Public Health Nursing Vol. 33 No. 1, pp. 82–89 0737-1209/© 2015 Wiley Periodicals, Inc doi: 10.1111/phn.12207
SPECIAL FEATURES: METHODS
A Comparison of Survey Measures and Biomarkers of Secondhand Tobacco Smoke Exposure among Nonsmokers Chizimuzo Okoli, PhD, MPH, MSN, RN University of Kentucky College of Nursing, Lexington, Kentucky Correspondence to: Chizimuzo Okoli, University of Kentucky College of Nursing, 517 College of Nursing Building, Lexington, KY 40536. E-mail: [email protected]
ABSTRACT Objectives: Secondhand tobacco smoke (SHS) exposure causes several adverse physical health outcomes. Conceptual differences in survey measures of psychosocial (SHS exposure from smokers in an individuals life) and physical (environments where an individual is exposed to SHS) SHS exposure exist. Few studies have examined the association between psychosocial and physical SHS exposures measures in comparison to biomarkers of SHS exposure. Design and Sample: A secondary analysis of cross-sectional data was examined among a convenience sample of 20 adults. Measures: Data included survey items on SHS exposure and hair nicotine and saliva cotinine levels. Spearman analysis was used to assess correlations among variables. Results: Medium and strong correlations were found among SHS exposure measures with the exception of saliva cotinine levels. Strong correlations were found among and between psychosocial and physical SHS exposure measures. Hair nicotine levels had medium strength associations with only perceived frequency of SHS exposure. Discussion: As psychosocial measures of exposure were associated with biomarkers, such measures (particularly perceived frequency of SHS exposure) should be added to surveys in addition to physical SHS exposure measures to enhance accuracy of SHS measurement. Future explorations with robust sample sizes should further examine the strength of relationship between psychosocial and physical SHS exposure measures. Key words: biomarkers, environmental tobacco smoke exposure, hair nicotine, psychosocial secondhand smoke exposure, saliva cotinine.
Background Secondhand tobacco smoke (SHS) is an important cause of diseases responsible for premature death in nonsmoking children and adults worldwide € (Oberg, Jaakkola, Woodward, Peruga, & Pr€ uss€ Ustun, 2011; U.S. Department of Health and Human Services, 2014). In the United States, between 1999 and 2007, there was a substantial decline in SHS exposure among nonsmokers from 52.9% to 40.1%. However, approximately 88 million nonsmokers remain exposed to SHS (Centers for Disease Control and Prevention, 2010). Composed of both side-stream and main stream tobacco smoke, SHS contains hundreds of chemicals among
82
which several are carcinogenic and even brief exposures confers some level of cardiovascular and respiratory health risk (Flouris, Vardavas, Metsios, Tsatsakis, & Koutedakis, 2010; U.S. Department of Health and Human Services, 2014). Understanding factors associated with SHS exposure, and eliminating such exposures continue to be important objectives of comprehensive tobacco control policy (Levy, Chaloupka, & Gitchell, 2004; World Health Organization, 2012). To date, several methods have been used to quantify SHS exposure, including surveys, biological markers, and personal/atmospheric monitoring (Avila-Tang et al., 2013; Lee, Arheart, Trapido,
Okoli : Measures of secondhand tobacco smoke exposure Soza-Vento, & Rodriguez, 2005; Repace, Al-Delaimy, & Bernert, 2006). A marker for SHS should be unique to tobacco smoke, present in sufficient quantities to permit detection at low smoking rates, emitted at similar rates from all cigarette brands, and maintained at a constant ratio to SHS for a wide range of cigarette brands and environmental conditions (Rothberg, Heloma, Svinhufvud, Kahkonen, & Reijula, 1998). Quantifications of SHS exposure should take into account the degree of the pollution concentration, and duration and time pattern of exposure (Jaakkola & Jaakkola, 1997). The accuracy of the exposure method depends on how well the actual exposure can be reconstructed from information assessed during the time of measurement (Jaakkola & Jaakkola, 1997). Each method to examine SHS exposure has inherent strengths and limitations and two methods, namely survey and biological measures will be the focus of this paper (Steenland & Savitz, 1997). Survey-based measures of SHS exposure are valuable because they are often brief, can estimate longer time intervals of exposure, and are typically easy to administer. However, these measures may be subject to inaccuracies related to self-report and recall bias. Survey measures also may not be conceptually congruent. For instance, certain questions may ask about psychosocial SHS exposure with words such as ‘exposure to smoking’, as compared to physical SHS exposure from ‘secondhand smoke exposure’ or ‘environmental tobacco smoke exposure’ (Choi, Ahluwalia, Harris, & Okuyemi, 2002; Eisner, Katz, & Yelin, 2001; Richardson et al., 2013). Where psychosocial