Article Text

Socioeconomic disparities in secondhand smoke exposure among US never-smoking adults: the National Health and Nutrition Examination Survey 1988–2010
  1. Wen Qi Gan1,2,3,
  2. David M Mannino1,
  3. Ahmedin Jemal4
  1. 1Department of Preventive Medicine and Environmental Health, University of Kentucky College of Public Health, Lexington, Kentucky, USA
  2. 2Department of Population Health, Hofstra North Shore-LIJ School of Medicine, Great Neck, New York, USA
  3. 3Feinstein Institute for Medical Research, North Shore-Long Island Jewish Health System, Great Neck, New York, USA
  4. 4Surveillance and Health Services Research, American Cancer Society, Inc, Atlanta, Georgia, USA
  1. Correspondence to Dr Wen Qi Gan, Department of Preventive Medicine and Environmental Health, University of Kentucky College of Public Health, 111 Washington Avenue, Suite 220, Lexington, KY 40536, USA; wenqi.gan{at}uky.edu

Abstract

Background Secondhand smoke (SHS) is a leading preventable cause of illness, disability and mortality. There is a lack of quantitative analyses on socioeconomic disparities in SHS; especially, it is not known how socioeconomic disparities have changed in the past two decades in the USA.

Objectives To examine socioeconomic disparities and long-term temporal trends in SHS exposure among US never-smoking adults aged ≥20 years.

Methods 15 376 participants from the National Health and Nutrition Examination Survey (NHANES) 1999–2010 were included in the analysis of socioeconomic disparities; additional 8195 participants from NHANES III 1988–1994 were included in the temporal trend analysis. SHS exposure was assessed using self-reported exposure in the home and workplace as well as using serum cotinine concentrations ≥0.05 ng/mL. Individual socioeconomic status (SES) was assessed using poverty-to-income ratio.

Results During the period 1999–2010, 6% and 14% of participants reported SHS exposure in the home and workplace, respectively; 40% had serum cotinine-indicated SHS exposure. Individual SES was strongly associated with SHS exposure in a dose–response fashion; participants in the lowest SES group were 2–3 times more likely to be exposed to SHS compared with those in the highest SES group. During the period 1988–2010, the prevalence declined over 60% for the three types of SHS exposure. However, for cotinine-indicated exposure, the magnitudes of the declines were smaller for lower SES groups compared with higher SES groups, leading to widening socioeconomic disparities in SHS exposure.

Conclusions SHS exposure is still widespread among US never-smoking adults, and socioeconomic disparities for cotinine-indicated exposure have substantially increased in the past two decades.

  • Secondhand Smoke
  • Socioeconomic Status
  • Disparities

Statistics from Altmetric.com

Introduction

Secondhand smoke (SHS) is a complex mixture of gases and fine particles generated from the burning end of a tobacco product such as a cigarette as well as exhaled by active smokers.1 ,2 It is also known as passive smoke, involuntary smoke and environmental tobacco smoke. In the USA, the most common source of SHS is cigarettes, followed by pipes, cigars and other tobacco products.1 Tobacco smoke contains more than 7000 chemicals, at least 250 are known to be toxic, about 70 are known to cause cancer.2 Inhaling SHS can cause serious health problems in non-smokers such as respiratory and cardiovascular diseases, and there is no safe level of exposure.1–7 The US Surgeon General estimates that for non-smoking adults who inhale SHS, the risk of lung cancer increases 20–30%, and the risk of heart disease increases 25–30%1; each year approximately 7330 and 33 950 US non-smoking adults die from lung cancer and coronary heart disease, respectively, because of inhaling SHS.3 SHS is a leading preventable cause of illness, disability and mortality in the USA and worldwide.

