All Party Parliamentary Group
on Smoking and Health
Call for Evidence

Smoking in private vehicles:
Comments on the Peer-reviewed literature

Imperial Tobacco Group

25 October 2011

http://www.imperial-tobacco.com

Republished with permission by the Libertarian Alliance

Overall comments

Imperial  Tobacco  believes  that  the  scientific evidence, taken  as  a  whole,  is  insufficient to establish that other people’s tobacco smoke is a cause of any disease.

The statistical population studies (epidemiology) which have led to claims that other people’s tobacco smoke is a risk to health are subject to methodological flaws. Taken as a whole, the studies do not demonstrate a difference in the incidence of disease between non-smokers who are exposed to other people’s tobacco smoke and non-smokers who are not. Combining the studies through a process called meta-analysis resolves none of the original flaws and creates additional flaws.

The peer-reviewed literature tends to fall into 3 categories, with some overlap:

i)          papers  that  survey  the  prevalence  of  Environmental  Tobacco  Smoke  (“ETS”)

exposure in vehicles;

ii)         quantitative measurements of ETS chemical markers in vehicles; and iii)      the claimed health effects from reported exposure to ETS in vehicles.

The  claimed  health  effects  reported  include  wheeze  (borderline  statistical  significance), myocardial infarction (no association), lung cancer (no association) and chronic bronchitis (reported statistically significant association). All of these studies are subject to the following flaws: i) exposure and disease endpoints are self-reported and likely to be inaccurate; ii) the retrospective nature of the studies leaves them open to recall bias which is likely to lead to over- estimations of relative risk; iii) residual confounding i.e. not properly adjusted for other risk factors. None of these studies is sufficient to establish that ETS in vehicles was a factor in any the claimed diseases.

Where levels of ETS markers are quantified e.g. PM2.5, comparisons have been made with air quality   guidelines  which   appear   to   have   been   exceeded  significantly.  However   these comparisons are invalid: air quality guidelines are 24-h averages whereas the studies were conducted over much shorter periods of exposure (typically 5-30 mins) and often peak, not average exposure levels, were reported.

The studies all assume that ETS constitutes a risk to health and many refer to theUSSurgeon General’s report of 2006, in which it was concluded that “there is no risk-free level of exposure to second-hand smoke”.

1.  Prevalence

Several papers reported on prevalence of ETS in vehicles without measuring the levels of exposure but simply asking subjects if they are exposed to cigarette smoke in a vehicle (Roberts et al, 1996; Gillespie et al, 2005; Marshall et al, 2006; Lee et al, 2008; Mbulo, 2008; Leatherdale et al, 2009; Wolfson et al, 2009; and Glover et al, 2011). In some papers there was an attempt to assess the levels by asking questions about the smokiness of the car or frequency of exposure. This is extremely inaccurate and open to recall bias. In none of the reports were any of the conclusions validated with biochemical verifications of the levels of exposure e.g. urinary cotinine.

The quality of much of this data is further called into question by the fact that many of the questionnaires had not been validated and several leading questions were included.

Interview techniques included telephone interviews, web surveys and face-to-face interviews. The accuracy of the answers is likely to vary with these techniques and is also influenced by the experience of the interviewer.

Although a GP’s diagnosis would always be considered best practice, most of these studies relied on self-reported disease endpoints which are subjective and likely to be inaccurate.

Traffic and climate conditions are known to influence the non-tobacco specific markers used but this was neither commented on nor adjusted for.

2.  Quantitative measurements of ETS markers

Several studies have reported quantitative measurements of ETS markers including PM2.5, nicotine, CO, and ultrafine particles or UFP (Muramatsu et al, 1987; Edwards et al, 2006; Rees et al, 2006; Vardarvas et al, 2006; Ott et al, 2008; Jones et al, 2009; Sendzik et al, 2009; Liu et al,

2010; Hongji S et al, 2010). It should be noted that most of these markers are not specific to ETS and tobacco.

The studies differed in quality but some attempted to include multiple ventilation parameters in moving vehicles in different types of traffic (Rees et al, 2006; Liu et al, 2011) whereas others simply burned cigarettes in stationary cars (Vardavas et al, 2006).

It has been noted by Pawson et al (2011) that ETS is a complex mixture and as such the concentrations of the chemicals vary according to different conditions of the internal volume of the car, the speed it is travelling at and the level of ventilation employed, and therefore it can be difficult to assess the actual exposure with any degree of precision.

