Environmental Tobacco Smoke Exposure Among Police Officers in Hong Kong

The 1992 US Environmental Protection Agency (EPA) review¹ on passive smoking contains strong and sufficient evidence confirming that exposure to environmental tobacco smoke (ETS), ie, passive smoking, can cause respiratory illnesses in children. However, other than lung cancer, evidence on the relationship between ETS and respiratory ill health in adults is scarce. Three epidemiological studies^{2 - 4} of this relationship were included in the EPA report. One of them studied women,² another studied both men and women,³ and the third⁴ did not explicitly state the sex of the subjects who were student nurses and so were mostly women. Furthermore, apart from the study by Schwartz and Zeger,⁴ which investigated ETS exposure among nursing school students with smoking roommates living in a residential hall, the other 2 studies^{2 - 3} investigated only ETS at home. So far no clear conclusion has been reached about the association between ETS exposure, particularly at work, and chronic respiratory symptoms in adults.¹

We conducted a literature search using the same strategy as in the Report of the UK Scientific Committee on Tobacco and Health⁵ to identify all ETS references related to respiratory symptoms. Because a comprehensive review on ETS before 1990 has been reported in the US EPA report, we searched the MEDLINE and EMBASE databases from January 1991 to October 1998 using the criteria passive or second-hand or second hand or involuntary; smok$ or tobacco$ or cigarette$; and cough or wheeze or breathless or phlegm or mucous where "$" indicates a wild character. Five epidemiological studies of chronic respiratory disease and ETS were found.^{6 - 10} Two of them^{7 - 8} studied only ETS exposure at home in women. The other 3 articles^{6 ,9 - 10} studied ETS exposure at home and/or at work in men and women. White et al⁶ reported that respiratory symptoms were more likely to be found among those with ETS exposure at work. Leuenberger et al⁹ reported a significant trend of increasing risk with total ETS at home and at work, but there was no separate analysis for exposure either at home or work since the exposure data on ETS were not collected separately for home and workplace. Although Jaakkola et al¹⁰ measured ETS exposure both at home and at work separately, only the relationship between total exposure level at home and at work and respiratory symptoms was reported with an increasing, but insignificant, trend for some symptoms. None of these studies have demonstrated a clear dose-response relationship, which is an important criterion for a causal association,¹¹ between ETS exposure at work and respiratory symptoms in those who have never smoked.

In November 1999, the US National Cancer Institute announced the availability of the most comprehensive report¹² on the health risks of ETS exposure ever conducted. However, no definite conclusion of the association between ETS and chronic respiratory symptoms in adults was reached.

The dearth of research on ETS in the workplace and chronic respiratory symptoms prompted this study, which examined the effects of ETS exposure at home and at work among adults in Hong Kong who have never smoked.

This study was conducted within the police force for the personnel department and traffic division of the Hong Kong Police Department following a request from officers for an inquiry into the potential risks of exposure to ambient air pollution. The officers were told that the purpose of the study was to measure aspects of their general health, and this is written in the introductory remark of the questionnaire. All uniformed officers in the traffic, foot patrol, and marine formations of the Hong Kong Police department were eligible for inclusion in the respiratory health survey. The survey was carried out over a 2-month period, among all regional and district traffic teams and foot patrol officers in December 1995 and among Marine police officers in January 1996. Participation was voluntary, there was no requirement for officers to attend the questionnaire sessions or to reveal their identity, and there were no sanctions for nonattendance.

A self-administered structured questionnaire was distributed to participants to complete under classroom conditions. The participants were asked to provide their unique police identity numbers but not their names. The questionnaire was presented in both Chinese and English and covered demographic characteristics, working history, smoking habit, exposure to ETS at home and at work, utilization of health care services, and respiratory health (Table 1).

Respiratory symptoms were elicited using the British Medical Research Council Respiratory Health Questionnaire.¹³ Completed questionnaires were placed into envelopes, sealed, and returned to the research team. The participants were reassured that no one (including their seniors and peers) other than designated researchers in the university department of community medicine would be permitted to gain access to the identity of any participant, and confidentiality was guaranteed.