Active cigarette smoking is strongly associated with individual socioeconomic status (SES); persons with lower SES are more likely to smoke cigarettes and other tobacco products.3 ,8–10 It is therefore plausible that non-smokers with lower SES are more likely to be exposed to SHS. Many previous studies reported socioeconomic disparities in SHS exposure among children11–16 and non-smoking adults.17–23 However, there is a lack of quantitative analyses on this issue in the USA17–19; especially it is not known how socioeconomic disparities in SHS exposure have changed in the past two decades. Furthermore, it has been reported that SHS exposure has been declining in the USA19 ,24; however, it is not clear whether SHS exposure has been equally declined for people from different SES groups. Therefore, we conducted this study to examine socioeconomic disparities and their long-term temporal trends in SHS exposure among never-smoking adults in the USA.

Methods

Study design and participants

The National Health and Nutrition Examination Survey (NHANES) is a series of cross-sectional surveys to assess the health and nutrition status of the US population. During each survey, a nationally representative sample of the civilian non-institutionalised population was selected using a complex, stratified, multistage probability sampling procedure. The survey included a household interview and a subsequent medical examination. During the household interview, various health-related questionnaires were administered by trained interviewers. During the medical examination, blood samples were collected for various laboratory measurements.25

In accordance with the National Center for Health Statistics (NCHS) analytic guidelines,26 we aggregated all available NHANES data from 1999 to 2010 to create a large combined data set. Never-smokers 20 years of age and older were selected for the analyses of the relationship between individual SES and SHS exposure. In the temporal trend analysis on the prevalence of SHS exposure, phase 1 (1988–1991) and phase 2 (1991–1994) of NHANES III data were also included to describe long-term temporal trends of SHS exposure. Informed consent was obtained from all participants.

Assessment of exposure to SHS

The current study was restricted to never-smoking adults, who had not smoked more than 100 cigarettes in their lifetime.

Self-reported exposure in the home: For NHANES 1999–2010, SHS exposure in the home was defined as never-smoking adults who answered ‘yes’ to the question: “Does anyone who lives here smoke cigarettes, cigars, or pipes anywhere inside this home?” For NHANES III 1988–1994, it was defined as never-smoking adults who reported that any household members smoked cigarettes inside their homes.

Self-reported exposure in the workplace: For NHANES 1999–2010, SHS exposure in the workplace was defined as never-smoking adults who “can smell the smoke from other people's cigarettes, cigars, or pipes” in the workplace. For NHANES III 1988–1994, it was defined as never-smoking adults who could smell other people's smoking in the workplace at least 1 h/day.

Serum cotinine-indicated exposure: Serum cotinine is a valid biomarker for SHS exposure, and has been widely used to measure SHS exposure in epidemiological studies.27–29 Based on the previous studies,1 ,24 never-smoking adults who had a serum cotinine concentration ≥0.05 ng/mL were regarded as exposed to SHS. In NHANES III 1988–1994 and NHANES 1999–2010, serum cotinine was measured by a high-performance liquid chromatography/atmospheric pressure ionisation tandem mass spectrometry (LC/MS/MS). This method has been described in detail elsewhere.30 ,31

Socioeconomic status

Poverty-to-income ratio (PIR) was used to measure each person's SES. The PIR is the ratio of annual household income to the family's corresponding poverty threshold published by the US Census Bureau in a given calendar year.32 PIR values ≥1 indicate participants living at or above the federal poverty level, whereas PIR values <1 indicate participants living below the federal poverty level. The PIR takes into account annual family income, family size and the minimum income needed to support the family in a specific year; therefore, PIR is able to more accurately (vs annual family income) reflect actual SES for participants living in different states and from different time periods (eTable 1). To examine a dose–response relationship between individual SES and SHS exposure, participants were divided into quintiles according to PIR levels, with quintile 1 representing the lowest SES group, and quintile 5 the highest SES group.

Statistical methods

The prevalence of SHS exposure was compared between different groups using the Rao-Scott χ2 test. The association between PIR levels and SHS exposure was examined using multiple logistic regression models adjusting for age (continuous), sex (man, women) and race/ethnicity (non-Hispanic Caucasian, non-Hispanic African-American, Mexican American or other); OR and 95% CI were calculated for each PIR group using the highest PIR group (quintile 5) as the reference category. In the temporal trend analysis, the absolute difference in the prevalence of SHS exposure between the lowest and highest PIR groups was calculated. The relative percentage change in the prevalence of SHS exposure between the later and early periods was calculated as the absolute difference in the prevalence of SHS exposure between the later and the early periods divided by the prevalence in the early period.

SAS survey procedures (SAS V.9.3, SAS Institute Inc) were used for the statistical analysis. Interview weights and examination weights were calculated and incorporated into the analyses of self-reported exposure and cotinine-indicated exposure, respectively, to account for the complex sampling design, differential probabilities of selection and non-response. All statistical tests were two-sided and a p value of less than 0.05 was considered to be statistically significant.

Results

A total of 15 376 never-smoking adults who participated in NHANES 1999–2010 were included in the analysis of socioeconomic disparities in SHS exposure. All these participants had information on household exposure, 8937 (58% of total participants) had information on workplace exposure and 13 698 (89% of total participants) had serum cotinine data. Additionally, never-smoking adults who participated in NHANES III 1988–1994 (8195 for household, 4467 for workplace and 6887 for cotinine-indicated exposure) were also included to examine temporal trends of SHS exposure over a period of 22 years.

During the period 1999–2010, 6% and 14% of participants were exposed to SHS in the home and workplace, respectively. Among 9355 participants with complete information on SHS exposure in the home and workplace, 21% were exposed to SHS in the home or workplace. Strikingly, 40% were exposed to SHS as indicated by serum cotinine levels. In the latest NHANES 2009–2010 survey, the corresponding prevalence was 4% for self-reported household exposure, 11% for self-reported workplace exposure and 30% for cotinine-indicated exposure. Table 1 shows that men (except for the household exposure), young people (eg, those aged <40 years), black people and people with lower education or lower PIR levels were more likely to be exposed to SHS.

Table 1

Prevalence (95% CI) of secondhand smoke exposure, NHANES 1999–2010

Overall, PIR levels were strongly associated with the prevalence of SHS exposure after adjustment for age, sex and race/ethnicity; compared with those in the highest PIR quintile, participants in the lowest PIR quintile were 2–3 times more likely to be exposed to SHS (table 2 and figure 1). The linear trend for the relationship was more evident for household exposure, but was less evident for workplace exposure (table 2 and figure 1). For cotinine-indicated exposure, the linear trend was particularly strong and statistically significant for all subgroups (table 2 and figure 1).

Table 2

ORs (95% CIs) for secondhand smoke exposure by quintiles of PIR levels, NHANES 1999–2010*

Figure 1

ORs and 95% CIs of SHS exposure by quintiles of PIR levels, NHANES 1999–2010. For each of the three groups, from the left to right, each error bar represents OR and 95% CI of SHS exposure for quintiles 4–1, respectively, compared with quintile 5 (the highest PIR group indicated by the left black dot). SHS, secondhand smoke; PIR, poverty-to-income ratio.

To examine temporal trends of SHS exposure, the participants were stratified by five time periods, including phase 1 (1988–1991) and phase 2 (1991–1994) of NHANES III, NHANES 1999–2002, NHANES 2003–2006 and NHANES 2007–2010. During the period 1988–2010, the prevalence of SHS exposure was substantially decreased to a similar extent for these three types of exposure: 68% for household exposure (eTable 2), 69% for workplace exposure (eTable 3) and 60% for cotinine-indicated exposure (eTable 4). For cotinine-indicated exposure, the magnitudes of the declines were dependent on individual PIR levels; lower PIR groups had smaller declines compared with higher PIR groups (figure 2 and eTable 4). For example, the prevalence decreased 71% for persons in the highest PIR quintile, but only 40% for persons in the lowest PIR quintile. As a result, the difference in the prevalence of SHS exposure between the lowest and highest PIR quintiles had substantially increased (figure 3 and eTable 4). This relationship was not observed for self-reported exposure in the home (eTable 2) and workplace (eTable 3). The visual impressions of figure 3 were consistent with the results of logistic regression analyses regarding the interaction between time period and family PIR in relation to SHS exposure. For cotinine-indicated exposure, the interaction was statistically significant (p<0.001) in unadjusted model as well as adjusted model including age, sex and race/ethnicity.

Figure 2

Relative declines in cotinine-indicated SHS exposure between 1988–1991 and 2007–2010 by quintiles of PIR levels. The p=0.018 for the linear trend test across quintiles of PIR levels. SHS, secondhand smoke; PIR, poverty-to-income ratio.

Figure 3

Prevalence of cotinine-indicated SHS exposure during the period 1988–2010 by quintiles of PIR levels. SHS, secondhand smoke; PIR, poverty-to-income ratio.

Discussion

Based on a large nationally representative sample of never-smoking adults, we found that exposure to SHS was still widespread in the USA, especially for cotinine-indicated exposure. Furthermore, we found that individual SES was strongly associated with SHS exposure in a dose–response fashion; persons in the lowest SES group were 2–3 times more likely to be exposed to SHS compared with those in the highest SES group. During the period 1988–2010, the prevalence of cotinine-indicated SHS exposure decreased 60%. However, the magnitudes of the declines were smaller for persons with lower SES compared with those with higher SES, leading to widening socioeconomic disparities in SHS exposure.

This study comprises self-reported SHS exposure and serum cotinine-indicated SHS exposure. The difference between these two types of exposure lies in that the former was reported subjectively by each person regarding SHS exposure at a specific location, homes or workplaces; whereas the latter was measured objectively by serum cotinine concentrations, which was able to reflect SHS exposure in various locations, including homes, workplaces and other locations such as cars, bars, restaurants and recreational settings. Because of differences in subjective judgement and potential under-reporting in self-reported exposure,33–35 as well as changes in interview questions concerning SHS exposure in the home and workplace between different NHANES surveys, we believe that serum cotinine-indicated SHS exposure is more valid and the corresponding findings are able to reflect a more complete picture in SHS exposure.24 ,27–29

In a study based on data from NHANES III and NHANES 1999–2002, Pirkle et al24 found that the median level of serum cotinine concentrations decreased 82% among US non-smoking adults, suggesting a substantial decline in SHS exposure among non-smoking adults. Furthermore, Chen et al36 examined serum cotinine concentrations among US non-smoking adults using the NHANES 2001–2006 data, and found that the median level of serum cotinine concentrations remained stable during the time period. The current study used a serum cotinine concentration of 0.05 ng/mL to identify persons exposed to SHS, and found that the prevalence of SHS exposure had steadily declined over a period of 22 years from 1988 to 2010 among US never-smoking adults. The major difference is that our study focused on the proportion of never-smoking adults exposed to SHS, whereas these two studies examined the degree of SHS exposure among non-smoking adults.24 ,36 Some previous studies also found that non-smoking adults with lower SES were more likely to be exposed to SHS.17–23 For example, based on the NHANES data, Max et al17 and two CDC reports18 ,19 found substantial SES disparities in SHS exposure among non-smokers in the USA. The current study described in detail the dose–response trends for the relationship between individual SES and SHS exposure across different demographic characteristics using the data from NHANES III 1988–1994 and NHANES 1999–2010. More importantly, our study examined long-term temporal trends for the socioeconomic disparities. This information is needed for developing more effective public health interventions to prevent SHS exposure.

Based on the 2001–2010 National Drug Strategy Household Surveys in Australia, Gartner and Hall11 investigated socioeconomic disparities in household SHS exposure among children under age 15. The authors found that the prevalence of SHS exposure was declined in all SES groups, but lower SES groups had much smaller declines compared with higher SES groups; the gap between the lowest and highest SES groups was thus increased. In a study based on the US National Health Interview Survey for children under age 18 in 1992 and 2000, Soliman et al12 found a similar situation that the declines were more substantial for children from higher SES families compared with children from lower SES families. Recently, Sims et al37 evaluated the effects of smoke-free legislation on salivary cotinine-indicated SHS exposure among non-smoking adults in England. The authors found that after the intervention, SHS exposure was substantially decreased in the population; however, the magnitudes of the declines varied by individual SES; persons with lower SES had smaller and non-significant declines compared with those with higher SES, leading to increased socioeconomic disparities in SHS exposure. Our results are consistent with these previous findings, suggesting that people of lower SES were more likely to be exposed to SHS, and they were less likely to eliminate SHS exposure afterward; socioeconomic disparities in SHS exposure therefore tend to increase over time.