The majority of studies have focused on PM2.5  levels and when comparing the studies there is considerable variation. For example:

No ventilation & motionless car:     3,800 to 13,000 μg/m3

Half open windows:                        162 to 12,000 μg/m3

Open windows:                               51- 5000 μg/m3

Many of the large variations reflect differences in the overall study design with the highest value in each range from studies in which cigarettes were left to burn in the vehicle during the analysis.

Pawson et al (2011) in their review of 6 such studies noted that the levels monitored in the vehicles fluctuated with traffic conditions, climatic conditions, speed and type of vehicle, duration of journey, number of passengers and ventilation conditions.

Where PM2.5 levels were monitored, researchers tended to quote either peak levels or else the average level over the short time monitored in the vehicle, typically 5-30 minutes. The authors then compared these levels with ambient air guideline limits such as those from the US Environmental Protection Agency (“EPA”). The EPA guideline limits are:

15 μg/m3         Annual

35 μg/m3         24h average

Typically, the 24-h average is calculated by taking the recorded level a number of times over an hour, adding these together and dividing by the number of readings. These hourly readings are then taken consecutively in groups of 24 and the 24 hourly averages calculated1 . It is therefore not valid to compare a measurement taken over less than one hour to a 24-hour average air quality standard. Indeed, Hongji et al (2010) note that there is no US EPA standard for PM2.5 and the standard available is for 24-h average levels while the exposure in vehicles is short-term.

Matt et al (2008) measured residual contamination of cars with ETS by quantifying nicotine from surface wipes and in dust and air samples. Smokers’ cars demonstrated significantly higher levels of nicotine compared to non-smokers. Fortmann et al (2010) appears to be an extension and development of the Matt (2008) study cited above.  Residual tobacco smoke was quantified by analysing nicotine from surface wipes and in dust and air samples; in addition car interiors were inspected for dustiness and signs of past smoking. The authors state that this should be considered a preliminary study. They also note that cleaning procedures employed in vehicles could contribute to the formation of secondary pollutants including carcinogens which are not present in ETS. In both studies, the number of cigarettes smoked in the car per week was self- reported and the ventilation parameters, speeds driven, and temperature were not controlled for.

Crucially Pawson et al (2011) stated that “…we are still short of a benchmark to indicate whether any of the reported levels may be declared safe or dangerous.”

1 EssexHealth Protection Unit (Dec 2010) Part of the Health Protection Agency. Air Quality Factsheet 6. Particles – PM10 & PM2.5. http://www.hpa.org.uk/web/HPAwebFile/HPAweb_C/1294740079203

3.  Claims about adverse health effects

Sly et al (2007) measured exposure of ETS in cars with risk of persistent wheeze in adolescents. They did measure lung function using spirometry and so didn’t depend on self-reported wheeze. However, the odds ratios (“ORs”) were close to or under 2. Relative risks (“RRs”) below 2 are subject to residual confounding and therefore this association is questionable. Unfortunately this is only a letter with no data and therefore the primary data have not been subjected to peer- review.

Kabir et al (2009) report on respiratory health effects in children after exposure to ETS in cars. The exposure to ETS was self-reported. ETS exposure in cars showed borderline significant association with wheeze and hey fever symptoms but not with asthma.

Kreuzer et al (2000) conducted a case-control study to investigate ETS exposure and lung cancer risk. ETS in cars was not associated with lung cancer (OR 1.15, CI 0.76-1.75). This is a sub- study of a larger study by Boffetta et al (1998) which actually found no association.

Muscatand Wynder (1995) conducted a hospital-based case-control study to investigate the potential association between myocardial infarction and ETS. ETS exposure in cars was not associated with myocardial infarction in men or women.

Evans et al (2009) surveyed 64,961 Canadian non-smokers aged 12yr+ who were asked about exposure  to  ETS  in  cars  and  chronic  bronchitis.  Authors  report  a  statistically  significant association of 2.25 (1.42-3.58). The disease endpoint was not verified by a GP diagnosis, to be diagnosed as having chronic bronchitis the patient must be observed for at least two years. The questions regarding ETS exposure were subject to recall bias.

Glover et al (2011) reported a statistically significant association between ETS exposure in cars and smoking initiation (RR 1.87, CI: 1.43-2.44). RRs below 2 are subject to residual confounding and therefore this association is questionable.

Brody et al (2011) investigated the occupancy of nicotinic receptors in the brain for a small number of smokers and non-smokers exposed to ETS. The authors reported no differences between smokers and non-smokers stating that nicotine from ETS crosses the blood-brain barrier and that this is likely to be higher in infants (although this was not verified). However, the car was stationary and the windows closed apart from <1cm opening on passenger side.