The following definitions of the dependent variables for respiratory symptoms were used: throat problems, usually having a sore or itchy throat or other throat discomfort; cough or phlegm in the morning, usually having cough or bringing up phlegm first thing in the morning; cough or phlegm during day or night, usually having cough or bringing up phlegm either during the day or at night; chronic cough or phlegm, usually having cough or bringing up phlegm on most days for as much as 3 months each year; any cough or phlegm, usually having cough and/or bringing up phlegm first thing in the morning, or during the day or at night; increased cough and phlegm, usually having a period of increased cough and phlegm lasting 3 weeks or more in the past 3 years; ever wheezing, ever experienced wheezing or whistling in the chest; blocked or running nose, usually having a blocked or running nose; any symptoms, having any of the symptoms defined above.

Utilization of health services was assessed by whether the subject had consulted a Western-trained physician or Chinese traditional physician during the past 14 days for respiratory illness.

A full lifetime smoking history was obtained. A never-smoker was defined as someone who had never smoked any more than 1 cigarette a day or 1 cigar a week or chewed an ounce of tobacco a month for a total accumulation of no more than 6 months. Exposure to ETS at home was defined as the presence of 1 or more smokers who lived in the same household as the study participant. Exposure to ETS at work was defined as the presence of 1 or more coworkers who smoked nearby each day, the number of cigarettes smoked by coworkers nearby per day, the number of hours of exposure per day, and an index of the daily amount of ETS exposure at work, which is defined as the number of cigarettes smoked nearby multiplied by the number of hours exposed to ETS per day, expressed as the number of cigarette-hours. Total exposure to ETS at home and at work is the total number of smokers at home and coworkers smoking nearby.

The analysis was based on those participants who had never smoked and was performed separately for women and men. In comparing the effects of ETS exposure at work on respiratory symptoms, multiple logistic regression was used to calculate odds ratios (ORs) with 95% confidence intervals (CIs), adjusted for age, marital status, educational attainment, police rank, type of police duties (traffic police, foot patrol, or marine police), time on the force, any exposure to a dusty environment in previous jobs, and ETS exposure at home. The analysis was performed using STATA.¹⁴

Of the 11,038 police officers eligible to join the survey, 9923 (90%) completed the questionnaire. Of the respondents, 416 subjects were excluded because of missing entries for sex or smoking status. After exclusion of the missing data, 46.1% of men and 12.3% of women were current smokers, and 2.4% of men and 0.6% of women were former smokers. Only the 5196 (4468 men and 728 women) never-smokers were included in this study. The majority were police constables (72.4%), the lowest rank of officers, followed by sergeants or senior sergeants (22.0%), and inspectors or above (5.5%). Their age ranged from 18 to 58 years.

The demographic characteristics and exposure to ETS at home or work of the never-smoking men and women are shown in Table 1. More women (38.7%) than men (19.7%) were exposed to ETS at home. The same proportion of men and women (80%) reported exposure to ETS at work. More than 70% of both men and women had 2 or more smokers at home and at work, and about half had 3 or more coworkers who smoked nearby.

For total exposure to ETS at home and at work, positive and significant dose-response relationships were found in men between the total number of smokers at home and at work and the risks of respiratory symptoms for all respiratory categories and physician consultation (Table 2). In women the pattern of trends, except for chronic cough, chronic phlegm, ever wheezing, and physician consultation, was similar to that in men.

For ETS at home, the presence of a significant excess of wheezing, nasal problems, and any symptoms was found in men and a significant excess of throat problems was found in women (Table 3). Men exposed at work had a significantly higher prevalence of respiratory symptoms and physician consultation. In women exposed at work, throat problems, morning cough, morning phlegm, nasal problems, and the presence of any symptoms all showed a significant excess compared with women who were not exposed. After adjusting for age, marital status, educational attainment, police rank, type of police officer, time on the force, previous job exposure to dusty environment, and other ETS exposure, the ORs for exposure to ETS at home were significant for wheezing in men and for throat problems in women (Table 3). The adverse effects of ETS exposure at work were more prominent in men than in women; all adjusted ORs were significantly greater than 1 in men. In women, except for ever wheezing, all ORs were greater than 1. However, 5 categories of symptoms were not significantly associated, probably due to the smaller sample size. Of those respiratory symptoms with insignificant ORs, the estimated sample size required for the logistic regression analysis, with a 5% 2-sided significance level and 80% power, ranges from 699 (for nasal problems) to 106,734 (for chronic cough) with a median of 2381 women.¹⁵

Significant trends for all respiratory categories and physician consultation were found in men with smoking coworkers together with positive dose-response relationships between the number of coworkers smoking nearby and the risks of respiratory symptoms (Table 4). The same pattern was observed among all respiratory categories and physician consultation and the number of cigarettes smoked by coworkers nearby per day (results are available on request), the number of hours exposed per day (results are available on request), and the exposure index in terms of cigarette-hours (Table 5).

In women, generally positive trends with the number of smoking coworkers (Table 4) and the exposure index (Table 5) were found. However, a negative but insignificant trend was observed between the number of smoking coworkers and ever wheezing categories, suggesting that women who had a history of wheezing may tend to avoid excessive ETS exposure at work.

This study showed clear associations and dose-response relationships between ETS exposure at work and respiratory symptoms in men and women, with the associations being highly significant in men. Although some significant excess risks were observed for ETS exposure at home, stronger effects were observed at work where the risk of ETS exposure was much greater because subjects spent more time among a larger number of smoking coworkers at work than at home.

A potential weakness of this study is the lack of validation of smoking status and ETS exposure level. Since smoking status was self-reported by the participants, it is possible that smokers might have been misclassified as nonsmokers and included in the analysis. However a meta-analysis of the validity of self-reported smoking suggests generally high levels of sensitivity (87%) and specificity (89%) for self-report.¹⁶ In our study, we compared the declared smoking histories and expired air carbon monoxide levels of a random sample of 110 male officers and found that none of the declared nonsmokers had carbon monoxide levels exceeding 10 ppm, whereas 50 of 55 of the smokers had values greater than 10 ppm. This showed that misclassification was unlikely. For women, Riboli et al¹⁷ estimated that the proportion of women misreporting their active smoking habit was between 1.9% to 3.4% in 13 centers from 10 countries. If we assume a 3.5% misclassification in this study, such misclassification could not explain all the excess risks associated with ETS. In addition, the similarity in the proportion of men and women (80%) in this study reporting exposure to ETS at work further supports reliability of the reporting. Hence, misclassification, if any, would be minimal and cannot account for the strong ETS effects observed in both men and women.

On the other hand, the ORs reported herein may be underestimated due to background exposure. It is likely that some nonsmokers may report no exposure to any form of ETS, even if they are exposed and hence are misclassified as unexposed nonsmokers.

Although no explicit validation of outcome measures for respiratory symptoms and physician consultation was performed, these measures have been commonly used elsewhere^{13 ,18} and have been found to achieve a high level of reliability and validity.

Levels of ETS exposure were not directly measured. Because smoking among police officers in their offices was permitted, the number of smokers was used as the main indicator of ETS exposure. It is likely to be a valid measure of ETS exposure for 2 reasons. First, Cummings et al¹⁹ found that the number of smoking friends and family members seen regularly by their subjects was the single best predictor of cotinine levels, which is a commonly accepted biochemical measure for home, workplace, and social exposure to ETS. Second, a clear dose-response relationship was observed between reports of levels of ETS exposure and respiratory symptoms in our study. The participants reported respiratory symptoms first and ETS exposure much later in the questionnaire. It was unlikely that their reported exposure to ETS was biased by their symptom reporting.

