Flowchart of participants in the Swiss Cohort Study on Air Pollution and Lung Diseases in Adults (SAPALDIA) (2002) who are included in the present study. SF-36 indicates 36-Item Short Form Health Survey; SHS, secondhand smoke.
36-Item Short Form Health Survey mean score differences in men and women associated with moderate and high secondhand smoke (SHS) exposure, adjusted for age, age squared, education level, professional category, nationality, comorbid conditions, categorical body mass index, categorical forced expiratory volume in 1 second (percentage of predicted), and study area. BP indicates bodily pain; GH, general health; MH, mental health; PF, physical functioning; RE, role emotional; RP, role physical; SF, social functioning; and VT, vitality. *P = .048. †P = .01. ‡P = .03. §P = .02. For all P values, the group of subjects not exposed to secondhand smoke was the reference group.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Bridevaux P, Cornuz J, Gaspoz J, et al. Secondhand Smoke and Health-Related Quality of Life in Never Smokers: Results From the SAPALDIA Cohort Study 2. Arch Intern Med. 2007;167(22):2516–2523. doi:10.1001/archinte.167.22.2516
Although secondhand smoke (SHS) has been linked with various respiratory conditions and symptoms, its association with health-related quality of life (HRQOL) is unknown.
A cross-sectional study was performed of 2500 never smokers in Switzerland who participated in the Swiss Cohort Study on Air Pollution and Lung Diseases in Adults and completed a 36-Item Short Form Health Survey (SF-36) in 2002. Using linear regression models adjusting for confounders, we measured the association between HRQOL and moderate or high SHS exposure (≤3 h/d or >3 h/d) compared with no SHS exposure. Data from men and women were analyzed separately and further stratified by source of SHS (home, workplace, and public spaces).
After adjustments, SHS was associated with reduced scores in all SF-36 domains. High SHS exposure predicted a greater reduction in HRQOL. Compared with nonexposed women, those with high SHS exposure at home had significantly lower scores on the physical functioning (−7.8, P < .001), role physical (−10.5, P = .02), bodily pain (−9.2, P = .01), and social functioning (−8.1, P = .007) domains. Exposed men had lower scores for the role physical domain (−20.0, P < .001) and a trend toward lower scores in other domains. In women, exposure to SHS at home was associated with a stronger negative effect on HRQOL than at work and in public spaces.
Secondhand smoke is associated with reduced HRQOL, more significantly so in women. Exposure to SHS at home and high levels of exposure are associated with lower SF-36 scores, suggesting a dose-response relationship.
Secondhand smoke (SHS) exposure is widespread and affects between 6% and 65% of the population in the European Community Respiratory Health Survey 2.1 In Switzerland, 25% of the adult population reported exposure to SHS in 2004. Secondhand smoke has been recognized as a risk factor for lung cancer,2 coronary heart disease,3 chronic respiratory symptoms,4,5 and low pulmonary function.6,7 Cross-sectional and longitudinal studies reported an association between SHS and asthma.4,8
Health-related quality of life (HRQOL) has been shown to be reduced among current smokers, with a greater reduction in physical and mental health domains.9,10 In a case-control study, even a short history of smoking was associated with a decrease in HRQOL scores in young smokers, which suggests that HRQOL decline may precede changes in pulmonary function tests.11
Overall, HRQOL is considered an important tool to comprehensively assess physical, social, and mental functioning of a population. For example, in a large cohort of veterans, decline in the physical component summary of the 36-Item Short Form Health Survey (SF-36) increased the risk of death and of hospitalization.12 This relationship was confirmed, independently of prevalent cardiovascular disease or cancer, in a large survey from the general population in Europe.13 Scores on the SF-36 also predicted total health care costs.14 Hence, even a minor reduction in HRQOL due to SHS may represent an early marker of a preventable increase in mortality, hospitalization, and health care costs at population levels. However, despite the known health effects of SHS among never smokers, the relationship between SHS and quality of life has not been delineated so far.
The goals of our study were (1) to investigate the association between SHS and HRQOL among never smokers, as evaluated by the SF-36, using a population-based survey: the Swiss Cohort Study on Air Pollution and Lung Diseases in Adults (SAPALDIA) and (2) to assess the association of various sources of SHS exposure with HRQOL.