SHS exposure assesses an individual’s perception of their exposure and the exposure from smokers in their life (such as partners, friends, or family members), physical SHS exposure denotes the environments in which an individual is exposed to smoking (such as work, home, school). Biomarkers of exposure, viewed as the ‘gold standard’ of measures, are often limited by an inability to adequately quantify duration of SHS exposure. For instance, where some biomarkers such as hair nicotine can measure longer SHS exposure durations, other biomarkers like serum cotinine are limited to shorter durations (Al-Delaimy, 2002; Benowitz et al., 2010; Eisner, Jacob, Benowitz, Balmes, & Blanc, 2009). Using an ecological approach, combining survey-based measures
83
in conjunction with biological and/or atmospheric/ personal measures can ideally enhance assessments of SHS exposure (Matt, Bernert, & Hovell, 2008). As biomarkers and atmospheric/personal measures of SHS may not be as readily available due to costs associated with equipment and diagnostic techniques, assessing the accuracy of survey measures is often warranted. However, little is known about the extent to which survey measures of SHS exposure are differentially related to biomarkers of chronic and acute SHS exposure. Hence, the aim of this exploratory study was to examine the associations between common survey items assessing psychosocial and physical SHS exposure with biomarkers of chronic and acute SHS exposure. Understanding which survey measures are associated with what type of exposure can further improve the utility of such measures in determining SHS exposure.
Methods Design and sample This study was based on secondary, cross-sectional data from baseline survey questions and biological markers of 11 male and 9 female participants obtained through quota sampling. The aim of the parent study was to examine the effect of prior SHS exposure on the responses to nicotine administration among nonsmokers (Okoli, Kodet, & Robertson, 2015). Eligibility for the study included (a) being 18 years and older, (b) not having used any tobacco products more than once in their lifetime and not having used any tobacco products in the past 10 years, (c) having at least 1 inch of hair on the scalp which is a requirement for hair nicotine analysis. Exclusion criteria were (a) being a current tobacco user, (b) self-reported history of cardiac complications, (c) self-reported history of a mental illness/neuropsychiatric disorders and/or having a prescription for pain treatment or for psychotropic medications, (d) being pregnant, (e) having undergone treatment for a substance/drug addiction, (f) currently using any recreational substances such as cocaine, amphetamine, opioids, marijuana, or had ever used such substances more than once in their lifetime or in the past 10 years, and (g) having dyed hair or completely bald, and (h) self-reported mothers use of tobacco products while pregnant with the potential participant.
84
Public Health Nursing
Volume 33
Number 1
Those eligible made an appointment at a University Clinical Trials Unit, where the research procedure was explained to each participant and informed consent was obtained. At baseline, participants completed a brief self-report survey containing items on demographic variables and measures of SHS exposure. After completing the survey, saliva and hair samples were obtained to determine current and past SHS smoke exposure, respectively. This study’s protocol and research procedures were approved by the University of Kentucky Institutional Review Board.
Measures Psychosocial SHS exposure was assessed using three measures of (a) exposure to smoking from partner and family members, (b) perceived frequency of SHS exposure, and (c) smokers in the home environment as follows: 1. Exposure to smoking from partner and family members was based on an adaptation from the work of Choi et al. (2002) in which respondents were asked to select the persons in their lives who smoke cigarettes from a list including mother, father, step-parent(s)/guardian(s), brother/sister, and husband/wife/common law/partner. Individuals who smoked were coded as 2, those who had quit were coded as 1, and those who had never smoked or their smoking status was unknown were coded as 0. The intention to provide a lower score if their partners or family members had quit smoking, as opposed to if they were currently smoking, was based on findings that smoking behavior is more likely if parents are current smokers as opposed to former smokers (den Exter Blokland, Engels, Hale, Meeus, & Willemsen, 2004). Responses for this measure ranged from 0 to 10, with higher scores indicating greater exposure to smoking from partner and family members. 2. Perceived frequency of SHS exposure was based on a one-item scale that assessed the degree of perceived frequency of exposure to SHS on a scale of 0 to 10 (Pilkington, Gray, & Gilmore, 2007; Richardson et al., 2013). Higher scores indicated greater perceived exposure.
January/February 2016
3. Smokers in home environment was based on a one-item measure that assessed the number of persons who smoke in the home with response choices as none, one, two, three, and four or more (Belanger et al., 2008). Because of low variability in responses, this measure was dichotomized to either “none” or “any”. Environmental SHS exposure was examined from three measures of (a) exposure in past 30 days, (b) exposure in past 7 days, and (c) total number of hours exposed in the last week as follows: 1. To assess SHS exposure in the past 30 days, participants chose from seven sources of potential exposure including cafe/restaurant, car, in your house, in others house, mall/shopping center, at/near school, and at work (Mack, Islam, Lee, Selzer, & Hodis, 2003; Okoli, Hall, Rayens, & Hahn, 2007; Richardson et al., 2013). Response choices for each source were never, once, a few times, more than a few times, and a lot. Responses for this measure ranged from 0 to 28 with higher scores indicating greater exposure. 2. To assess SHS exposure in the past 7 days, participants were asked a question similar to SHS exposure in the past 30 days with the exception that they were asked about their past 7 days exposure. Response for this measure also ranged from 0 to 28. 3. Total number of hours exposed in the last week was determined by asking participants to estimate their exposure per venue over the past week; and a summary score of total hours of exposure was determined. Chronic SHS exposure and acute biomarkers of SHS exposure were examined as follows: 1. Chronic SHS exposure, defined as past-month exposure, was assessed using hair nicotine based on previously established protocols (Al-Delaimy, 2002; Al-Delaimy, Crane, & Woodward, 2002; Hahn et al., 2006; Kim et al., 2014). Analysis of
Okoli : Measures of secondhand tobacco smoke exposure
85
hair samples was performed by the high-performance liquid chromatography with electro chemical detection (HPLC-ECD) method with a lower level of detection of 0.05 ng/mg (Mahoney & Al-Delaimy, 2001). Hair nicotine is considered a ‘gold standard’ by meeting the criteria specified for a marker of SHS exposure (Rothberg et al., 1998).
classified according to the following recommended convention scale:
SPECIAL FEATURES: METHODS
A Comparison of Survey Measures and Biomarkers of Secondhand Tobacco Smoke Exposure among Nonsmokers Chizimuzo Okoli, PhD, MPH, MSN, RN University of Kentucky College of Nursing, Lexington, Kentucky Correspondence to: Chizimuzo Okoli, University of Kentucky College of Nursing, 517 College of Nursing Building, Lexington, KY 40536. E-mail: [email protected]
ABSTRACT Objectives: Secondhand tobacco smoke (SHS) exposure causes several adverse physical health outcomes. Conceptual differences in survey measures of psychosocial (SHS exposure from smokers in an individuals life) and physical (environments where an individual is exposed to SHS) SHS exposure exist. Few studies have examined the association between psychosocial and physical SHS exposures measures in comparison to biomarkers of SHS exposure. Design and Sample: A secondary analysis of cross-sectional data was examined among a convenience sample of 20 adults. Measures: Data included survey items on SHS exposure and hair nicotine and saliva cotinine levels. Spearman analysis was used to assess correlations among variables. Results: Medium and strong correlations were found among SHS exposure measures with the exception of saliva cotinine levels. Strong correlations were found among and between psychosocial and physical SHS exposure measures. Hair nicotine levels had medium strength associations with only perceived frequency of SHS exposure. Discussion: As psychosocial measures of exposure were associated with biomarkers, such measures (particularly perceived frequency of SHS exposure) should be added to surveys in addition to physical SHS exposure measures to enhance accuracy of SHS measurement. Future explorations with robust sample sizes should further examine the strength of relationship between psychosocial and physical SHS exposure measures. Key words: biomarkers, environmental tobacco smoke exposure, hair nicotine, psychosocial secondhand smoke exposure, saliva cotinine.
Background Secondhand tobacco smoke (SHS) is an important cause of diseases responsible for premature death in nonsmoking children and adults worldwide € (Oberg, Jaakkola, Woodward, Peruga, & Pr€ uss€ Ustun, 2011; U.S. Department of Health and Human Services, 2014). In the United States, between 1999 and 2007, there was a substantial decline in SHS exposure among nonsmokers from 52.9% to 40.1%. However, approximately 88 million nonsmokers remain exposed to SHS (Centers for Disease Control and Prevention, 2010). Composed of both side-stream and main stream tobacco smoke, SHS contains hundreds of chemicals among
82
which several are carcinogenic and even brief exposures confers some level of cardiovascular and respiratory health risk (Flouris, Vardavas, Metsios, Tsatsakis, & Koutedakis, 2010; U.S. Department of Health and Human Services, 2014). Understanding factors associated with SHS exposure, and eliminating such exposures continue to be important objectives of comprehensive tobacco control policy (Levy, Chaloupka, & Gitchell, 2004; World Health Organization, 2012). To date, several methods have been used to quantify SHS exposure, including surveys, biological markers, and personal/atmospheric monitoring (Avila-Tang et al., 2013; Lee, Arheart, Trapido,
Okoli : Measures of secondhand tobacco smoke exposure Soza-Vento, & Rodriguez, 2005; Repace, Al-Delaimy, & Bernert, 2006). A marker for SHS should be unique to tobacco smoke, present in sufficient quantities to permit detection at low smoking rates, emitted at similar rates from all cigarette brands, and maintained at a constant ratio to SHS for a wide range of cigarette brands and environmental conditions (Rothberg, Heloma, Svinhufvud, Kahkonen, & Reijula, 1998). Quantifications of SHS exposure should take into account the degree of the pollution concentration, and duration and time pattern of exposure (Jaakkola & Jaakkola, 1997). The accuracy of the exposure method depends on how well the actual exposure can be reconstructed from information assessed during the time of measurement (Jaakkola & Jaakkola, 1997). Each method to examine SHS exposure has inherent strengths and limitations and two methods, namely survey and biological measures will be the focus of this paper (Steenland & Savitz, 1997). Survey-based measures of SHS exposure are valuable because they are often brief, can estimate longer time intervals of exposure, and are typically easy to administer. However, these measures may be subject to inaccuracies related to self-report and recall bias. Survey measures also may not be conceptually congruent. For instance, certain questions may ask about psychosocial SHS exposure with words such as ‘exposure to smoking’, as compared to physical SHS exposure from ‘secondhand smoke exposure’ or ‘environmental tobacco smoke exposure’ (Choi, Ahluwalia, Harris, & Okuyemi, 2002; Eisner, Katz, & Yelin, 2001; Richardson et al., 2013). Where psychosocial SHS exposure assesses an individual’s perception of their exposure and the exposure from smokers in their life (such as partners, friends, or family members), physical SHS exposure denotes the environments in which an individual is exposed to smoking (such as work, home, school). Biomarkers of exposure, viewed as the ‘gold standard’ of measures, are often limited by an inability to adequately quantify duration of SHS exposure. For instance, where some biomarkers such as hair nicotine can measure longer SHS exposure durations, other biomarkers like serum cotinine are limited to shorter durations (Al-Delaimy, 2002; Benowitz et al., 2010; Eisner, Jacob, Benowitz, Balmes, & Blanc, 2009). Using an ecological approach, combining survey-based measures
83
in conjunction with biological and/or atmospheric/ personal measures can ideally enhance assessments of SHS exposure (Matt, Bernert, & Hovell, 2008). As biomarkers and atmospheric/personal measures of SHS may not be as readily available due to costs associated with equipment and diagnostic techniques, assessing the accuracy of survey measures is often warranted. However, little is known about the extent to which survey measures of SHS exposure are differentially related to biomarkers of chronic and acute SHS exposure. Hence, the aim of this exploratory study was to examine the associations between common survey items assessing psychosocial and physical SHS exposure with biomarkers of chronic and acute SHS exposure. Understanding which survey measures are associated with what type of exposure can further improve the utility of such measures in determining SHS exposure.
Methods Design and sample This study was based on secondary, cross-sectional data from baseline survey questions and biological markers of 11 male and 9 female participants obtained through quota sampling. The aim of the parent study was to examine the effect of prior SHS exposure on the responses to nicotine administration among nonsmokers (Okoli, Kodet, & Robertson, 2015). Eligibility for the study included (a) being 18 years and older, (b) not having used any tobacco products more than once in their lifetime and not having used any tobacco products in the past 10 years, (c) having at least 1 inch of hair on the scalp which is a requirement for hair nicotine analysis. Exclusion criteria were (a) being a current tobacco user, (b) self-reported history of cardiac complications, (c) self-reported history of a mental illness/neuropsychiatric disorders and/or having a prescription for pain treatment or for psychotropic medications, (d) being pregnant, (e) having undergone treatment for a substance/drug addiction, (f) currently using any recreational substances such as cocaine, amphetamine, opioids, marijuana, or had ever used such substances more than once in their lifetime or in the past 10 years, and (g) having dyed hair or completely bald, and (h) self-reported mothers use of tobacco products while pregnant with the potential participant.
84
Public Health Nursing
Volume 33
Number 1
Those eligible made an appointment at a University Clinical Trials Unit, where the research procedure was explained to each participant and informed consent was obtained. At baseline, participants completed a brief self-report survey containing items on demographic variables and measures of SHS exposure. After completing the survey, saliva and hair samples were obtained to determine current and past SHS smoke exposure, respectively. This study’s protocol and research procedures were approved by the University of Kentucky Institutional Review Board.
Measures Psychosocial SHS exposure was assessed using three measures of (a) exposure to smoking from partner and family members, (b) perceived frequency of SHS exposure, and (c) smokers in the home environment as follows: 1. Exposure to smoking from partner and family members was based on an adaptation from the work of Choi et al. (2002) in which respondents were asked to select the persons in their lives who smoke cigarettes from a list including mother, father, step-parent(s)/guardian(s), brother/sister, and husband/wife/common law/partner. Individuals who smoked were coded as 2, those who had quit were coded as 1, and those who had never smoked or their smoking status was unknown were coded as 0. The intention to provide a lower score if their partners or family members had quit smoking, as opposed to if they were currently smoking, was based on findings that smoking behavior is more likely if parents are current smokers as opposed to former smokers (den Exter Blokland, Engels, Hale, Meeus, & Willemsen, 2004). Responses for this measure ranged from 0 to 10, with higher scores indicating greater exposure to smoking from partner and family members. 2. Perceived frequency of SHS exposure was based on a one-item scale that assessed the degree of perceived frequency of exposure to SHS on a scale of 0 to 10 (Pilkington, Gray, & Gilmore, 2007; Richardson et al., 2013). Higher scores indicated greater perceived exposure.
January/February 2016
3. Smokers in home environment was based on a one-item measure that assessed the number of persons who smoke in the home with response choices as none, one, two, three, and four or more (Belanger et al., 2008). Because of low variability in responses, this measure was dichotomized to either “none” or “any”. Environmental SHS exposure was examined from three measures of (a) exposure in past 30 days, (b) exposure in past 7 days, and (c) total number of hours exposed in the last week as follows: 1. To assess SHS exposure in the past 30 days, participants chose from seven sources of potential exposure including cafe/restaurant, car, in your house, in others house, mall/shopping center, at/near school, and at work (Mack, Islam, Lee, Selzer, & Hodis, 2003; Okoli, Hall, Rayens, & Hahn, 2007; Richardson et al., 2013). Response choices for each source were never, once, a few times, more than a few times, and a lot. Responses for this measure ranged from 0 to 28 with higher scores indicating greater exposure. 2. To assess SHS exposure in the past 7 days, participants were asked a question similar to SHS exposure in the past 30 days with the exception that they were asked about their past 7 days exposure. Response for this measure also ranged from 0 to 28. 3. Total number of hours exposed in the last week was determined by asking participants to estimate their exposure per venue over the past week; and a summary score of total hours of exposure was determined. Chronic SHS exposure and acute biomarkers of SHS exposure were examined as follows: 1. Chronic SHS exposure, defined as past-month exposure, was assessed using hair nicotine based on previously established protocols (Al-Delaimy, 2002; Al-Delaimy, Crane, & Woodward, 2002; Hahn et al., 2006; Kim et al., 2014). Analysis of
Okoli : Measures of secondhand tobacco smoke exposure
85
hair samples was performed by the high-performance liquid chromatography with electro chemical detection (HPLC-ECD) method with a lower level of detection of 0.05 ng/mg (Mahoney & Al-Delaimy, 2001). Hair nicotine is considered a ‘gold standard’ by meeting the criteria specified for a marker of SHS exposure (Rothberg et al., 1998).
classified according to the following recommended convention scale:
