 (0.87,1.28) 1.00

0.92 (0.45,1.86) 1.00

0.87 (0.71,1.08)

1.01 (0.83,1.22) 1.00

0.82 (0.66,1.01) 1.00

0.86 (0.70,1.06)

0.86 (0.68,1.08) 1.00

1.04 (0.86,1.25) 1.00

0.77 (0.38,1.58) 1.00

0.93 (0.74,1.16)

1.00 (0.82,1.22) 1.00

0.74 (0.59, 0.92) 1.00

0.93 (0.74,1.18) 0.84 (0.66,1.08) 1.00

1.12 (0.92,1.36) 1.00

a Adjusted for time spent cycling on the road. b Adjusted for all variables.

c 2006 New Zealand Deprivation Index with decile ten the most deprived neighbourhood and decile one the least. d Restricted to 1731 participants who reported cycling in the dark.

Strengths and limitations

This is one of the very few prospective cohort studies involving cyclists and used record linkage to obtain objective information on bicycle crashes from multiple databases. This resource efficient method of data collection was also designed to minimise potential biases associated with loss to follow-up (Greenland, 1977) and self-reports (af Wahlberg et al., 2010; Jenkins et al., 2002; Tivesten et al., 2012).

While emigration during follow-up is a potential issue in using the linked data, this accounted for less than 2% of the participants resurveyed in 2009 and may not substantially influence outcome occurrences (Kristensen and Bjerkedal, 2010). This analysis excludes very minor crashes not requiring either medical or police attention, which represents 70% or more of self-reported crashes (de Geus et al., 2012; Hoffman et al., 2010; Tin Tin et al., 2013). Case ascertainment may also be affected by personal, social and health service factors (Cryer and Langley, 2008; Lyons et al., 2005) as well as inaccuracies in individual data sources (Davie et al., 2008; Health Outcomes International Pty Ltd., 2005; McDonald et al., 2009) and in record linkage. Notwithstanding these limitations, the reasonably high specificity of the linked data enhanced the ability of this study (compared with previous research) to provide unbiased risk ratios (Blakely and Salmond, 2002; Howe, 1998). Moreover, probabilistic bias analyses were undertaken to account for residual biases.

Our analysis used exposure data collected at baseline to predict the risk of future crashes. Participants may have changed their cycling behaviours during follow-up. In the resurvey conducted in 2009,44% of the responders reported the same amount of cycling, 23% reported more cycling, 28% reported less cycling and 5% reported no cycling. Exposure misclassification of this kind is likely to underestimate risk estimates (Andersen, 2004).

Finally, our participants are not representative of all New Zealand cyclists. Compared with adult cyclists who participated in a national survey conducted in 2007/08 (Sport New Zealand, 2009), the study sample has more over 35 year olds (64% vs. 78%), males (60% vs.

72%) and non-Maori (89% vs. 96%) but fewer who reside in low deprivation (first two quintiles of deprivation scores) areas (85% vs. 61%). These differences may have minimal impact on risk estimates (Lash et al., 2009) but limit generalizability of incidence rate estimates.

interpretation

This study, based on multiple data sources, identified many more crashes than previously published New Zealand data (Ministry of Transport, 2012b; Tin Tin et al., 2010). The Auckland region, which has the lowest prevalence of active travel in the country (Tin Tin et al., 2009), had a higher risk of on-road bicycle crashes. Given differences in definitions and methodologies of data collection, analysis and presentation, it is hard to make comparisons with studies elsewhere (Appendix C), but it appears that exposure-based injury rates are lower in countries or regions with a higher level of cycling. This phenomenon, described as "safety in numbers", has been reported in many places (Ekman, 1996; Jacobsen, 2003; Leden et al., 2000; Robinson, 2005; Tin Tin et al., 2011). However, regardless of the prevalence of cycling, the health benefits gained from regular cycling outweigh additional injuries or deaths from crashes (Holm et al., 2012; Lindsay et al., 2011; Rojas-Rueda et al., 2012).

Previous studies reported demographic differences in cycling injuries but the results varied. Males and children were over-represented in official statistics (Amoros et al., 2011; Boufous et al., 2012; Tin Tin et al., 2010; Yan et al., 2011) but not in self-reports (de Geus et al., 2012; Heesch et al., 2011; Hoffman et al., 2010). In this study, the risk of on-road crashes was higher in older age groups and the risk of collisions appeared to be higher in younger cyclists and males. There was a lower risk of all crashes in overweight or obese cyclists.