The SAPALDIA cohort (N = 9651) was enrolled in 1991 and consisted of a random sample of the population aged 18 to 60 years across 8 study areas chosen to represent the diverse climate, degrees of urbanization, and levels of air pollution in Switzerland (Geneva, Basel, Lugano, Aarau, Wald, Payerne, Davos, and Montana).15 In 2002, the follow-up study included 8047 subjects (83.4%) from the 1991 sample. As part of this study, 5695 subjects (70.8%) returned the SF-36. During a person-to-person examination, trained interviewers administered an extensive health inventory questionnaire addressing the subject's history of respiratory symptoms, allergic diseases, living and working environment, smoking habits, socioeconomic status, and comorbidities as well as exposure to SHS. Because the effect of active smoking on HRQOL is considerable in current and former smokers, subtle additional changes due to SHS among those groups may be overlooked. Thus, to avoid confounding by smoking, we based our study sample on 2500 adults who answered the SF-36 and had never smoked.
The primary predictor for this study was exposure to SHS during the 12-month period before the interview. Subjects were considered to have been exposed to SHS if they answered yes to the question, “Have you been regularly exposed to tobacco smoke in the last year?” Exposure to SHS was further detailed by collecting data on its sources (“Have you been regularly exposed to tobacco smoke at home during the last 12 months? [yes/no] At your workplace? [yes/no] In bars or restaurants? [yes/no]”), as well as the duration of exposures in hours per day (“During how many hours per day are you exposed to tobacco smoke at home? At your workplace? In bars or restaurants? In total?”). The duration of SHS exposure was categorized in the following groups: 0 h/d (reference group), 3 h/d or less (moderate exposure), and more than 3 h/d (high exposure) in accordance with a previous study.4 Few subjects reported exposure of more than 3 h/d in bars or restaurants; they were therefore included in the group of subjects reporting moderate exposure. Mean daily SHS exposure was also used as a continuous variable for each different source and for total exposure.
The subjects filled out SF-36 questionnaires between April 2001 and March 2003. Because Switzerland is a multilingual country, German, French, and Italian validated versions of the questionnaire were administered.
The outcomes of interest were the 8 domains of the SF-36 (physical functioning, role physical, bodily pain, general health, vitality, social functioning, role emotional, and mental health). Each domain explores a different aspect of HRQOL. In particular, the physical functioning domain is related to the ability to perform physical activities, including strenuous effort. The role physical measures the extent of disability, such as problems with work, due to physical problems. The aspects of social life are addressed by the social functioning scale, which examines the ability to perform social activities given the potential interference of physical or emotional problems.
Age, sex, nationality, education level, professional status, comorbid conditions, and physician-diagnosed asthma were derived from questionnaires. Body height and weight were measured with participants not wearing any shoes or coats. Pulmonary function tests were performed by trained pulmonary technicians, according to the American Thoracic Society standards.
To assess for potential response bias, characteristics of subjects who did and did not answer the SF-36 were compared. The SHS-exposed and nonexposed subjects were compared with unpaired t test and χ2 test when appropriate. Then, univariate and multivariate linear regression models were constructed using the 8 domains of the SF-36 as dependent variables and SHS exposure categories as the main independent variables. Ninety-two questionnaires (3.7%) had 1 or more missing answers and were excluded from the analysis.
Age was found to have a nonlinear effect and thus was entered into the models with a quadratic term. Low education, non-Swiss citizenship, low professional status, diabetes mellitus, body mass index (BMI) (calculated as weight in kilograms divided by height in meters squared), and study area were found to be predictors of both SHS and HRQOL. Percentage of predicted forced expiratory volume in 1 second (FEV1), as well as self-reported comorbid conditions (any chronic disease, congestive heart failure, hypertension, diabetes, osteoarthritis, stroke, and depression), were associated with HRQOL scores and controlled for. The study area was included as a random-effects variable in our multivariate models.
Further confounding was also assessed by introducing supplementary covariates into the model. Asthma, season of the year (questionnaire completion date), and occupational exposure were tested as variables. These factors did not modify the association between SHS and the different quality-of-life domains and were not included.
An analysis using an interaction term (sex × SHS) showed an effect modification. Thus, subsequent models were stratified on sex.
Finally, to isolate a potentially more pronounced effect of one or another source of SHS exposure (at home, workplace, or in bars/restaurant), analysis stratifying on SHS sources was performed.
To better anchor the clinical significance of SHS exposure and HRQOL in our population, odds ratios of reporting regular outpatient visits during the 12-month period preceding the survey were calculated, adjusted for the aforementioned covariates.
Of the 5695 SF-36 responders, 2500 (43.9%) had never smoked and were eligible for the analyses. Among those, 388 subjects (15.5%) reported regular exposure to SHS (Figure 1).