The current study has some limitations that should be noted. As mentioned before, self-reported SHS exposure might not accurately reflect actual SHS exposure because of the following reasons. First, self-reported SHS exposure was subjective, and subject to recall bias. As active smoking had become more socially unacceptable, it was possible that some persons might under-report active smoking and thus SHS exposure.33–35 Second, household SHS exposure was defined as household members who smoked inside their homes; it was not certain whether never-smoking household members were actually exposed and how long they were exposed. Also, it was not known whether non-household members such as visitors smoked inside their homes. Third, as mentioned in the Methods section, the interview questions concerning self-reported SHS exposure were slightly different between NHANES III 1988–1994 and NHANES 1999–2010. As a result, the estimated prevalence would be potentially smaller in the early period (1988–1991), and the relative percentage changes would potentially underestimate the true magnitudes of declines in household (eTable 2) and workplace (eTable 3) exposure. To partly solve the problem, we also calculated relative percentage changes using NHANES 1999–2002 as the early period.

Additionally, serum cotinine as an indicator of SHS exposure has some limitations. First, serum cotinine has a half-life of approximately 16 h in the body, it could only reflect SHS exposure occurred within recent several days.38 Second, some studies have shown racial/ethnic differences in nicotine and cotinine metabolism, for example, serum cotinine concentrations tend to be higher in black than white active smokers and non-smokers exposed to SHS.39–41 But the racial/ethnic difference would not substantially affect our results because we used a cotinine concentration of 0.05 ng/mL as the cut-off to identify persons exposed to SHS. Third, although serum cotinine was able to objectively reflect SHS exposures in various locations, it was not able to pinpoint specific locations where exposures occurred. Finally, it should be noted that although the method of serum cotinine measurements had been continuously improved in NHANES, the limit of detection of 0.05 ng/mL was maintained in each cycle of NHANES,24 therefore the changes in sensitivity of the laboratory method would not affect our results.

Conclusions

The study found that SHS exposure is still widespread; 40% of US never-smoking adults were exposed to SHS as indicated by serum cotinine levels. There are considerable socioeconomic disparities in SHS exposure; persons with lower SES were more likely to be exposed to SHS compared with those with higher SES. During the period 1988–2010, the prevalence of cotinine-indicated SHS exposure decreased 60%. However, the magnitudes of the declines were smaller for lower SES groups compared with higher SES groups; the socioeconomic disparities in SHS exposure had thus substantially increased.

Given the widespread SHS exposure and the serious health consequences, SHS is still an important public health problem. More effective interventions are needed to reduce and eliminate SHS exposure in the USA. As noted in the 2006 Surgeon General's Report,1 people of lower SES are less likely to have smoke-free home rules and less likely to be protected by smoke-free workplace policies. Future SHS interventions need to pay more attention to people of lower SES to better control SHS exposure in the USA.

What this paper adds

  • Secondhand smoke (SHS) exposure is still widespread in the USA. During the period 1999–2010, 40% of US never-smoking adults were exposed to SHS as indicated by serum cotinine levels.

  • Individual socioeconomic status was strongly associated with SHS exposure in a dose–response fashion; persons in the lowest socioeconomic group were 2–3 times more likely to be exposed to SHS compared with those in the highest socioeconomic group.

  • In the past two decades, the prevalence of serum cotinine-indicated SHS exposure declined by 60%. However, the magnitudes of the declines were smaller for lower socioeconomic groups compared with higher socioeconomic groups, leading to widening socioeconomic disparities in SHS exposure.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

Footnotes

  • Contributors WQG, AJ and DMM conceptualised the study. WQG acquired and analysed the data, and takes responsibility for the accuracy of the data analysis. WQG wrote the manuscript. DMM and AJ critically revised the manuscript for important intellectual content. All authors have read and approved the final version of the manuscript.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval The NHANES was reviewed and approved by the NCHS Institutional Review Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.