In this study, commuting with a bicycle did not predict an increased risk of on-road crashes, in accordance with previous Australian research (Heesch et al., 2011). It is noteworthy because bicycle commuting, as a means to engage in regular physical activity, is more likely to be adopted and sustained compared with traditional exercise programmes

Incidence of bicycle-motor vehicle collisions in the Taupo Bicycle Study, New Zealand, 2006-2011.

No Rate per 1000 person-years (95% CI) Rate per million hour cycled-years ( 95% CI) Crude hazard ratio ( 95% CI) Adjusted hazard ratioa 95% CI) Adjusted hazard ratiob (95% CI)

Total 120 10.43 (8.64,12.47) 38.22 31.69,45.70)

Age (years)

16-35 32 12.46 8.52,17.59) 44.14 30.19, 62.32) 1.00 1.00 1.00

36-50 57 9.48 7.18,12.29) 35.45 26.85,45.93) 0.76 ( 0.49,1.17) 0.78 ( 0.51,1.22) 0.84 (0.52, 1.35)

51 + 31 10.58 7.19,15.01) 38.43 26.11,54.54) 0.85 ( 0.52,1.39) 0.83 0.50, 1.37) 0.89 (0.50, 1.58)

Gender

Male 93 11.12 8.98,13.63) 39.71 32.05,48.65) 1.30 ( 0.85, 2.00) 1.23 ( 0.80, 1.90) 1.18 (0.75, 1.87)

Female 27 8.57 (5.65, 12.48) 33.85 22.31,49.26) 1.00 1.00 1.00

Ethnicity

Maori 4 8.51 2.32, 21.79) 33.53 9.14, 85.86) 0.81 0.30, 2.20) 0.86 ( 0.31, 2.37) 0.94 (0.34, 2.59)

Non-Maori 116 10.51 (8.68, 12.60) 38.41 31.74,46.06) 1.00 1.00 1.00

Education

High/secondary school or less 23 9.60 6.09,14.41) 32.33 20.49,48.51) 0.78 ( 0.48,1.27) 0.87 ( 0.54,1.41) 0.89 (0.55, 1.45)

Polytechnic 28 9.58 6.37,13.84) 35.18 23.38, 50.85) 0.82 ( 0.52,1.28) 0.87 ( 0.55,1.35) 0.97 (0.62, 1.53)

University 68 11.03 8.57,13.99) 41.81 32.47, 53.01) 1.00 1.00 1.00

Missing 1 36.26 0.92, 202.01) 166.67 4.22, 928.61)

Body Mass Index

< 25 69 11.37 8.84,14.39) 39.16 30.47, 49.56) 1.00 1.00 1.00

25 + 50 9.30 6.90, 12.26) 36.76 27.28, 48.46) 0.82 ( 0.57,1.19) 0.91 ( 0.62,1.32) 0.96 (0.65, 1.42)

Missing 1 15.90 0.40, 88.57) 58.14 1.47, 323.93)

NZDep2006 scoresc

1-3 66 11.43 8.84,14.55) 42.09 32.55, 53.55) 1.00 1.00 1.00

4-7 42 10.35 7.46,13.99) 38.41 27.68, 51.91) 0.89 ( 0.60,1.32) 0.91 0.61, 1.35) 1.01 (0.68, 1.50)

8-10 12 7.83 4.05, 13.68) 28.07 14.50, 49.03) 0.69 ( 0.37,1.28) 0.67 ( 0.36, 1.25) 0.74 (0.39, 1.38)

Missing 0 0.00 0.00

Urbanicity of residence

Main urban area 107 11.99 9.83,14.49) 43.58 35.72, 52.66) 1.00 1.00 1.00

Others 13 5.33 2.84,9.11) 20.51 10.92, 35.07) 0.43 0.24, 0.78) 0.44 0.25, 0.80) 0.53 (0.29, 0.98)

Missing 0 0.00 0.00

Region of residence

Auckland 54 13.34 10.02,17.41) 47.00 35.30, 61.32) 1.00 1.00 1.00

Wellington 27 11.28 7.43,16.40) 44.95 29.62, 65.39) 0.85 ( 0.53,1.35) 0.90 ( 0.56,1.43) 0.97 (0.60, 1.55)

Others 39 7.93 5.64,10.84) 29.12 20.71, 39.80) 0.60 ( 0.39, 0.91) 0.59 0.39, 0.90) 0.79 (0.50, 1.23)

Missing 0 0.00 0.00

Years of cycling

<1 6 6.54 2.40,14.23) 32.27 11.84, 70.23) 0.59 ( 0.25,1.38) 0.71 ( 0.30,1.68) 0.69 (0.29, 1.67)

1-4 61 10.47 8.01,13.45) 40.33 30.85, 51.81) 0.94 ( 0.65,1.36) 1.03 ( 0.71,1.51) 0.96 (0.65, 1.43)