Comparison of subjects who completed the SF-36 (n = 5695) and those who did not (n = 2352) showed that responders were younger (51.8 vs 52.8 years; P = .004) and more likely to be male (2787 [48.9%] vs 1080 [45.9%]; P = .01), Swiss citizens (4988 [87.6%] vs 1865 [79.3%]; P < .001), and better educated (low education level, 774 [13.6% vs 410 [17.4%]; P < .001). Asthma diagnosed by a physician was reported less frequently among responders (411 [7.2%] vs 204 [8.7%]; P < .001), but no difference in the FEV1 to forced vital capacity ratio was observed. Responders were less likely to be current smokers 1384 [24.3%] vs 620 [26.4%]; P < .001), but SHS exposure among never smokers was similar (388 of 2500 [15.5%] vs 105 of 744 [14.1%]; P = .35).
Table 1 compares the characteristics of subjects with (n = 388) and without (n = 2112) reported exposure to SHS during the last 12 months. Subjects exposed to SHS were generally younger, had a higher BMI, were non–Swiss citizens, were less educated, and had a lower professional status. They more frequently reported exposure to dust, gas, or fumes. In addition, comorbid conditions were slightly more frequent and reached statistical significance for diabetes.
The source and duration of SHS exposure by sex are described in Table 2. Although the total duration of SHS exposure was similar in women and men, there were some sex-related differences. Women were more exposed at home than men, whose exposure mainly took place at work. Mean SHS exposure time was 2 hours 54 min/d, independent of the source. Self-reported total exposure was lower than the sum of home, workplace, and bar/restaurant exposure. This was not unexpected because each question had been asked separately.
Table 3 provides the age-adjusted SF-36 scores by sex and SHS exposure for never smokers. In men, significantly lower HRQOL scores were present on the role physical domain, whereas a trend toward lower scores was present on all 7 other domains. In women, significantly lower scores were present on all domains except bodily pain.
Adjusted differences in SF-36 scores comparing subjects with moderate or high SHS exposure with the nonexposed are detailed by sources of exposure and sex in Table 4. For women, SHS exposure at home was generally associated with lower HRQOL. Women with home exposure of more than 3 h/d were particularly affected. In contrast, SHS exposure in the workplace among women was associated with significantly lower HRQOL scores only for the physical functioning score. For SHS exposure in bars and restaurants, women reported a lower HRQOL than the nonexposed subjects on all domains, with statistical significance reached for physical functioning, vitality, and social functioning.
Multivariate regression models using the duration of SHS (in hours per day) as a continuous variable showed that, in women, each 1-hour increase in home SHS exposure resulted in a decrease in the domains of physical functioning (−1.16 points/h; P = .02), role physical (−1.52 points/h; P = .03), bodily pain (−1.54 points/h; P = .07), social functioning (−1.40 points/h; P = .03), and mental health (−0.80 points/h; P = .02). The relationship was similar but less marked and not significant for exposure in the workplace and in bars or restaurants. For men, although a similar negative effect of SHS was observed on almost all subscales, only total duration of home SHS was significantly correlated with role physical (−5.29 points/h; P < .001).
Mean adjusted differences in SF-36 scores associated with moderate and high SHS are shown in Figure 2 for men and women separately. Women with moderate SHS exposure had a significantly lower HRQOL on physical functioning and vitality domains, whereas among those with high SHS exposure, the association was larger and significant on physical functioning and social functioning domains.
The potential role of asthma as an effect modifier of the association of SHS with HRQOL was explored for men and women separately. For men with a diagnosis of asthma (77 of 1011 [7.6%]), exposure to SHS was associated with a trend toward more physical limitations. For women, asthma (96 of 1488 [6.5%]) did not interact with SHS to lower HRQOL. Thus, an effect modification of SHS by asthma could be suspected but not confirmed, perhaps because of the small number of participants with asthma. In addition, testing the interaction term asthma × SHS proved not to be significant in either men (P = .17) or women (P = .10). Similarly, FEV1 did not show a statistical interaction with SHS exposure (P = .22 for men and P = .59 for women).
Table 5 shows the adjusted odds ratios of reporting regular physician visits among never smokers stratified by level of SHS exposure and physical functioning scores. In 3 nested models adjusting for only demographic variables (model 1), socioeconomic variables (model 2), and comorbidities, BMI, and lung function tests (model 3), subjects exposed to SHS were more likely to report regular outpatient visits. The relationship was significant and stronger for those exposed more than 3 h/d in models 1 and 2. Also, more frequent physician visits were reported in subjects with lower HRQOL, compared with those with physical functioning scores greater than 90 points.
Our study indicates that SHS exposure is associated with lower HRQOL scores in most SF-36 domains for women. In the physical domains, there was a dose-response relationship, with higher daily SHS exposure resulting in greater reduction of HRQOL, the association being stronger when women were exposed at home. The observed reduction in HRQOL was large enough to be perceived.
