BMJ  2003;327:1455-1456 (20 December), doi:10.1136/bmj.327.7429.1455

Hazardous journey

Car colour and risk of car crash injury: population based case control study

¹ Section of Epidemiology and Biostatistics, School of Population Health, University of Auckland, Private Bag 92019, Auckland, New Zealand, ² George Institute for International Health, University of Sydney, Sydney, Australia

Globally, road traffic crashes kill about 3000 people a day.¹ Identification of modifiable risk factors is an important step in reducing this burden. Previous research suggests that white or light coloured cars are less likely to be involved in a crash, than cars of other colours.² We investigated the effect of car colour on the risk of a serious injury from a crash, using a population based case control study designed to identify and quantify modifiable risk factors.

Participants, methods, and results

The Auckland car crash injury study was conducted in the Auckland region of New Zealand between April 1998 and June 1999.^{3 4} The study population comprised all drivers of cars on public (urban and rural) roads in the region. Cases (n = 571) were all car drivers involved in crashes in which one or more of the occupants of the car were admitted to hospital or died (response rate 93%). Controls (n = 588) were car drivers identified by cluster sampling of drivers from randomly selected sites on the road network, at randomly selected times, representative of all time spent driving in the study region during the study (response rate 79%). Data on car colour were available for all of the controls and all but four of the cases.

White was the most prevalent colour in both cases and controls. White, black, grey, red, and silver colour groups were relatively homogeneous with respect to shades included in each group. There was no consistent pattern in distribution of car colour by age of vehicle.

In the multivariable analysis, we assessed the effect of the following potential confounders: age of driver, sex, educational level, ethnicity, alcohol consumption (in previous six hours), use of recreational drugs, seat-belt use, average time spent driving each week, vehicle speed, vehicle age, engine size, registration, warrant of fitness and vehicle insurance, driving licence status, road type, weather, and ambient light conditions (day, night, twilight). Confounders were included in the final model if they resulted in a change in the odds ratio of more than 5% and had no more than 5% missing data.⁵

We found a significant reduction in the risk of serious injury in silver cars compared with white cars in both the univariate analysis (odds ratio = 0.5 (95% confidence interval 0.3 to 0.8)) and the multivariable analysis (adjusted odds ratio = 0.4 (0.2 to 0.9); table). There was a significant increased risk of a serious injury in brown vehicles after confounders had been adjusted for (odds ratio = 2.1 (1.1 to 4.2)) and the risks for black and green cars were also raised (adjusted odds ratio = 2.0 (1.0 to 4.2) and 1.8 (1.0 to 3.6) respectively). However, green and brown colour groups were heterogeneous in terms of shades of colours included. The risk of a serious injury in yellow, grey, red, and blue cars was not significantly different from that in white cars.

View this table: [in this window] [in a new window]   Association of car colour with car crash injury in Auckland

 

Comment

Silver cars were about 50% less likely to be involved in a crash resulting in serious injury than white cars. The design and methods are a reasonable approach to study the association between modifiable risk factors and injury from car crashes. The association between silver car colour and reduced risk of serious injury persisted after we had adjusted for major confounding factors, but the possibility of residual confounding remains. The extent to which these results are generalisable to other settings is open to question. Increasing the proportion of silver cars could be an effective passive strategy to reduce the burden of injury from car crashes.

---

We thank the study participants and I Civil, R Dunn, and J Bailey for help with the design of the study.

Contributors: SF was responsible for this analysis and writing the paper; JC, ER, RN, SA, and RJ contributed to the design and conduct of the study and the writing of the paper; ER and JC contributed to this analysis. JC is the guarantor.

Funding: The Health Research Council of New Zealand funded the Auckland car crash injury study and the Road Safety Research Trust funded this analysis.

Competing interests: None declared.

References

1. Peden M, McGee K, Sharma G. The injury chart book: a graphical overview of the global burden of injuries. Geneva: World Health Organization, 2002.
2. Lardelli-Claret P, De Dios Luna-Del-Castillo J, Juan Jimenez-Moleon J, Femia-Marzo P, Moreno-Abril O, et al. Does vehicle color influence the risk of being passively involved in a collision? Epidemiology 2002;13: 721-4.[Medline]
3. Connor J, Norton R, Ameratunga S, Robinson E, Civil I, Dunn R, et al. Driver sleepiness and the risk of serious injury to car occupants: a population based case-control study. BMJ 2002;324: 1125-8.[Abstract/Free Full Text]
4. Connor J, Norton R, Ameratunga S, Robinson E, Wigmore B, Jackson R. Prevalence of driver sleepiness in a random population-based sample of car driving. Sleep 2001;24: 688-94.[ISI][Medline]
5. Greenland S. Modeling and variable selection in epidemiologic analysis. Am J Public Health 1989;79: 340-9.[Abstract/Free Full Text]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Reddit    Technorati    What's this?

Rapid Responses:

Read all Rapid Responses

A Silver Car Catastrophe in Surrey

mike c harvey
bmj.com, 19 Dec 2003 [Full text]

Will the authors please supply important data

Judith Ronat
bmj.com, 20 Dec 2003 [Full text]

Statistics are revealing! Do they really help?

Rajan T.D.
bmj.com, 20 Dec 2003 [Full text]

Re: A Silver Car Catastrophe in Surrey

Dennis E. Staszak
bmj.com, 20 Dec 2003 [Full text]

Does the season affect color accident rates?

Patricia I Cushman
bmj.com, 20 Dec 2003 [Full text]

Unconvincing

Theodore D. Sternberg
bmj.com, 20 Dec 2003 [Full text]

The Invisible Silver Car

ken doe
bmj.com, 20 Dec 2003 [Full text]

Theory as to why darker colors such as black and green are more dangerous

Randy C. Kady
bmj.com, 20 Dec 2003 [Full text]

but who was responsible?

Nicholas D Moore
bmj.com, 20 Dec 2003 [Full text]

Other confounders?

Melissa K Raven
bmj.com, 20 Dec 2003 [Full text]

Anything but silver

Phillip J. Colquitt
bmj.com, 22 Dec 2003 [Full text]

Lies, Damned Lies and Statistics

David I Saunders
bmj.com, 23 Dec 2003 [Full text]

Effect of car color was not tested

David A. Ziegler
bmj.com, 24 Dec 2003 [Full text]

Not everything that shines under the sun is Silver

Shlomo M Monnickendam
bmj.com, 26 Dec 2003 [Full text]

Other factors

Mario J G Cordeiro
bmj.com, 27 Dec 2003 [Full text]

Not definitive but interesting

Michael Low
bmj.com, 28 Dec 2003 [Full text]

Does vehicle size, environmental light, or weather inluence accident rates?

Ferdinand Frauscher, et al.
bmj.com, 29 Dec 2003 [Full text]

The distinction between victims and villains

Y Adi, et al.
bmj.com, 9 Jan 2004 [Full text]

Statisticaly worthless

chris t sutor
bmj.com, 12 Jan 2004 [Full text]

Statsitically of some value, just hard to interpret

Matthew L Grove
bmj.com, 13 Jan 2004 [Full text]

A silver machine…

Peter Morrell
bmj.com, 15 Jan 2004 [Full text]

Author's response

Sue Furness
bmj.com, 18 Jan 2004 [Full text]

Re: Statisticaly worthless

Laith Al-Hasani
bmj.com, 2 May 2005 [Full text]

Another study

Malcolm Tattersall
bmj.com, 18 Jan 2006 [Full text]