Although ETS has been extensively studied in relationship to lung cancer and coronary heart disease in adults and respiratory illnesses in children, there are few reports from the West and none identified from Asia on the relationship between ETS exposure at work and respiratory symptoms in adults, particularly in never-smoking men. In the West, this may be due to the greater awareness of the problem of ETS and more restrictions on smoking in the workplace. In 1991, 80% of Canadian workers had smoking restrictions in their workplace.²⁰ A 1991 survey of company smoking policies in the United States showed that 85% of firms had smoking policies, and of these, 34% had complete bans and another 34% prohibited smoking in all open work areas.²¹ The Occupational Safety and Health Administration proposed in April 1994 that buildings that allow smoking should provide designated smoking areas in separated and enclosed rooms wherein the air would be exhausted directly to the outside.²² Unfortunately, such legislation is not yet enacted,²³ and the tobacco industry will continue to fight against it by arguing that there is a lack of strong evidence of adverse effects due to ETS at work. However the lack of evidence is probably due to the greater number of smoking restrictions at work in the West than in Asia and hence a lower level of ETS, and so the adverse effects of ETS exposure at work are more difficult to measure.

In Asia, the level of awareness is low and few workplaces have smoking bans or restrictions. For example, in mainland China, 41.5% of men and 19.3% of women were exposed to ETS at work and the prevalence of ETS among government employees was 70.8%,²⁴ which was the highest among all occupations studied.

In the Hong Kong general population aged 15 years and older, 26.7% of men and 3.1% of women were estimated to be smokers in 1996,²⁵ and a telephone survey in 1994 through 1996 found that among the adult working population aged 25 through 74 years, 47.3% (95% CI, 45.6%-49.1%) of men and 26.5% (95% CI, 25.2%-27.7%) of women were exposed to ETS at work.²⁶ Compared with these figures, the prevalence of ETS at work in the sample of police officers participating in our study was much higher, about 80% in women and men, and more than 30% were exposed to more than 3 smoking coworkers nearby. This was a result of both the higher smoking prevalence (46.1% of men; 12.3% of women) in the police force and the lack of restrictions on smoking in the workplace.

The difference in absolute rates of physician consultation for respiratory illnesses in the past 14 days in nonsmoking men between those exposed and not exposed to ETS at work was 4.5% (20.1-15.6), and the attributable risk of ETS exposure at work was 23% ([1.30-1]/1.30). The attributable risk for any respiratory symptoms was 57% ([2.33-1]/2.33) in the nonsmoking men exposed to ETS at work. Nonsmokers exposed to ETS may not realize that their health is affected by their coworkers who smoke nearby. They may accept that smoking at work is the norm and do not attempt to protect themselves from exposure to ETS. In addition to the burden of respiratory illnesses and loss of productivity, public sector health care resources are also consumed, leading to economic losses to the community.

After an internal report was made to the personnel and traffic departments of the force, police administration supported measures promoting a healthful lifestyle within the force aimed at controlling smoking at work and reducing smoking overall. These measures included initiating a trial smoking intervention program for both prevention and cessation in the police training school, offering smoking cessation programs for members of the force, re-enforcing the existing no-smoking policy for officers on duty, and review of the adequacy of regular health assessments for training and occupational health surveillance in the force.

During the past 18 months the senior management has taken steps to create a new nonsmoking culture in the police force. A smoke-free workplace campaign was launched, which included the circulation of documents to all commanders of major regional formations for their views on the development of new regulations and a code of practice on smoking in the workplace. In particular the current regulation that forbids uniformed officers from smoking anywhere in the view of the public was extended to all members of the force, including plainclothes officers. In addition, smoking has been banned in all police buildings and is only permitted in specially designated areas.

Although exposure to ETS at work is involuntary, it is easily preventable. The prevention of smoking in the workplace can significantly and rapidly improve the respiratory health of workers.²⁷ There is a need for a total ban on smoking in the workplace to protect all workers in both developed and developing countries. This study provides evidence of the serious health hazards due to ETS exposure at work to support tobacco control and smoking prevention measures. Delays in the implementation of stronger legislation in the United States and other Western countries can have adverse implications for tobacco control in developing countries in which smoking is still permitted in most workplaces.