Other studies have reported sex differences in the effects of SHS and smoking. Langhammer et al16 have shown that exposure to SHS during childhood led to a higher risk of wheezing or breathlessness in women but not in men. In the same study, more respiratory symptoms and lower self-rated health were reported by women, after controlling for smoking status and lung function. Cross-sectional studies found lower lung function among women exposed to smoking, as well as a lower cough threshold and higher smoking-related bronchial hyperresponsiveness.17-20
The association between SHS and bodily pain in women is intriguing. It may be due to a true relationship mediated by conditions that eventually cause pain. It could be confounded by some socioeconomic factors (eg, housekeepers carrying heavy loads) that are related to SHS and bodily pain, or it may be due to chance alone. Further studies are needed to elucidate this association.
Sources of SHS exposure were an important determinant of HRQOL in our study. In an earlier report, Leuenberger et al4 found a particularly strong effect of SHS exposure on respiratory symptoms at work, although this exposure has been drastically reduced ever since because of smoking bans in many workplaces. Recent studies have shown that home SHS exposure was the strongest predictor of serum or ambient cotinine level compared with workplace or social exposure.21,22 The negative effect of SHS exposure at home is likely to be caused by a higher concentration of side-stream smoke due to the relatively small size of private houses or apartments, and possibly by a greater number of cigarettes smoked in a close environment.23 This hypothesis is supported by our data showing that the negative effect on HRQOL of 1 hour of SHS at home is greater than the same duration of exposure in the workplace or in bars/restaurants. This fact may explain the larger HRQOL score differences seen in women, mainly exposed at home, compared with men.
Previous epidemiologic studies have shown an association between the risk of having a diagnosis of asthma and SHS exposure, especially in women.4,8 Lung function of patients with asthma was worse when exposed to SHS in one study.7 In a recent report, our group24 was able to show that silent bronchial hyperresponsiveness is a risk factor for the development of respiratory symptoms and lung function decline in otherwise healthy individuals persistently exposed to SHS.
In this study, few subjects reported having asthma, and they were apparently not affected by SHS in terms of HRQOL. One explanation could be that asthmatic subjects, in particular those who are more sensitive to SHS, try to avoid SHS exposure and thus paradoxically report an equal HRQOL.
The cross-sectional design of our study cannot exclude the effects of unmeasured confounders. For instance, cardiovascular disease and cancer have both been associated with SHS and are important predictors of SF-36 scores.2,3 However, the prevalence of these 2 conditions is low among nonsmokers, and particularly women, who experienced the larger reduction in HRQOL associated with SHS. In our multivariate models, self-reported congestive heart failure, hypertension, and diabetes mellitus were controlled for and only slightly changed the coefficients of SHS exposure. We cannot exclude the effects of undiagnosed chronic obstructive pulmonary disease or asthma interacting with SHS to reduce HRQOL. Nevertheless, such an effect is unlikely to have been of significant magnitude in our study given that FEV1 was adjusted for in the multivariate models. Finally, socioeconomic differences not entirely captured by our control for study area, level of education, professional category, and citizenship may have played a role.
Our study has several limitations. First, responders to the SF-36 questionnaires had a lower BMI and a higher socioeconomic status than did nonresponders and thus may not be representative of the whole population of subjects exposed to SHS, who have a lower socioeconomic status and a higher BMI. However, such a selection bias would lead to an underestimation of the true effect of SHS on HRQOL. Second, SHS exposure was assessed by questionnaires, which could lead to misclassification of true exposure. In particular, self-reported duration of exposure is subject to recall errors and may mirror the perceived burden of SHS more than actual duration. Nonetheless, previous studies found a linear correlation between self-reported SHS exposure and biomarkers such as salivary or serum cotinine, indicating that questionnaires about reported SHS exposure correctly classify the relative levels of SHS exposure among nonsmokers.21,25 Third, it is possible that a tobacco- or respiratory disease–specific questionnaire would be more sensitive to detect the effects of SHS exposure than the SF-36 generic instrument applied in this study.
Strengths of our study include the size of the sample and its random selection from the general population of Switzerland. The multicultural and geographic diversity of the sample reinforces the external validity of our results. Furthermore, the nonspecific interview-based questionnaire, centered on a variety of potential environmental hazards and not only on SHS, reduces the risk of a recall bias. Finally, in our sample of never smokers, lower HRQOL and SHS were both associated with health care utilization, which suggests that the health effects of SHS are clinically meaningful.
