OBJECTIVE: No study has ever reported the association between persistent respiratory symptoms and exposure to secondhand smoke (SHS) in adolescent smokers. The impact of SHS exposure on child health could be largely underestimated by not taking into account such effects. We investigated the association between exposure to SHS and respiratory symptoms among adolescent current smokers.
METHODS: A total of 32506 students aged 11 to 20 years from 85 randomly selected secondary schools in Hong Kong completed a self-administered questionnaire that included persistent respiratory symptoms (for 3 consecutive months in the past 12 months), number of days of SHS exposure per week at home and outside home, smoking status, amount of active smoking, and other basic demographic characteristics and socioeconomic status.
RESULTS: Adolescent current smokers who were exposed to SHS at home 1 to 4 and 5 to 7 days/wk were 50% (95% confidence interval [CI]: 3%–121%) and 77% (95% CI: 5%–199%) more likely, respectively, to report respiratory symptoms compared with those who were unexposed (P = .01 for trend). The corresponding figures for exposure outside home were 41% (95% CI: 3%–94%) and 85% (95% CI: 31%–161%; P = .004 for trend). Such associations were also observed among never-smokers, but they were weaker than those among current smokers (P < .01 for interaction).
CONCLUSIONS: This is the first evidence that SHS exposure is associated with increased risks for persistent respiratory symptoms among adolescent current smokers. Health promotion programs should aim at SHS reduction as well as smoking cessation among adolescent smokers.
It is well established that secondhand smoke (SHS) exposure causes premature deaths and serious diseases in adults.1 Almost all of the evidence is based on nonsmokers so as to eliminate the effects of active smoking. The harm of SHS is deemed trivial among active smokers, and any positive findings would be seriously doubted even though SHS exposure is often more common in smokers than in nonsmokers. If SHS can harm smokers, then tobacco control measures should include warning smokers of the risks from SHS and protecting them from SHS exposures.
In 2005, we reported the first dose-response evidence of respiratory ill health in adult current smokers as a result of SHS exposure at work from 1 to ≥4 smokers with odds ratios from 1.53 to 1.96.2 Before this, we had shown similar dose-response relations among nonsmokers in the same study.3 Our PubMed search from January 1960 to February 2009 using the criteria “passive or second-hand or involuntary” and “smok$ or tobacco$ or cigarette$” and “active smok$ or current smok$” and “adolescen$ or youth$ or teenage$ or juvenescen$” ($ indicates ≥1 character) yielded no report of the effects of SHS on adolescent current smokers. The pattern and intensity of smoking and SHS exposure probably differ between adults and adolescents; the effects of SHS exposure may also be different in adults and in adolescents. Serious smoking-related diseases in late adulthood are probably not as relevant to adolescent smokers as the much more common respiratory symptoms. We therefore investigated the association between SHS exposure and respiratory symptoms among adolescent smokers in Hong Kong.
During February 2003 to April 2004, a youth smoking survey that used a self-administered questionnaire was conducted in Hong Kong with high response rate (98%) from students in 85 secondary schools that were randomly selected with a probability proportional to school enrollment size pursuant to the requirement of the Global Youth Tobacco Survey.4 All classes in form 1 (equivalent to US grade 6) and 2 random classes in forms 2 to 5 (grades 7–10) of the participating schools (in total 1012 classes) were surveyed.
Students' smoking status was classified as never-smokers, experimenters (smoked once or a few times), ex-smokers (smoked in the past but not now), <1 cigarette per day, 1 to 6 cigarettes per day, and ≥6 cigarettes per day.5 Current smokers were defined as having smoked on ≥1 day in the past 30 days.4 Students reported the number of days in the past 7 days that they were exposed to SHS at home and outside home in 2 separate questions. Other questions related to smoking habits and perceptions toward SHS exposure were asked. Respiratory symptoms were defined as persistent cough or sputum for 3 consecutive months during the past 12 months, as had been used in previous studies of smoking, passive smoking, and respiratory symptoms in secondary school students.5 Basic demographic characteristics and socioeconomic status were also recorded.