5+ 53 11.24 8.42, 14.70) 36.98 27.70,48.37) 1.00 1.00 1.00

Missing 0 0.00 0.00

Ever ride off-road

Yes 38 8.64 6.12, 11.86) 35.42 25.06,48.61) 0.76 ( 0.52,1.12) 0.79 ( 0.54,1.17) 0.76 (0.51,1.14)

No 81 11.49 9.12, 14.28) 39.51 31.38,49.11) 1.00 1.00 1.00

Missing 1 16.04 0.41, 89.37) 60.02 1.52, 334.44)

Ever ride in the dark

Yes 84 10.86 8.66, 13.45) 36.54 29.15,45.24) 1.13 ( 0.76,1.68) 0.98 ( 0.66,1.47) 0.71 (0.45,1.14)

No 36 9.61 6.73, 13.30) 43.17 30.23, 59.76) 1.00 1.00 1.00

Missing 0 0.00 0.00

Ever ride in a bunch

Yes 94 11.53 9.32, 14.11) 38.60 31.20, 47.24) 1.50 (0.96, 2.35) 1.32 (0.84, 2.07) 1.14 (0.71,1.83)

No 25 7.66 4.95, 11.30) 36.69 23.74, 54.16) 1.00 1.00 1.00

Missing 1 11.12 0.28, 61.96) 42.91 1.09, 239.10)

Cycle to work at least once a week

Yes 44 12.13 8.82, 16.29) 37.76 27.43, 50.69) 1.30 (0.89,1.91) 1.15 (0.78,1.69) 1.13 (0.74,1.72)

No 73 9.49 7.44, 11.93) 37.97 29.76, 47.74) 1.00 1.00 1.00

Missing 3 15.73 3.24,45.98) 58.07 11.98,169.72)

Type of bicycle most commonly used

Road 114 11.42 9.42,13.72) 40.20 33.16,48.29) 1.00 1.00 1.00

Others 6 4.07 1.49, 8.85) 20.59 7.56,44.81) 0.35 ( 0.15, 0.81) 0.39 (0.17, 0.89) 0.43 (0.18,1.01)

Missing 0 0.00 0.00

Crash in the past 12 months

Yes 56 15.75 11.90, 20.45) 53.17 40.16, 69.05) 2.02 ( 1.40, 2.90) 1.92 (1.33, 2.76) 1.86 (1.27, 2.70)

No 62 7.82 5.99,10.02) 29.83 22.87, 38.24) 1.00 1.00 1.00

Missing 2 87.02 10.54, 314.33) 247.43 29.96, 893.79)

Rate per 1000 person-years (95% CI)

Rate per million hour cycled-years (95% CI)

Crude hazard ratio (95% CI)

Adjusted hazard ratioa (95% CI)

Adjusted hazard ratiob (95% CI)

119 10.48 (8.68,12.55) 1 8.86 (0.22,49.35) 0 0.00

38 11.15 (7.89,15.31)

54 9.33 (7.01,12.18)

28 12.61 (8.38,18.23)

0 0.00

Always wear helmet Yes No

Missing

Wear fluorescent colours Always Sometimes Never Missing

Always use front and back lights in the darkd Yes 70 10.96 (8.54,13.84)

No 14 10.43 (5.70,17.51)

Missing 0 0.00

Use reflective materials in the darkd Always 43

Sometimes 17

Never 24

Missing 0

11.51 (8.33,15.51) 7.66 (4.46,12.26) 13.68 (8.77, 20.36) 0.00

Ever listen to music while riding

Yes 24

Missing 0

12.78 (8.19,19.01) 10.03 (8.13,12.25) 0.00

38.34 (31.76,45.88) 37.28 (0.94, 207.69) 0.00

40.60 (28.73, 55.72) 33.77 (25.37, 44.07) 48.27 (32.07, 69.76) 0.00

35.76 (27.88,45.18) 41.08 (22.46, 68.93) 0.00

37.44 (27.10, 50.43) 25.57 (14.89,40.93) 50.15 (32.14, 74.63) 0.00

43.55 (27.91, 64.80) 37.27 (30.19,45.51) 0.00

1.24 (0.17, 9.08) 1.00

0.89 (0.54,1.45) 0.73 (0.46,1.16) 1.00

1.05 (0.59,1.i 1.00

0.83 (0.50,1.38) 0.55 (0.25,1.04) 1.00

1.25 (0.79,1.96) 1.00

1.15 (0.16, 8.45) 1.00

0.87 (0.53,1.42) 0.70 (0.44,1.11) 1.00

0.97 (0.54,1.74) 1.00

0.78 (0.47,1.30) 0.50 (0.26, 0.95) 1.00

1.22 (0.78,1.93) 1.00

0.81 (0.49,1.32) 1.12 (0.66,1.90) 1.00

0.88 (0.48,1.61) 1.00

0.88 (0.49,1.57) 0.51 (0.26, 0.99) 1.00

1.23 (0.76,1.97) 1.00

a Adjusted for time spent cycling on the road. b Adjusted for all variables.

c 2006 New Zealand Deprivation Index with decile ten the most deprived neighbourhood and decile one the least. d Restricted to 1731 participants who reported cycling in the dark.