4.  Other papers

MacKenzie and Freeman (2010) traced the “…second-hand smoke is 23 times more toxic in a vehicle than in the home” statistic, which has been widely reported in the media and peer- reviewed literature to a quote in the “Rocky Mountain News” newspaper in Denver, Colorado. No empirical evidence to support this statement was located. However, these authors then focus on the paper by Rees et al, 2006 which attempts to quantify PM2.5 under various driving conditions with windows ‘open’ (all four windows opened) or ‘closed’ (driver’s window opened 5cm). The authors reported mean respirable particle (“RSP”) concentrations of 272 μg/m3 when closed and

51 μg/m3 when opened. Measurements were taken over 5 min but were still compared to the EPA

24-h average guidelines.

Zhou et al (2000) reported measurements in “passenger cars” but these are actually trains carrying 150 passengers in which 25% were smoking. All the windows were shut; therefore the paper is not relevant to the discussion on private vehicles.

Martin et al (2006) conducted a prevalence study by observation. Teams of observers were placed at 5 sites inWellington,New Zealandto record smoking prevalence in cars. They reported

4.1% point prevalence smoking in cars and attempted to note if windows were open. However, it would be difficult to impossible for an observer standing 1-2m from the car to ascertain the level of ventilation.

Park et al (1998) measured air exchange rate and ETS levels (measured as RSP) in a stationary vehicle and then used this to simulate driving cycles. No fans, air conditioning or recirculation were used. Based on modelled data the authors claimed that levels of RSP and formaldehyde could be potentially harmful to sensitive populations e.g. asthmatics

Conclusion

This provides a comprehensive review of the available peer-reviewed scientific literature which relates to  smoking in  vehicles.   Some papers attempted to  survey the  prevalence of  ETS exposure in vehicles, and some attempted to equate a number of claimed health effects with a self-reported exposure to ETS in vehicles.

All of these studies are subject to significant flaws: exposure and disease endpoints were self- reported and likely to be inaccurate; studies are open to recall bias which is likely to lead to over- estimations of relative risk; and the studies were not properly adjusted for other risk factors. None of these studies is sufficient to establish that ETS in vehicles was a factor in any of the claimed diseases.

Where levels of ETS markers were quantified, e.g. PM2.5, comparisons have been made with air quality   guidelines  which   appear   to   have   been   exceeded  significantly.  However   these comparisons are not valid: air quality guidelines are 24-h averages whereas the studies were conducted over much shorter periods of exposure (typically between 5-30 minutes) and often peak, not average exposure levels, were reported.

The peer reviewed science, taken as a whole, does not justify a ban on smoking in private vehicles on public health grounds.

Contact:

We would welcome any opportunity to discuss the issues presented in this document in more detail. If you would like to do so please contact:

Dr Steve Stotesbury
Head of Regulatory Science
Imperial Tobacco Limited
PO Box 244
Upton Road
Bristol
BS99 7UJ

References

Boffetta P et al (1998). Multicenter case-control study of exposure to environmental tobacco smoke and lung cancer inEurope. J Natl Cancer Inst. 901440-50.

Brody AL et al (2011). Effect of secondhand smoke on occupancy of nicotinic acetylcholine receptors in brain. Arch Gen Psychiatry. 2011 Sep;68(9):953-60. Epub 2011 May 2.

Edwards R et al (2006). Highly hazardous air quality associated with smoking in cars: New

Zealandpilot study. N Z Med J. 2006 Oct 27;119(1244).

Evans J, Chen Y.   The association between home and vehicle environmental tobacco smoke

(ETS) and chronic bronchitis in a Canadian population: the Canadian Community Health Survey,

2005. Inhal Toxicol. 2009 Feb;21(3):244-9.

Fortmann AL, Romero RA, Sklar M, Pham V, Zakarian J, Quintana PJ, Chatfield D, Matt GE. Residual tobacco smoke in used cars: futile efforts and persistent pollutants. Nicotine Tob Res.

2010 Oct;12(10):1029-36. Epub 2010 Aug 30.

Gillespie  J  et  al  (2005).  Secondhand smoke  in New  Zealand  homes  and  cars:  exposure, attitudes, and behaviours in 2004. N Z Med J. 2005 Dec 16;118(1227):U1782.

Glover M et al (2011). Driving kids to smoke? Children’s reported exposure to smoke in cars and early smoking initiation. Addict Behav. 2011 Nov;36(11):1027-31. Epub 2011 Jun 22.

Hongji S et al (2010). Impact of smoking on in-vehicle fine particle exposure during driving. Atmospheric Environment 44(28):3465-4368.

Jones MR et al (2009). Secondhand tobacco smoke concentrations in motor vehicles: a pilot study. Tob Control. 2009 Oct;18(5):399-404. Epub 2009 Aug 25.

Kabir Z et al (2009) Second-hand smoke exposure in cars and respiratory health effects in children. Eur Respir J. 2009 Sep;34(3):629-33. Epub 2009 Apr 8.

Kreuzer M, Krauss M, Kreienbrock L, Jöckel KH, Wichmann HE.  Environmental tobacco smoke and lung cancer: a case-control study inGermany. Am J Epidemiol. 2000 Feb 1;151(3):241-50.

Leatherdale STand Ahmed R. (2009). Second-hand smoke exposure in homes and in cars among Canadian youth: current prevalence, beliefs about exposure, and changes between 2004 and 2006. Cancer Causes Control. 209 Aug;20(6):855-65.

Lee DJ et al (2008). Respiratory effects of secondhand smoke exposure among young adults residing in a “clean” indoor air state. J Community Health. 2008 Jun;33(3):117-25.

Liu S and Zhu Y (2010). A case study of exposure to ultrafine particles from secondhand tobacco smoke   in    an    automobile.   Indoor   Air.    2010    Oct;20(5):412-23.   doi:    10.1111/j.1600-

0668.2010.00665.x. Epub 2010 Jul 16.

MacKenzie R and Freeman B (2010). Second-hand smoke in cars: How did the “23 times more toxic” myth turn into fact?. CMAJ. 2010 May 18;182(8):796-9. Epub 2010 Apr 12.

Marshall L et al (2006). Youth tobacco surveillance–United States, 2001-2002. Surveill Summ.

2006 May 19;55(3):1-56

Martin J, George R, Andrews K, Barr P, Bicknell D, Insull E, Knox C, Liu J, Naqshband M, Romeril K, Wong D, Thomson G, Wilson N. Observed smoking in cars: a method and differences by socioeconomic area. Tob Control. 2006 Oct;15(5):409-11.

Matt GE et al (2008). Residual tobacco smoke pollution in used cars for sale: air, dust, and surfaces. Nicotine Tob Res. 2008 Sep;10(9):1467-75.

Mbulo L (2008). Changes in exposure to secondhand smoke among youth inNebraska, 2002-

2006. rev Chronic Dis. 2008 Jul;5(3):A84.

MuscatJE and Wynder EL (1995). Exposure to environmental tobacco smoke and the risk of heart attack. Int J Epidemiol. 1995 Aug;24(4):715-9

Muramatsu  M,  et  al  (1987).  Estimation  of  personal  exposure  to  ambient  nicotine  in  daily environment. Int Arch Occup Environ Health. 59(6):545-50.

Ott W, Klepeis N, Switzer P. Air change rates of motor vehicles and in-vehicle pollutant concentrations from secondhand smoke. J Expo Sci Environ Epidemiol. 2008 May;18(3):312-25. Epub 2007 Jul 18.

Park JH et al (1998). Measurement of air exchange rate of stationary vehicles and estimation of in-vehicle exposure. J Expo Anal Environ Epidemiol. 1998 Jan-Mar;8(1):65-78.

Pawson R et al (2011). Myths, facts and conditional truths: what is the evidence on the risks associated with smoking in cars carrying children? CMAJ. 2011 Jul 12;183(10):E680-4. Epub

2011 Jan 10.

Rees VW et al (2006). Measuring air quality to protect children from secondhand smoke in cars. Am J Prev Med. 2006 Nov;31(5):363-8.

Roberts LM et al (1996). Children’s exposure to environmental tobacco smoke in private vehicles. Med J Aust. 1996 Sep 16;165(6):350.

Sendzik T et al (2009). An experimental investigation of tobacco smoke pollution in cars. Nicotine

Tob Res. 2009 Jun;11(6):627-34. Epub 2009 Apr 7.

Sly PD et al (2007). Exposure to environmental tobacco smoke in cars increases the risk of persistent wheeze in adolescents. Med J Aust. 2007 Mar 19;186(6):322.

Vardavas  CI  et  al  (2006).  Environmental  tobacco  smoke  exposure  in  motor  vehicles:  a preliminary study. Am J Prev Med. 2006 Nov;31(5):363-8.

Wolfson M, McCoy TP, Sutfin EL.  College students’ exposure to secondhand smoke. Nicotine

Tob Res. 2009 Aug;11(8):977-84. Epub 2009 Jun 10.

Zhou R (2000). Comparison of environmental tobacco smoke concentrations and mutagenicity for several indoor environments. Mutat Res. 2000 Feb 16;465(1-2):191-200.