After exclusion of the poorly answered questionnaires, such as those with response sets and excessive missing answers, 36612 remained (98% of the original sample). Of these, 32506 (89%) students aged 11 to 20 years were analyzed after successive exclusion of those who were older than 20 (n = 127) and missing information on respiratory symptoms (n = 210), SHS exposure (n = 397), smoking status (n = 211), socioeconomic information (n = 3124), and gender (n = 37).
Logistic regression that was based on generalized estimating equations to control for cluster effect within the same school was used to calculate adjusted odds ratios for respiratory symptoms as a result of SHS exposure. Exposures at home and outside home were analyzed as separate variables. For the analysis of home exposure, students who were exposed to SHS outside home ≥2 days/wk were excluded to reduce the effects of exposure outside home, and any residual effects were controlled by adjustments. Similarly, for the analysis of outside home exposure, students who were exposed to SHS at home ≥2 days/wk were excluded. We did not exclude students with SHS exposure 1 day/wk at the other site for sample size considerations and because the effect, if any, would likely be small.
Other covariate adjustments included gender, age (in years), highest parental education, and type of housing. For the analyses among current smokers, which were our main focus, the cumulative amount of smoking calculated by the number of years of smoking times the number of cigarettes smoked per day was also adjusted for. To examine the consistency of findings, we also repeated the analyses among never-smokers. The potential moderating effect of smoking status (current versus never) on the association between respiratory symptoms and SHS exposure (any versus none, at home, and outside home) was assessed. Ethical approval for the study was granted by the institutional review board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster.
Of all participants, 47% were boys and 56% were aged ≤15. One (24%) in 4 had ever smoked, including current smokers (9%), ex-smokers (2%), and experimenters (13%). Most current smokers reported exposure to SHS at home (51%) and outside home (85%) and realized the harmful effects of SHS (85%). Among the older current smokers aged ≥16 (n = 1394), the starting age of smoking monthly was mainly 14 to 15 years (37%), followed by ≥16 (25%), 12 to 13 (21%), 10 to 11 (7%), 8 to 9 (6%), and ≤7 (4%). A substantial proportion of all current smokers purchased cigarettes by themselves directly from shops or street kiosks (48%), smoked 2 to 5 cigarettes per day (33%), smoked in public places such as parks or streets (47%), and were living with ≥2 smokers (32%). Males and females aged 15 to 19 had daily smoking rates of 3% and 2%, respectively, which were very similar to those from government household survey on the general population,6 confirming that our sample should be representative of the general population (Cohen effect size7 w = 0.01).
Table 1 shows that higher prevalence of respiratory symptoms was associated with being male, lower parental education, and living in temporary housing. Students who tried or used to smoke but not now, smoked <1 cigarette per week, and smoked ≥1 cigarette per week had excess risk (ER) and of respiratory symptoms of 23% (95% confidence interval [CI]: 13%–35%), 77% (95% CI: 52%–107%), and 207% (95% CI: 167%–253%), respectively, compared with those who never smoked (P < .001 for trend).
Table 2 shows that current smokers who were exposed to SHS at home for 1 to 4 and 5 to 7 days/wk had ER of respiratory symptoms of 50% (95% CI: 3%–121%) and 77% (95% CI: 5%–199%), respectively, compared with those who were unexposed (P = .01 for trend). The corresponding figures for SHS exposure outside home were 41% (95% CI: 3%–94%) and 85% (95% CI: 31%–161%; P = .004 for trend). ERs were also observed among never-smokers who were exposed to SHS at home and outside home, although the effect sizes were significantly smaller (P = .004 for interaction) than those among current smokers.
This study has provided the first evidence that SHS exposure is associated with respiratory symptoms with a dose-response relationship among adolescent smokers. This association is consistent with that found among adult smokers in our Police Health Study2 and those in numerous studies of nonsmoking adults1–3, 8 and children.1, 5, 8, 9 The strength of the association found (Table 1) is far from trivial and comparable to that in a study of secondary school students in Hong Kong.5 Positive association between SHS exposure and respiratory symptoms was also observed among never-smokers in this study, lending support that the findings were internally coherent.
We found that the association between SHS exposure and respiratory symptoms was stronger in adolescent smokers than never-smokers, a significant interaction (P < .01). The association remained stronger in smokers than never-smokers after additional adjustments for number of smokers at home and the proportion of friends who smoke (P < .02 for interaction) or the acceptability of SHS exposure (P < .01 for interaction). One explanation is that adolescent smokers are more susceptible to SHS than never-smokers, but this is uncertain because the same underlying difference, if present, would have prevented them from smoking in the first place. Given the reported same number of days of exposure per week, the same number of home smokers and smoking friends, and the same acceptability of SHS exposure, a more plausible explanation is that despite the aforementioned similarities, adolescent smokers were actually exposed to more SHS than never-smokers because the former were more likely to have stayed longer and closer to other smokers, whereas the latter could have moved away to reduce exposure. The longer duration of SHS exposure at home or outside home among smokers than never-smokers could also result in higher cumulative exposures because of the prolonged settling time of fine particles (PM1 and PM2.5), which represent most of the SHS particles.10 Additional investigations are needed, but cotinine assays on body fluids cannot separate exposure from active smoking and SHS exposure among active smokers, in adolescents and in adults.
The effect of SHS exposure on respiratory symptoms would depend on the level of exposure among those exposed. We found substantial ERs even in Hong Kong, where smoking prevalence was relatively low11 and partial smoking bans in public areas were already in place at the time of the study. In populations with high smoking prevalence and weak tobacco control measures, the risks should be higher; however, smoke-free measures might inevitably cluster smokers in specified areas or designated smoking rooms, potentially exposing them to higher levels of SHS. The effects of smoking bans on SHS exposure of smokers, especially adolescent smokers, therefore warrant additional investigations, and this highlights the importance of our findings as many countries, especially China, step up their tobacco control measures under the World Health Organization Framework Convention on Tobacco Control.
Smokers need to know that SHS exposure is harmful even to themselves and other smokers, and the argument that smokers can be allowed to work in smoking rooms (because they are not additionally harmed by SHS) is not acceptable. Note that the health risks of SHS exposures could be equivalent to that for mild active smoking (1 or 2 cigarettes per day).12
Our study has several limitations. First, all data were based on self-report, including smoking, SHS exposure, and respiratory symptoms, because objective measurements or validation was impractical given the size of the study. Although smoking is a sensitive issue among adolescents, the anonymous nature of the survey should have provided sufficient assurance about confidentiality. The observed associations, as expected and frequently reported, of smoking and SHS exposure with socioeconomic variables and respiratory symptoms indirectly supported the validity of the data. Second, temporality cannot be ascertained because of the cross-sectional nature of the survey; however, the notion of reverse causation that students with respiratory symptoms deliberately increased their noxious exposure to SHS seems improbable. If anything, students might avoid SHS exposure to ameliorate their respiratory symptoms, thereby diluting the associations. Finally, few people would be completely unexposed to SHS in densely populated Hong Kong even now (smoking in most indoor public and work places has been banned in Hong Kong since January 1, 2007), not to mention at the time of the study, when smoking was still allowed in public places such as restaurants; therefore, the reference groups who reported no SHS exposure have probably underestimated their exposure, and the risk for respiratory symptoms on the basis of these groups would also be underestimated.2, 3, 13
Our findings would promote tobacco control measures by alerting clinicians and public health professionals that the SHS effects on current smokers should not be ignored. Clinicians should advise smoking adolescent patients to avoid SHS exposure as well as to quit smoking and should remind the patients that SHS is even more harmful to smokers than to nonsmokers, especially adolescents. Tobacco control advocators can also make use of our evidence to wrestle with the smoke-free exemption policies and the arguments from opponents of tobacco control, for example, allowance of smoking rooms inside nonsmoking indoor area, approval for smoking bars and restaurants after smoke-free legislation is in force, and allowing smokers to work in smoking areas or smoking rooms.
Increased SHS exposure is associated with increased respiratory symptoms in adolescent current smokers. Adolescent smokers are subjected to the adverse health effects of both active and passive smoking. The impact of SHS exposure on public health would be underestimated by not taking into account the effects of SHS on current smokers. Health promotion programs should aim at SHS reduction as well as smoking cessation among adolescent smokers. Alerting active smokers about the harmful effects of SHS from others on themselves may motivate more smokers to support banning of smoking in public and other places. Our article provides not just important clinical and public health implications but also a keystone on the path toward a complete ban of smoking in our living environment for the benefits of the next generations.
This study was supported by the Hong Kong Council on Smoking and Health and the Department of Health.
Thanks to M.K. Lai for project management; L.M. Ho, PhD, for statistical advice; and Hong Kong Council on Smoking and Health and the Department of Health (Government of Hong Kong S.A.R.) for funding this study.
- Accepted June 4, 2009.
- Address correspondence to Sai-Yin Ho, PhD, School of Public Health, 5/F WMWM Blk, Faculty of Medicine Building, 21 Sassoon Rd, Hong Kong. E-mail:
Financial Disclosure: The authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject:
SHS causes premature deaths and diseases among nonsmokers.
What This Study Adds:
Increased SHS exposure is associated with increased respiratory symptoms in adolescent current smokers. SHS effects on respiratory symptoms are much larger on current adolescent smokers than on never-smokers.
- ↵US Department of Health and Human Services. The Health Consequences of Involuntary Exposure to Tobacco Smoke: A Report of the Surgeon General. Atlanta, GA: US Department of Health and Human Services, Centers for Disease Control and Prevention, Coordinating Center for Health Promotion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health; 2006
- ↵Lam TH, Ho LM, Hedley AJ, et al. Secondhand smoke and respiratory ill health in current smokers. Tob Control.2005;14 (5):307– 314
- ↵The Global Youth Tobacco Survey Collaborative Group (GYTSCG). Tobacco use among youth: a cross country comparison. Tob Control.2002;11 (3):252– 270
- ↵Lam TH, Chung SF, Betson CL, Wong CM, Hedley AJ. Respiratory symptoms due to active and passive smoking in junior secondary school students in Hong Kong. Int J Epidemiol.1998;27 (1):41– 48
- ↵Census and Statistics Department. Thematic Household Survey Report No. 26. Social Surveys Section, Census and Statistics Department; 2006. Available at: www.censtatd.gov.hk/products_and_services/products/publications/statistical_report/social_data/index_cd_B1130226_dt_detail.jsp. Accessed September 17, 2009
- ↵Cohen J. Statistical Power Analysis for the Behavioral Sciences. Rev ed. New York, NY: Academic Press; 1977
- ↵Lam TH, Hedley AJ, Chung SF, Macfarlane DJ, Child Health and Activity Research Group. Passive smoking and respiratory symptoms in primary school children in Hong Kong. Hum Exp Toxicol.1999;18 (4):218– 223
- ↵World Health Organization. Country Profiles: Tobacco or Health. Tobacco-Free Initiative, Western Pacific Region2000. Available at: www.wpro.who.int/internet/resources.ashx/TFI/country+profiles+2000.pdf. Accessed September 17, 2009
- ↵Pechacek TF, Babb S. How acute and reversible are the cardiovascular risks of secondhand smoke? BMJ.2004;328 (7446):980– 983
- Copyright © 2009 by the American Academy of Pediatrics