(Hillsdon et al., 1995) but is deterred by safety concerns for many people (Mackie, 2009; van Bekkum et al., 2011).

While many cyclists feel safer in a group than alone (O'Connor and Brown, 2010), our findings showed that participants who ever rode in a bunch had a higher crash risk. The data did not allow us to determine if the crashes occurred while riding in a bunch. Consequently, it was not possible to distinguish risk factors associated with cycling in a peloton (such as high speeds or reduced warning of road hazards) from characteristics of bunch riders, who tend to be more experienced and, possibly, take greater risks in traffic (Johnson et al., 2009). This is an area for future research.

This study revealed that cyclists with a bicycle crash history were more likely to experience crash episodes during follow-up. This does not fit the findings from a US study (Hoffman et al., 2010) but is consistent with "accident proneness" which assumes that injuries tend to cluster within persons. This concept was introduced decades ago (Farmer and Chambers, 1926; Greenwood and Woods, 1919) and confirmed in a meta-analysis (Visser et al., 2007) but was challenged by a recent study (Hamilton et al., 2011). A broader term "accident liability" emphasises the role of multiple factors in injury causation (Farmer and Chambers, 1926; Kune, 1985). These are beyond the scope of this analysis but are worthy of further evaluation.

While conspicuity aids are effective in improving detection and recognition time by drivers (Kwan and Mapstone, 2009), the effect of such measures on cyclist safety is not yet conclusive. In this analysis, using lights reduced the risk of on-road crashes but the effectiveness of other conspicuity aids was not clear as in a US cohort study (Hoffman et al., 2010). The protective effect of fluorescent colours found in our previous analysis may be due to failure to exclude off-road crashes (Thornley et al., 2008). In any case, our study design did not allow us to account for details of the circumstances of the crash, such as weather, lighting, road and traffic conditions. Cyclists' acute behaviour, that is, immediately prior to a crash, may be more relevant to crash risk and was examined in a case-control study (Hagel et al., 2012). The study observed that the risk of collisions

was increased by wearing fluorescent clothing but decreased by wearing white or coloured clothing. The findings, however, may be complicated by potential biases due to differential misclassification of exposure, traffic risk and other risk behaviours. These issues will need to be considered in future research.

Conclusion

Bicycle crashes are relatively common in this cohort and the risk varies by demographic and cycling characteristics. In particular, the risk of on-road injuries is higher in the region with the lowest level of active travel, supporting the safety in numbers effect. Bunch riding and previous crash experience also place cyclists at risk of all crashes. These factors and the possible protective effect of conspicuity aids are worthy of exploration in future research and cycle safety initiatives.

Abbreviations

ACC Accident Compensation Corporation HR Hazard Ratio

Conflict of interest statement

The authors declare that there are no conflicts of interest.

Acknowledgments

We thank the participating cyclists and organisers of the Lake Taupo Cycle Challenge for their support, and Professor John Langley, Professor Anthony Rodgers and Dr Simon Thornley for their initial contribution to the study. Our thanks also go to the Accident Compensation Corporation, Ministry of Health and New Zealand Transport Agency for the provision of bicycle crash data. This study was funded by grant 09/142 from the Health Research Council of New Zealand.

Incidence of off-road bicycle crashes in the Taupo Bicycle Study, New Zealand, 2006-2011.

No Rate per 1000 person-years (95% CI) Crude hazard ratio ( 95% CI) Adjusted hazard ratioa (95% CI) Adjusted hazard ratiob (95% CI)

Total 581 50.48 (46.46, 54.76)

Age (years)

16-35 134 52.16 43.71,61.78) 1.00 1.00 1.00

36-50 294 48.49 43.49, 54.85) 0.94 ( 0.77,1.15) 1.03 (0.83, 1.26) 1.03 (0.83,1.29)

51 + 153 52.20 44.26, 61.15) 1.00 ( 0.79,1.26) 1.14 (0.90, 1.44) 1.16 (0.89,1.50)

Gender

Male 443 52.99 48.17, 58.16) 1.21 ( 1.00,1.47) 1.18 (0.97, 1.43) 1.05 (0.86,1.29)

Female 138 43.82 36.82, 51.78) 1.00 1.00 1.00

Ethnicity

Maori 23 48.93 31.02, 73.43) 0.97 ( 0.64,1.47) 0.95 (0.62, 1.44) 1.08 (0.70,1.65)

Non-Maori 558 50.54 (46.44, 54.92) 1.00 1.00 1.00

Education

High/secondary school or less 122 50.94 (42.30, 60.82) 1.06 ( 0.85,1.32) 1.21 (0.92, 1.60) 1.13 (0.91,1.40)

Polytechnic 170 58.16 (49.75, 67.59) 1.21 ( 1.00,1.47) 1.29 (1.02, 1.65) 1.28 (1.05,1.56)

University 285 46.23 41.02, 51.93) 1.00 1.00 1.00

Missing 4 145.03 ( 39.52, 371.33)

Body Mass Index

<25 350 57.66 51.77, 64.03) 1.00 1.00 1.00

25-30 193 43.42 37.51, 50.00) 0.75 ( 0.63, 0.90) 0.80 (0.67, 0.96) 0.78 (0.65, 0.94)

30 + 36 38.65 27.07, 53.51) 0.67 0.47, 0.94) 0.73 (0.52, 1.04) 0.76 (0.54,1.08)

Missing 2 31.79 3.85,114.85)

NZDep2006 scoresc

1-3 297 51.45 45.76, 57.65) 1.00 1.00 1.00

4-7 195 48.06 41.55, 55.30) 0.93 ( 0.77,1.11) 0.92 (0.76, 1.10) 0.97 (0.81,1.17)

8-10 85 55.47 44.31, 68.59) 1.08 ( 0.85,1.38) 1.05 (0.83, 1.35) 1.12 (0.87,1.44)

Missing 4 27.15 7.40, 69.50)

Urbanicity of residence

Main urban area 464 52.01 47.38, 56.96) 1.00 1.00 1.00

Others 113 46.30 38.16, 55.67) 0.89 ( 0.72,1.09) 0.87 (0.71, 1.07) 0.89 (0.71,1.12)

Missing 4 27.15 7.40, 69.50)

Region of residence

Auckland 201 49.66 43.03, 57.02) 1.00 1.00 1.00

Wellington 127 53.03 44.21, 63.10) 1.07 ( 0.86,1.34) 0.99 (0.79, 1.25) 1.00 (0.79,1.25)

Others 249 50.61 44.52, 57.30) 1.02 ( 0.85,1.23) 0.95 (0.79, 1.15) 0.98 (0.80,1.21)

Missing 4 27.15 7.40, 69.50)

Years of cycling

<1 28 30.51 20.28, 44.10) 0.51 0.34, 0.75) 0.51 (0.34, 0.76) 0.62 (0.41, 0.93)

1-4 267 45.82 40.49, 51.66) 0.76 (0.64, 0.90) 0.80 (0.67, 0.94) 0.83 (0.69, 0.99)

5+ 284 60.24 53.44, 67.67) 1.00 1.00 1.00

Missing 2 39.55 4.79,142.88)

Ever ride in the dark

Yes 439 56.76 51.57, 62.32) 1.51 (1.25,1.83) 1.38 (1.14, 1.68) 1.20 (0.98,1.48)

No 141 37.62 31.67, 44.37) 1.00 1.00 1.00

Missing 1 36.26 0.92, 202.01)

Ever ride in a bunch

Yes 474 58.13 53.01,63.61) 1.86 (1.50, 2.31) 1.84 (1.47, 2.29) 1.60 (1.27, 2.02)

No 102 31.24 25.47, 37.92) 1.00 1.00 1.00

Missing 5 55.60 18.05,129.76)

Type of bicycle most commonly used

Road 513 51.41 47.05, 56.05) 1.00 1.00 1.00

Mountain 48 54.91 40.48, 72.80) 1.06 (0.79,1.44) 0.77 (0.56,1.06) 0.88 (0.64,1.23)

Others 17 28.30 16.49,45.31) 0.56 (0.34, 0.91) 0.56 (0.34,0.91) 0.63 (0.39,1.04)

Missing 3 54.39 11.22,158.94)

Crash in the past 12 months

Yes 221 62.15 54.22, 70.90) 1.37 (1.16,1.63) 1.26 (1.06,1.50) 1.20 (1.01,1.43)

No 359 45.27 40.71, 50.20) 1.00 1.00 1.00

Missing 1 43.51 1.10,242.41)

Always wear helmet

Yes 575 50.66 46.60, 54.97) 1.15 (0.47, 2.81) 1.24 (0.51, 3.02) 1.17 (0.48, 2.83)

No 5 44.28 14.38,103.35) 1.00 1.00 1.00

Missing 1 21.75 0.55,121.21)

Always use front and back lights in the darkd

Yes 360 56.35 50.68, 62.49) 0.95 (0.75,1.22) 1.02 (0.79, 1.31) 0.98 (0.76,1.26)

No 79 58.89 46.62, 73.39) 1.00 1.00 1.00

Missing 0 0.00

No Rate per 1000 person-years (95% CI) Crude hazard ratio (95% CI) Adjusted hazard ratioa (95% CI) Adjusted hazard ratiob (95% CI)

Ever listen to music while riding

Yes 111 59.10 (48.62,71.18) 1.19 (0.96,1.47) 1.09 (0.88,1.34) 1.14 (0.92,1.42)

No 469 49.02 (44.68,53.67) 1.00 1.00 1.00

Missing 1 15.54 (0.39, 86.58)

a Adjusted for time spent cycling in general and % time spent cycling off-road. b Adjusted for all variables.

c 2006 New Zealand Deprivation Index with decile ten the most deprived neighbourhood and decile one the least. d Restricted to 1731 participants who reported cycling in the dark.

Appendix A, B and C. Supplementary data

Supplementary data to this article can be found online at http:// dx.doi.org/10.1016/j.ypmed.2013.05.001.

References

Accident Compensation Corporation, 2012. Annual Report 2012. ACC, Wellington.

af Wahlberg, A.E., Dorn, L., Kline, T., 2010. The effect of social desirability on self reported and recorded road traffic accidents. Transp. Res. F Psychol. Behav. 13,106-114.

Amoros, E., Chiron, M., Thelot, B., Laumon, B., 2011. The injury epidemiology of cyclists based on a road trauma registry. BMC Public Health 11, 653. http://dx.doi.org/ 10.1186/471-2458-11-653.

Andersen, L.B., 2004. Relative risk of mortality in the physically inactive is underestimated because of real changes in exposure level during follow-up. Am. J. Epidemiol. 160,189-195.

Andersen, L.B., Schnohr, P., Schroll, M., Hein, H.O., 2000. All-cause mortality associated with physical activity during leisure time, work, sports, and cycling to work. Arch. Intern. Med. 160,1621-1628.

Aultman-Hall, L., Kaltenecker, M.G., 1999. Toronto bicycle commuter safety rates. Accid. Anal. Prev. 31, 675-686.

Bassett Jr., D.R., Pucher, J., Buehler, R., 2008. Walking, cycling, and obesity rates in Europe, North America, and Australia. J. Phys. Act. Health 5, 795-814.

Beck L.F., Dellinger, A.M., O'Neil, M.E., 2007. Motor vehicle crash injury rates by mode of travel, United States: using exposure-based methods to quantify differences. Am. J. Epidemiol. 166, 212-218.

Blakely, T., Salmond, C., 2002. Probabilistic record linkage and a method to calculate the positive predictive value. Int. J. Epidemiol. 31, 1246-1252.

Boufous, S., de Rome, L., Senserrick, T., Ivers, R., 2012. Risk factors for severe injury in cyclists involved in traffic crashes in Victoria, Australia. Accid. Anal. Prev. 49, 404-409.

Buehler, R., Pucher, J., 2012. Walking and cycling in Western Europe and the United States. TR News 280, 34-42.

Cryer, C., Langley, J., 2008. Developing Indicators of Injury Incidence That can be Used to Monitor Global, Regional and Local Trends. Injury Prevention Research Unit, University of Otago, Dunedin.

Davie, G., Langley, J., Samaranayaka, A., Wetherspoon, M.E., 2008. Accuracy of injury coding under ICD-10-AM for New Zealand public hospital discharges. Inj. Prev. 14, 319-323.

Davie, G., Samaranayaka, A., Langley, J.D., Barson, D., 2011. Estimating person-based injury incidence: accuracy of an algorithm to identify readmissions from hospital discharge data. Inj. Prev. 17, 338-342.

de Geus, B., Vandenbulcke, G., Int Panis, L., et al., 2012. A prospective cohort study on minor accidents involving commuter cyclists in Belgium. Accid. Anal. Prev. 45, 683-693.

Ekman, L., 1996. On the treatment of flow in traffic safety analysis — a non-parametric approach applied on vulnerable road users. Bulletin, 136. Institutionen for Trafikteknik, Lunds (Tekniska Hogskola, Lund).

Elvik, R., Mysen, A.B., 1999. Incomplete accident reporting: meta-analysis of studies made in 13 countries. Transp. Res. Rec. 1665,133-140.

Farmer, E., Chambers, E., 1926. A psychological study of individual differences in accident rates. Report No. 38. Industrial Fatigue Research Board, London.

Garrard, J., Greaves, S., Ellison, A., 2010. Cycling injuries in Australia: road safety's blind spot? J. Australas. Coll. Road Saf. 21, 37-43.

Greenland, S., 1977. Response and follow-up bias in cohort studies. Am. J. Epidemiol. 106,184-187.

Greenwood, M., Woods, H., 1919. The incidence of industrial accidents upon individuals with special reference to multiple accidents. Report No. 4. Industrial Fatigue Research Board, London.

Hagel, B., Ruest, N., Morgunov, N., et al., 2012. Do visibility aids reduce the risk of motor-vehicle injury in bicyclists? Inj. Prev. 18, A36-A37.

Hamilton, G.M., Meeuwisse, W.H., Emery, C.A., Steele, R.J., Shrier, I., 2011. Past injury as a risk factor: an illustrative example where appearances are deceiving. Am. J. Epidemiol. 173, 941-948.

Health Outcomes International Pty Ltd., 2005. Methods and Systems Used to Measure and Monitor Occupational Disease and Injury in New Zealand: NOHSAC Technical Report 2. National Occupational Health and Safety Advisory Committee (NOHSAC), Wellington.

Heesch, K.C., Garrard, J., Sahlqvist, S., 2011. Incidence, severity and correlates of bicycling injuries in a sample of cyclists in Queensland, Australia. Accid. Anal. Prev. 43, 2085-2092.

Higgins, P.A.T., 2005. Exercise-based transportation reduces oil dependence, carbon emissions and obesity. Environ. Conserv. 32, 197-202.

Hillsdon, M., Thorogood, M., Anstiss, T., Morris, J., 1995. Randomised controlled trials of physical activity promotion in free living populations: a review. J. Epidemiol. Commun. Health 49, 448-453.

Hoffman, M.R., Lambert, W.E., Peck, E.G., Mayberry, J.C., 2010. Bicycle commuter injury prevention: it is time to focus on the environment. J. Trauma 69,1112-1119.

Holm, A.L., Glumer, C., Diderichsen, F., 2012. Health impact assessment of increased cycling to place of work or education in Copenhagen. BMJ Open 2, e001135. http://dx.doi.org/10.1136/bmjopen-2012-35.

Howe, G.R., 1998. Use of computerized record linkage in cohort studies. Epidemiol. Rev. 20,112-121.

Jacobsen, P.L., 2003. Safety in numbers: more walkers and bicyclists, safer walking and bicycling. Inj. Prev. 9, 205-209.

Jenkins, P., Earle-Richardson, G., Slingerland, D.T., May, J., 2002. Time dependent memory decay. Am. J. Ind. Med. 41, 98-101.

Johnson, M., Oxley, J., Cameron, M., 2009. Cyclist Bunch Riding: A Review of the Literature. Monash University Accident Research Centre, Victoria.

Kingham, S., Koorey, G., Taylor, K., 2009. Attracting the next 10% of cyclists with the right infrastructure. New Zealand Cycling Conference, New Plymouth.

Kristensen, P., Bjerkedal, T., 2010. Dealing with emigration in cohort studies: follow-up of mortality and cancer incidence among Norwegians born between 1967 and 1976. Eur. J. Epidemiol. 25,155-161.

Kune, J.B., 1985. Accident liability. Br. J. Ind. Med. 42, 336-340.

Kwan, I., Mapstone, J., 2009. Interventions for increasing pedestrian and cyclist visibility for the prevention of death and injuries. Cochrane Database Syst. Rev. 4.

Langley, J.D., Dow, N., Stephenson, S., Kypri, K., 2003. Missing cyclists. Inj. Prev. 9, 376-379.

Lash, T., Fox, M., Fink, A., 2009. Applying Quantitative Bias Analysis to Epidemiological Data. Springer, New York.

Leden, L., Garder, P., Pulkkinen, U., 2000. An expert judgment model applied to estimating the safety effect of a bicycle facility. Accid. Anal. Prev. 32, 589-599.

Lindsay, G., Macmillan, A., Woodward, A., 2011. Moving urban trips from cars to bicycles: impact on health and emissions. Aust. N. Z.J. Public Health 35, 54-60.

Litman, T., 2012. Evaluating Transportation Equity: Guidance for Incorporating Distributional Impacts in Transportation Planning. Victoria Transport Policy Institute, Victoria.

Lyons, R.A., Brophy, S., Pockett, R., John, G., 2005. Purpose, development and use of injury indicators. Int. J. Inj. Contr. Saf. Promot. 12, 207-211.

Mackie, H., 2009. 'I want to ride my bike': overcoming barriers to cycling to intermediate schools. New Zealand Transport Agency Research Report No. 380. New Zealand Transport Agency, Wellington.

McDonald, G., Davie, G., Langley, J., 2009. Validity of police-reported information on injury severity for those hospitalized from motor vehicle traffic crashes. Traffic Inj. Prev. 10,184-190.

Ministry of Transport, 2012a. Comparing Travel Modes. Ministry of Transport, Wellington.

Ministry of Transport, 2012b. Risk on the Road. Pedestrians, Cyclists and Motorcyclists. Ministry of Transport, Wellington.

Moritz, W., 1997. Survey of North American bicycle commuters: design and aggregate results. Transp. Res. Rec. 1578, 91-101.

Munster, D., Koorey, G., Walton, D., 2001. Role of road features in cycle-only crashes in New Zealand. Transfund New Zealand Research Report No. 211. Transfund New Zealand, Wellington.

O'Connor, J.P., Brown, T.D., 2010. Riding with the sharks: serious leisure cyclist's perceptions of sharing the road with motorists. J. Sci. Med. Sport 13, 53-58.

Oja, P., Titze, S., Bauman, A., et al., 2011. Health benefits of cycling: a systematic review. Scand. J. Med. Sci. Sports 21,496-509.

Robinson, D.L., 2005. Safety in numbers in Australia: more walkers and bicyclists, safer walking and bicycling. Health Promot. J. Austr. 16,47-51.

Rojas-Rueda, D., de Nazelle, A., Teixido, O., Nieuwenhuijsen, M.J., 2012. Replacing car trips by increasing bike and public transport in the greater Barcelona metropolitan area: a health impact assessment study. Environ. Int. 49,100-109.

Schafer, J., 1997. Analysis of Incomplete Multivariate Data. Chapman & Hall, London.

Sport New Zealand, 2009. Sport and Recreation Profile: Cycling — Findings from the 2007/08 Active New Zealand Survey. Sport New Zealand, Wellington.

Tercero, F., Andersson, R., 2004. Measuring transport injuries in a developing country: an application of the capture-recapture method. Accid. Anal. Prev. 36,13-20.

Thornley, S.J., Woodward, A., Langley, J.D., Ameratunga, S.N., Rodgers, A., 2008. Conspicuity and bicycle crashes: preliminary findings of the Taupo Bicycle Study. Inj. Prev. 14,11-18.

Tin Tin, S., Woodward, A., Thornley, S., Ameratunga, S., 2009. Cycling and walking to work in New Zealand, 1991-2006: regional and individual differences, and pointers to effective interventions. Int. J. Behav. Nutr. Phys. Act. 6, 64. http://dx.doi.org/10.1186/ 479-5868-6-64.

Tin Tin, S., Woodward, A., Ameratunga, S.N., 2010. Injuries to pedal cyclists on New Zealand roads, 1988-2007. BMC Public Health 10, 655. http://dx.doi.org/10.1186/ 471-2458-10-655.

Tin Tin, S., Woodward, A., Thornley, S., Ameratunga, S., 2011. Regional variations in pedal cyclist injuries in New Zealand: safety in numbers or risk in scarcity? Aust. N. Z. J. Public Health 35, 357-363.

Tin Tin, S., Woodward, A., Ameratunga, S., 2013. Completeness and accuracy of crash outcome data in a cohort of cyclists: a validation study. BMC Public Health 13, 420. http://dx.doi.org/10.1186/1471-2458-13-420.

Tivesten, E., Jonsson, S., Jakobsson, L., Norin, H., 2012. Nonresponse analysis and adjustment in a mail survey on car accidents. Accid. Anal. Prev. 48,401-415.

van Bekkum, J.E., Williams, J.M., Morris, P.G., 2011. Cycle commuting and perceptions of barriers: stages of change, gender and occupation. Health Educ. 111, 476-497.

Visser, E., Pijl, Y.J., Stolk, R.P., Neeleman, J., Rosmalen, J.G.M., 2007. Accident proneness, does it exist? A review and meta-analysis. Accid. Anal. Prev. 39, 556-564.

Wardlaw, M., 2002. Assessing the actual risks faced by cyclists. Traffic Eng. Contr. 43, 420-424.

Winters, M., Davidson, G., Kao, D.N., Teschke, K., 2011. Motivators and deterrents of bicycling: comparing influences on decisions to ride. Transportation 38, 153-168.

Yan, X., Ma, M., Huang, H., Abdel-Aty, M., Wu, C., 2011. Motor vehicle-bicycle crashes in Beijing: irregular maneuvers, crash patterns, and injury severity. Accid. Anal. Prev. 43,1751-1758.