In conclusion, results of our study suggest that SHS is associated with a significant decrease of HRQOL scores in women, particularly in the physical health domain, as measured by the SF-36. This observed association seems dose dependent, notably in women exposed to SHS at home. These findings may contribute to prompting future public health policy to focus on counseling about SHS at home and not only on controlling SHS in public places.
Correspondence: Margaret W. Gerbase, MD, PhD, Division of Pulmonary Medicine, University Hospitals of Geneva, 24 rue Micheli-du-Crest, 1211 Geneva, Switzerland (firstname.lastname@example.org).
Accepted for Publication: August 18, 2007.
Author Contributions: Drs Bridevaux and Gerbase had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Bridevaux, Cornuz, Gaspoz, Ackermann-Liebrich, Leuenberger, Rochat, and Gerbase. Acquisition of data: Ackermann-Liebrich, Schindler, Leuenberger, and Gerbase. Analysis and interpretation of data: Bridevaux, Cornuz, Gaspoz, Burnand, Schindler, Rochat, and Gerbase. Drafting of the manuscript: Bridevaux and Cornuz. Critical revision of the manuscript for important intellectual content: Cornuz, Gaspoz, Burnand, Ackermann-Liebrich, Schindler, Rochat, and Gerbase. Statistical analysis: Bridevaux, Gaspoz, and Schindler. Obtained funding: Gaspoz, Ackermann-Liebrich, Leuenberger, Rochat, and Gerbase. Administrative, technical, and material support: Gaspoz, Ackermann-Liebrich, Leuenberger, and Rochat. Study supervision: Cornuz, Burnand, Ackermann-Liebrich, Schindler, Leuenberger, Rochat, and Gerbase.
Financial Disclosure: None reported.
Funding/Support: This study was supported by the Swiss National Science Foundation (grants 4026-28099, 3347CO-108796, 3247BO-104283, 3247BO-104288, 3247BO-104284, 32-65896.01, 32-59302.99, 3252720.97, and 32-4253.94); the Federal Office for Forest, Environment and Landscape; the Federal Office of Public Health; the Federal Office of Roads and Transport; the canton's government of Aargau, Basel-Stadt, Basel-Land, Geneva, Luzern, Ticino, and Zurich; the Swiss Lung League; and the canton's Lung League of Basel Stadt/Basel Landschaft, Geneva, Ticino, and Zurich.
Additional Information: SAPALDIA Basel is part of the European Community Respiratory Health Survey.
Additional Contributions: We are indebted to the study participants, the technical and administrative support at coordinating centers, and the medical teams and fieldworkers at the local study sites. This work would not have been possible without their enthusiastic commitment.
Thierry Rochat, MD (pneumology), Ursula Ackermann-Liebrich, MD (epidemiology), Jean-Michel Gaspoz, MD, MSc (cardiology), Philippe Leuenberger, MD (pneumology), L. J. Sally Liu, ScD (exposure), Nicole M. Probst Hensch, MPH, PhD (epidemiology/genetic and molecular biology), and Christian Schindler, PhD (statistics).
Jean-Claude Barthélémy, MD (cardiology), Werner Berger, MD (genetic and molecular biology), Robert Bettschart, MD (pneumology), Andreas Bircher, MD (allergology), Otto Brändli, MD (pneumology), Martin Brutsche, MD, PhD (pneumology), Luc Burdet, MD (pneumology), Martin Frey, MD (pneumology), Margaret W. Gerbase, MD, PhD (pneumology), Diane Gold, MD, DTM&H (epidemiology/cardiology/pneumology), Werner Karrer, MD (pneumology), Roland Keller, MD (pneumology), Bruno Knöpfli, MD (pneumology), Nino Künzli, MD, MPH, PhD (epidemiology/exposure), Urs Neu, PhD (exposure), Laurent Nicod, MD (pneumology), Marco Pons, MD (pneumology), Eric Russi, MD (pneumology), Peter Schmid-Grendelmeyer, MD (allergology), Joël Schwartz, PhD (epidemiology), Peter Straehl, MPH, PhD (exposure), Jean-Marie Tschopp, MD (pneumology), Arnold von Eckardstein, MD (clinical chemistry), Jean-Pierre Zellweger, MD (pneumology), and Elizabeth Zemp Stutz, MD, MPH (epidemiology).
Scientific Collaborators at Coordinating Centers
Pierre-Olivier Bridevaux, MD, MSc (pneumology), Ivan Curjuric, MD (epidemiology), Denise Felber Dietrich, MD, PhD (cardiology), Armin Gemperli, PhD (statistics), Medea Imboden, PhD (genetic and molecular biology), Dirk Keidel, MSc (statistics), and Patricia Städele-Kessler, MHS (statistics).
Create a personal account or sign in to: