Abstract
Objective
This study examined the joint influence of work- and household-related variables on smoking behavior among a population representative sample of blue-collar workers with live-in partners.
Methods
The study used data on 1,389 blue-collar workers from the Tobacco Use Supplement to the United States Current Population Survey 2002 to 2003 longitudinal overlap sample. Unadjusted and adjusted logistical regression analyses, which employed sampling and replicate weights to account for sampling design, were run to estimate independent and joint effects of the predictors.
Results
In adjusted analyses, partner smoking (OR=4.97, 95%CI=3.02–8.18) and complete and partial home smoking policy (OR=0.16, 95%CI=0.09–0.29 and OR=0.39, 95%CI=0.23–0.68, respectively) were significant predictors of smoking status, but worksite smoking policies and presence of a young child under 5 in the household were not (p>0.05). Baseline complete home smoking ban was a significant predictor of subsequent cessation (OR=3.49, 95%CI=1.19–10.23), while partner smoking status, workplace smoking policy, and the presence of a young child in the home did not predict cessation (p>0.05).
Conclusion
Household-related variables were significant predictors of smoking status and cessation among blue-collar workers. Current efforts to decrease smoking in this group, which are mostly focused on work-related risk factors, should consider how to incorporate household risk factors.
Keywords: smoking, occupational class, blue-collar, partner smoking
Introduction
The continuing decline of smoking in the US, to a current population prevalence of 19.3%, is cited as one of the greatest health advances in the past three decades (Schroeder, 2008; Syamlal et al., 2011). However, smoking prevalence among some blue-collar workers, such as construction and mining workers, is about 30% (Syamlal et al., 2011). Blue-collar workers have a younger age of smoking initiation, smoke more heavily, and have lower smoking cessation rates (Barbeau et al., 2004; Giovino et al., 2000). Blue-collar worksites are also less likely to provide social support for smoking cessation, and less likely to have workplace smoking policies or health insurance coverage for smoking cessation treatment (Barbeau et al., 2001; Shopland et al., 2004; Sorensen et al., 2002). The lack of workplace smoking policies is particularly important because these policies may increase smoking cessation attempts and success. A review of literature found that workers in workplaces with strict smoking policies were less likely to be smokers and more likely to quit smoking (Brownson et al., 2002). These findings were echoed in a longitudinal study; workers at worksites that adopted or maintained worksite smoking policies during the study period were almost twice as likely to quit smoking as workers whose worksites did not (Bauer et al., 2005).
Despite many studies on blue-collar workers and smoking, several gaps remain. First, these studies have not considered household-related risk factors that have been shown to impact smoking behaviors, such as home smoking restrictions, which, whether self-imposed or not, have been associated with decreased likelihood of smoking (Borland et al., 2006; Farkas et al., 1999; Hyland et al., 2009). Similarly, the smoking status of live-in partners has been shown to influence smoking behaviors (Dollar et al., 2009; Franks et al., 2002; Manchon Walsh et al., 2007). One study of blue-collar apprentices revealed that, compared to non-smokers, smokers were 13 times more likely to have partners who smoke (Okechukwu et al., 2010). Another potentially important household-related predictor of smoking behavior is the presence of young children because households with young children are more likely to restrict smoking within the home (Borland et al., 2006; Hawkins and Berkman, 2011).
Only one study has combined knowledge about work- and household-related smoking risk factors to investigate smoking outcomes (Farkas et al., 1999). Farkas and colleagues (1999) found that restricting smoking in homes and workplaces and living with a smoker were significant predictors of cessation attempts and success. However, the authors did not consider occupational class and used cross sectional data. The current study uses a population representative longitudinal sample of blue-collar workers to investigate the impact of both work- and household-related factors on smoking and smoking cessation and also evaluates the relative contribution of each factor to these outcomes.
Methods
Data Sources
Data for this study are from the 2002 and 2003 Tobacco Use Supplement to the Current Population Survey (TUS-CPS). The TUS-CPS uses a multistage probability sampling design to capture smoking behaviors, employment, and demographic information for a representative sample of US households (CPS, 2006; Soulakova et al., 2009). The TUS-CPS conducted a special cessation supplement in February 2003, which, because it overlapped with the February 2002 sample, provided longitudinal data.
The study sample included blue-collar workers (classified using the US Standard Occupational Classification System codes), who were ages 18 or older with live-in spouses or partners (n=1389). The final population representative sample of 1,008 blue-collar workers was reached after excluding 381 subjects who were missing the key outcome and exposure variables of respondent smoking status (n=5), partner smoking status (n=104), workplace smoking policy (n=279), and home smoking policy (n=5). Some subjects had missing values for more than one variable. Given the proportion of respondents with missing information on household income (6%), a prediction algorithm established by the CPS was used to impute income (Clark, 2011).
Study Measures
Outcomes
Individuals were classified as smokers at baseline if they reported lifetime smoking of at least 100 cigarettes and smoking “every day” or “some days” in the 2002 survey (CDC, 1994). A smoker is considered to have quit if they reported smoking in 2002, but reported smoking “not at all” in the 2003 survey.
Exposures
The same criteria for respondent smoking status were used to define partner smoking status. Home smoking policy was based on answers to the question “Which statement best describes the rules about smoking inside your home?” Choices were “no one is allowed to smoke anywhere inside the home” (complete ban), “smoking is allowed in some places or at some times inside the home” (partial restriction) and “smoking is permitted anywhere inside the home” (no ban).
TUS-CPS classified Workplace smoking policy based on answers to the following questions: “Does your place of work have an official policy that restricts smoking in any way?” “Which of these best describes your place of work’s smoking policy for indoor public or common areas?” and “Which of these best describes your place of work’s smoking policy for work areas?” Classifications were complete ban (official policy banning smoking in all public or work areas), partial restriction (official policy but smoking allowed in some or all public or work areas) and no ban (no official policy). Those who reported that they work mostly outdoors or travel to different sites were grouped into their own separate category.
Presence of young children
TUS-CPS respondents who reported having a child under the age of 5 in the household were classified as “yes” while others were classified as “no.”
Covariates
Socio-demographic characteristics
Gender classifications were male or female. Annual household income was divided into four categories (less than $25,000, $25,000–$49,999, $50,000–$74,999, and $75,000 or greater), and educational attainment was categorized as high school graduate or non-high school graduate. Responses to the two race and ethnicity questions were combined to form four racial/ethnic categories: non-Hispanic white; non-Hispanic black; Hispanic; and other.
Statistical Analyses
All analyses were conducted in 2012 with a significance level of 0.05. SAS version 9.2 was used to obtain descriptive statistics. Models for each of the outcome variables were estimated using SAS-Callable SUDAAN version 10, which allows for calculation of standard errors and confidence intervals that are valid for the TUS complex sampling design. The values of all predictor variables and covariates were obtained from responses to the 2002 survey.
Model building began with the estimation of simple logistic regression models to obtain the unadjusted effect of each exposure. Multiple logistic regression analyses were then used to examine the independent and joint influence of exposure variables on smoking behaviors, controlling for age, gender, race/ethnicity, education, and household income. Interactions among independent variables were also examined. Each model used a balanced repeated replication Fay’s adjustment factor of 4 and the TUS-supplied longitudinal sampling and replicate weights to account for possible within-state clustering, the multistage sampling schema, and nonresponse.
The adequacy statistic, which reports the percentage of explainable variation in the outcome variable that can be explained by each predictor variable, was used to rank the main exposures by explanatory value. Total explainable variation is −2*log likelihood (−2LL) of the full model that contains all predictors from a given set. The explainable variation for each predictor is the −2LL of the unadjusted model. Higher adequacy values indicate more explanatory power.
Results
The smoking prevalence among this sample of blue-collar workers was 26.3% in 2002 and 24.5% by 2003 (Table 1). Approximately 20% of respondents reported having smokers as partners in 2002; while 68.7% reported a complete home smoking ban, 17.4% reported a partial restriction, and 13.9% reported no ban. Thirty-seven percent of respondents reported a complete workplace smoking ban, 14.5% reported a partial restriction, 11.1% reported no restriction, and 38.0% reported working outdoors or traveling to different sites.
Table 1.
Sociodemographic and smoking characteristics of blue collar workers and their partners, Tobacco Use Supplement – Current Population Survey Overlap Sample 2002–2003 (N=1008)
| n | %a | |
|---|---|---|
| 2002 Smoking Status | ||
| Smoker | 268 | 26.3 |
| Non-smoker | 740 | 73.7 |
| 2003 Smoking Status | ||
| Smoker | 238 | 24.5 |
| Non-smoker | 770 | 75.5 |
| Successful Cessation Attempts | ||
| Yes | 58 | 21.1 |
| No | 210 | 78.9 |
| Partner Smoking Status | ||
| Smoker | 219 | 19.8 |
| Non-smoker | 789 | 80.2 |
| Partner Status | ||
| Spouse | 923 | 92.1 |
| Partner | 85 | 7.9 |
| Race/Ethnicity | ||
| White, non-Hispanic | 803 | 64.9 |
| Black, non-Hispanic | 60 | 9.7 |
| Other | 18 | 2.9 |
| Hispanic | 127 | 22.6 |
| Educational Attainment | ||
| Non-high school graduate | 174 | 23.3 |
| High school graduate | 834 | 76.7 |
| Annual Family Income | ||
| Less than $25,000 | 139 | 16.9 |
| $25,000 – $49,999 | 400 | 40.2 |
| $50,000 – $74,999 | 289 | 27.6 |
| $75,000 or greater | 180 | 15.3 |
| Child under the age of 5 | ||
| No | 769 | 70.3 |
| Yes | 239 | 29.7 |
| Gender | ||
| Female | 187 | 14.8 |
| Male | 821 | 85.2 |
| Work Location | ||
| Mostly Indoors | 606 | 59.2 |
| Mostly Outdoors | 221 | 20.7 |
| Travel to Different Buildings/Sites | 157 | 17.3 |
| Other | 24 | 2.8 |
| Occupation Description | ||
| Precision production, craft and repair occupations | 467 | 45.8 |
| Machine operators, assemblers and inspectors | 220 | 21.2 |
| Transportation and material moving occupations | 199 | 19.0 |
| Handlers, equipment cleaners, helpers and laborers | 122 | 14.0 |
| Home Smoking Policy | ||
| No ban | 166 | 13.9 |
| Partial restriction | 178 | 17.4 |
| Complete ban | 664 | 68.7 |
| Work Place Smoking Policy | ||
| No ban | 110 | 11.1 |
| Partial restriction | 147 | 14.5 |
| Complete ban | 373 | 36.5 |
| Work outdoors/Travel to different sites | 378 | 38.0 |
| Age | ||
| Mean (SD) | 43.2 (11.1) | |
Percentages are weighted.
Smoking Status
Partner smoking status and home smoking policy were significantly associated with smoking status in both unadjusted and adjusted analyses (Table 2). In the single predictor multivariable model (model 1), partner smoking was associated with 8.6 times greater odds of being a smoker (95%CI: 5.64–13.09), which decreased to 5.0 in the final model (95%CI 3.02–8.18). Complete home smoking ban was associated with significantly lower odds of smoking (OR: 0.09, 95%CI: 0.05–0.16); as was a partial restriction (OR: 0.35, 95%CI: 0.20–0.63) (model 2). Both remained significant predictors of smoking in the final model. Worksite smoking policy and presence of a young child were not significantly associated with smoking status in unadjusted or adjusted analysis.
Table 2.
Association of work-related and home related risk factors with respondent smoking at baseline among blue collar workers, Tobacco Use Supplement – Current Population Survey Overlap Sample 2002–2003a
| N = 1008
|
||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Unadjusted Models | Model 1 | Model 2 | Model 3 | Model 4 | Final Model | |||||||||||||
| OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | |||||||
| Partner Smoking Status | ||||||||||||||||||
| Smoker | 8.15 | 5.50 | 12.07 | 8.59 | 5.64 | 13.09 | 4.97 | 3.02 | 8.18 | |||||||||
| Non-smoker | REF | - | - | REF | - | - | REF | - | - | |||||||||
| Home Smoking Policy | ||||||||||||||||||
| Complete ban | 0.10 | 0.06 | 0.17 | 0.09 | 0.05 | 0.16 | 0.16 | 0.09 | 0.29 | |||||||||
| Partial restriction | 0.39 | 0.22 | 0.70 | 0.35 | 0.20 | 0.63 | 0.39 | 0.23 | 0.68 | |||||||||
| No ban | REF | - | - | REF | - | - | REF | - | - | |||||||||
| Work Place Smoking Policy | ||||||||||||||||||
| Complete ban | 0.66 | 0.37 | 1.17 | 0.64 | 0.35 | 1.16 | 0.83 | 0.39 | 1.78 | |||||||||
| Partial restriction | 1.04 | 0.55 | 1.96 | 0.96 | 0.51 | 1.79 | 1.17 | 0.55 | 2.50 | |||||||||
| Work outdoors/Travel to different sites | 0.83 | 0.48 | 1.42 | 0.85 | 0.49 | 1.46 | 1.23 | 0.62 | 2.43 | |||||||||
| No ban | REF | - | - | REF | - | - | REF | - | - | |||||||||
| Child under the age of 5 | ||||||||||||||||||
| Yes | 0.95 | 0.65 | 1.40 | 0.85 | 0.50 | 1.43 | 1.20 | 0.66 | 2.19 | |||||||||
| No | REF | - | - | REF | - | - | REF | - | - | |||||||||
| Age | 1.00 | 0.98 | 1.01 | 1.01 | 0.99 | 1.02 | 0.99 | 0.97 | 1.01 | 1.00 | 0.98 | 1.01 | 0.99 | 0.97 | 1.01 | 1.01 | 0.98 | 1.03 |
| Gender | ||||||||||||||||||
| Male | 0.85 | 0.54 | 1.35 | 1.25 | 0.72 | 2.17 | 1.01 | 0.62 | 1.65 | 0.83 | 0.52 | 1.31 | 0.90 | 0.58 | 1.41 | 1.13 | 0.64 | 1.98 |
| Female | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - |
| Race/Ethnicity | ||||||||||||||||||
| White, non-Hispanic | 1.54 | 0.97 | 2.45 | 1.47 | 0.75 | 2.89 | 1.25 | 0.63 | 2.49 | 2.20 | 1.20 | 4.04 | 2.25 | 1.23 | 4.14 | 1.04 | 0.52 | 2.05 |
| Black, non-Hispanic | 1.55 | 0.70 | 3.47 | 1.79 | 0.64 | 5.00 | 2.11 | 0.77 | 5.78 | 2.02 | 0.83 | 4.92 | 2.12 | 0.86 | 5.20 | 1.80 | 0.62 | 5.20 |
| Other | 1.12 | 0.24 | 5.26 | 1.61 | 0.28 | 9.44 | 0.79 | 0.10 | 6.31 | 1.78 | 0.38 | 8.23 | 1.72 | 0.37 | 8.12 | 1.00 | 0.13 | 7.83 |
| Hispanic | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - |
| Educational Attainment | ||||||||||||||||||
| Non-high school graduate | 0.93 | 0.57 | 1.54 | 1.04 | 0.53 | 2.02 | 0.91 | 0.48 | 1.70 | 0.90 | 0.51 | 1.59 | 0.92 | 0.53 | 1.61 | 0.94 | 0.47 | 1.90 |
| High school graduate | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - |
| Annual Family Income | ||||||||||||||||||
| Less than $25,000 | 2.76 | 1.47 | 5.17 | 3.50 | 1.72 | 7.12 | 2.70 | 1.23 | 5.93 | 4.03 | 1.98 | 8.21 | 4.10 | 1.99 | 8.44 | 2.63 | 1.21 | 5.69 |
| $25,000 – $49,999 | 2.03 | 1.22 | 3.39 | 2.09 | 1.19 | 3.67 | 1.74 | 0.95 | 3.18 | 2.24 | 1.32 | 3.81 | 2.25 | 1.32 | 3.85 | 1.69 | 0.93 | 3.07 |
| $50,000 – $74,999 | 1.26 | 0.71 | 2.23 | 1.10 | 0.58 | 2.08 | 1.05 | 0.54 | 2.06 | 1.30 | 0.73 | 2.32 | 1.30 | 0.73 | 2.31 | 0.94 | 0.48 | 1.86 |
| $75,000 or greater | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - |
Smoking Cessation
In unadjusted models of smoking cessation (Table 3), having a smoking partner led to significantly lower odds of cessation (OR: 0.38, 95%CI: 0.19–0.74). A complete home smoking ban significantly predicted cessation (OR: 4.34, 95%CI: 1.76–10.72), as did a partial restriction (OR: 3.16, 95%CI: 1.25–7.97), in comparison to no ban. Workplace smoking policy and presence of a young child were not significant predictors of smoking cessation.
Table 3.
Association of work-related and home related risk factors from 2002 with respondent cessation in 2003 among blue collar workers, Tobacco Use Supplement – Current Population Survey Overlap Sample 2002–2003a
| N = 268
|
||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Unadjusted Models | Model 1 | Model 2 | Model 3 | Model 4 | Final Model | |||||||||||||
| OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | OR | 95% CI | |||||||
| Partner Smoking Status | ||||||||||||||||||
| Smoker | 0.38 | 0.19 | 0.74 | 0.47 | 0.20 | 1.12 | 0.54 | 0.22 | 1.35 | |||||||||
| Non-smoker | REF | - | - | REF | - | - | REF | - | - | |||||||||
| Home Smoking Policy | ||||||||||||||||||
| Complete ban | 4.34 | 1.76 | 10.72 | 3.89 | 1.39 | 10.86 | 3.49 | 1.19 | 10.23 | |||||||||
| Partial restriction | 3.16 | 1.25 | 7.97 | 3.05 | 0.97 | 9.53 | 2.85 | 0.83 | 9.85 | |||||||||
| No ban | REF | - | - | REF | - | - | REF | - | - | |||||||||
| Work Place Smoking Policy | ||||||||||||||||||
| Complete ban | 0.54 | 0.18 | 1.58 | 0.49 | 0.14 | 1.70 | 0.39 | 0.11 | 1.38 | |||||||||
| Partial restriction | 0.24 | 0.04 | 1.61 | 0.32 | 0.06 | 1.90 | 0.27 | 0.04 | 1.87 | |||||||||
| Work outdoors/Travel to different sites | 0.65 | 0.25 | 1.71 | 0.56 | 0.19 | 1.72 | 0.50 | 0.16 | 1.57 | |||||||||
| No ban | REF | - | - | REF | - | - | REF | - | - | |||||||||
| Child under the age of 5 | ||||||||||||||||||
| Yes | 0.87 | 0.35 | 2.19 | 0.62 | 0.21 | 1.81 | 0.48 | 0.16 | 1.48 | |||||||||
| No | REF | - | - | REF | - | - | REF | - | - | |||||||||
| Age | 1.01 | 0.98 | 1.05 | 1.01 | 0.97 | 1.05 | 1.02 | 0.98 | 1.06 | 1.01 | 0.97 | 1.06 | 1.00 | 0.97 | 1.05 | 1.00 | 0.96 | 1.05 |
| Gender | ||||||||||||||||||
| Male | 3.86 | 1.24 | 12.02 | 2.67 | 0.77 | 9.28 | 2.82 | 0.77 | 10.37 | 2.90 | 0.80 | 10.49 | 3.08 | 0.87 | 10.94 | 2.64 | 0.75 | 9.32 |
| Female | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - |
| Race/Ethnicity | ||||||||||||||||||
| White, non-Hispanic | 0.27 | 0.11 | 0.67 | 0.32 | 0.12 | 0.85 | 0.43 | 0.16 | 1.18 | 0.26 | 0.09 | 0.74 | 0.25 | 0.09 | 0.67 | 0.38 | 0.12 | 1.20 |
| Black, non-Hispanic | 0.11 | 0.02 | 0.83 | 0.18 | 0.02 | 1.33 | 0.15 | 0.02 | 1.27 | 0.19 | 0.03 | 1.34 | 0.15 | 0.02 | 1.11 | 0.17 | 0.02 | 1.48 |
| Other | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
| Hispanic | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - |
| Educational Attainment | ||||||||||||||||||
| Non-high school graduate | 1.88 | 0.76 | 4.68 | 1.06 | 0.38 | 2.95 | 1.29 | 0.44 | 3.72 | 1.13 | 0.43 | 3.00 | 1.21 | 0.45 | 3.21 | 1.10 | 0.35 | 3.40 |
| High school graduate | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - |
| Annual Family Income | ||||||||||||||||||
| Less than $25,000 | 0.74 | 0.18 | 2.99 | 0.51 | 0.13 | 1.98 | 0.61 | 0.14 | 2.70 | 0.52 | 0.10 | 2.56 | 0.49 | 0.10 | 2.33 | 0.63 | 0.12 | 3.26 |
| $25,000 – $49,999 | 0.38 | 0.11 | 1.33 | 0.44 | 0.13 | 1.47 | 0.55 | 0.15 | 2.06 | 0.45 | 0.11 | 1.88 | 0.43 | 0.11 | 1.67 | 0.53 | 0.13 | 2.11 |
| $50,000 – $74,999 | 0.55 | 0.15 | 2.00 | 0.78 | 0.21 | 2.90 | 0.82 | 0.20 | 3.33 | 0.85 | 0.21 | 3.39 | 0.70 | 0.17 | 2.80 | 1.09 | 0.27 | 4.42 |
| $75,000 or greater | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - | REF | - | - |
After adjusting for covariates, partner smoking status was no longer significant (OR: 0.47, 95%CI: 0.20–1.12). Complete home smoking ban remained significant (OR: 3.89, 95%CI: 1.39–10.86), but partial restriction did not (OR: 3.05, 95%CI: 0.97–9.53). Workplace smoking policy and the presence of a young child remained non-significant. Only complete home smoking ban significantly predicted smoking cessation in the final model (OR: 3.49, 95%CI 1.19–10.23). None of the interaction terms tested was significant.
Adequacy Statistics
The total explainable variance for the multivariable model of smoking status and smoking cessation was 247.53 and 47.19 respectively (−2*log likelihood) (table 4). Examining the predictive ability of each independent variable, home smoking policy had the highest explainable variance for both smoking status and smoking cessation (63.4% and 30.8% respectively); this was followed by partner smoking status (62.5% and 20.9% respectively). Workplace smoking policy and the presence of a young child had very low explanatory value for both smoking status and smoking behavior.
Table 4.
Adequacy statistic for final model of work-related and home related risk factors from 2002 predicting respondent cessation in 2003 among blue collar workers, Tobacco Use Supplement – Current Population Survey Overlap Sample 2002–2003a
| Baseline Smoking Status | Smoking Cessation | |||
|---|---|---|---|---|
|
| ||||
| Variable | −2LL | Adequacy (%) | −2LL | Adequacy (%) |
| Total explainable variance by the entire set of predictors | 247.53 | N/A | 47.19 | N/A |
| Partner smoking status | 154.74 | 62.5 | 9.85 | 20.9 |
| Home smoking policy | 157.95 | 63.4 | 14.52 | 30.8 |
| Workplace smoking policy | 5.87 | 2.4 | 6.42 | 13.6 |
| Child under the age of 5 | 0.09 | 0.04 | 0.16 | 0.34 |
Model controlled for age, gender, race/ethnicity, educational attainment, and household income.
Discussion
This study examined the combined influence of work- and household-related risk factors for smoking status and cessation using a longitudinal population-representative sample of blue-collar workers. Home smoking ban and partner’s smoking predicted smoking status but workplace smoking policy and presence of a child under five did not. Household-related predictors had the highest explainable variance in the final predictive models, but only complete home smoking ban was significantly associated with smoking cessation in longitudinal analyses.
Workplace smoking policy did not predict smoking in either cross-sectional or longitudinal analysis. This finding is contrary to Farkas and colleagues’ analysis of cross-sectional data from the 1992 TUS-CPS (1999). The current study only involved blue-collar workers, but Farkas and colleagues conducted their analysis among all workers and did not control for occupational class. To understand differences between the two findings, the present study repeated the analysis done by Farkas and colleagues by using only 2002 cross-sectional data from all workers. In that analysis, workplace smoking policy predicted cessation when occupational class was not controlled for in the model. Other studies that did not consider household predictors have found associations between workplace policy and smoking (Brownson et al., 2002; Ham et al., 2011); but one study, which investigated specific workplaces instead of clustering all workplaces together found that the association between workplace policy and smoking differ by types of workplaces (Bitler et al., 2010).
This study did not find any association between presence of young children in households and smoking behavior. This finding is surprising because other studies found that presence of children in the household increases odds of home smoking ban, which is associated with smoking cessation (Hawkins and Berkman, 2011). Similar to present study, most studies have found that smokers are more likely to have partners who smoke, but studies of partner smoking cessation have been mixed. Two studies of non-US populations found an association between partner smoking and smoking cessation (Manchon Walsh et al., 2007; Monden et al., 2003) while a study with a US population did not find significant associations between partner smoking and smoking cessation (Franks et al., 2002). More studies are needed on the issue.
One limitation of the current study is potential exposure misclassification. For example, workers’ own smoking behaviors may influence their views of smoking policies at home and work. There is currently no objective repository of home smoking rules. Objective measures of workplace policies, such as state and local clean indoor air laws, exist. However, these laws might not reflect actual exposures for many workers—especially blue-collar workers. First, private workplaces might institute smoking policies even when state and local policies do not exist. A recent qualitative analysis of construction workers’ online discussions revealed that they are exposed to a variety of smoking restrictions based on job sites and contractors (Bondy and Bercovitz, 2011). This calls into question the classification of smoking policy for such blue-collar workers who are exposed to multiple workplaces over the course of the year. As we found here, there is a relatively large percentage of blue-collar workers who fall into this category. Second, most workers have to rely on their employer for enforcement of workplace smoking policies, and it is possible that there is differential enforcement of smoking policy by occupational class (Bitler et al., 2010).
Another study limitation is the use of self-reported smoking status without biochemical validation. However, the need for such validation in population-based studies has been questioned, and there is no expectation of potential differential misclassification of smoking by any of the exposure variables (Gorber et al., 2009; Murray et al., 2002). Unmeasured confounding is still a possibility, but multivariable analyses controlled for many potential confounding variables. Since this data were collected in 2002 and 2003, 26 US states have enacted workplace smoking bans and the number of home smoking bans has also increased (Mills et al., 2009). Nonetheless, these temporal trends are not expected to impact current findings regarding the strong effects of household-related risk factors on smoking and smoking cessation among blue-collar workers.
One strength of this analysis is the use of a longitudinal sample, which, though older, allowed for the establishment of temporal precedence and eliminated concerns about reverse causation with regards to smoking cessation. Also, the study was able to examine the influence of both work- and household-related risk factors for smoking in blue-collar workers, the occupational grouping with the highest prevalence of smoking. To our knowledge, this is the first examination of the influence of both work- and household-related risk factors for smoking among blue-collar workers.
Conclusion
Current efforts to decrease smoking among blue-collar workers have been focused on understanding and ameliorating work-related risk factors (Albertsen et al., 2006; Albertsen et al., 2003, 2004; Moher et al., 2005; Okechukwu et al., 2009; Sorensen et al., 2004). While this was a welcomed departure from the earlier focus on individual-level risk factors, these findings suggest that household risk factors are important drivers of smoking behaviors among blue-collar workers. Home smoking bans have tripled since the first TUS-CPS in 1992/1993; however, these bans are more likely in high education households (Mills et al., 2009). Blue-collar workers generally have lower education. Further longitudinal studies are needed to explicate possible occupational class differences in home smoking bans. Intervention studies are also needed to understand how to translate these findings into intervention activities.
Highlights.
We used longitudinal population representative data from blue-collar workers.
We examined household and work predictors of smoking status and smoking cessation.
Complete home smoking ban and partner smoking predicted smoking status.
Only complete home smoking ban predicted subsequent smoking cessation.
Presence of a child and workplace smoking policies were not significant predictors.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- Albertsen K, Borg V, Oldenburg B. A systematic review of the impact of work environment on smoking cessation, relapse and amount smoked. Prev Med. 2006;43:291–305. doi: 10.1016/j.ypmed.2006.05.001. [DOI] [PubMed] [Google Scholar]
- Albertsen K, Hannerz H, Borg V, Burr H. The effect of work environment and heavy smoking on the social inequalities in smoking cessation. Public Health. 2003;117:383–388. doi: 10.1016/S0033-3506(03)00103-3. [DOI] [PubMed] [Google Scholar]
- Albertsen K, Hannerz H, Borg V, Burr H. Work environment and smoking cessation over a five-year period. Scand J Public Health. 2004;32:164–171. doi: 10.1080/14034940310017779. [DOI] [PubMed] [Google Scholar]
- Barbeau EM, Li YI, Sorensen G, Conlan KM, Youngstrom R, Emmons K. Coverage of smoking cessation treatment by union health and welfare funds. Am J Public Health. 2001;91:1412–1415. doi: 10.2105/ajph.91.9.1412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Barbeau EM, Krieger N, Soobader MJ. Working class matters: socioeconomic disadvantage, race/ethnicity, gender, and smoking in NHIS 2000. Am J Public Health. 2004;94:269–278. doi: 10.2105/ajph.94.2.269. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bauer JE, Hyland A, Li Q, Steger C, Cummings KM. A longitudinal assessment of the impact of smoke-free worksite policies on tobacco use. Am J Public Health. 2005;95:1024–1029. doi: 10.2105/AJPH.2004.048678. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bitler M, Carpenter C, Zavodny M. Effects Of Venue-Specific State Clean Indoor Air Laws On Smoking-Related Outcomes. Health Economics. 2010;19:1425. doi: 10.1002/hec.1559. [DOI] [PubMed] [Google Scholar]
- Bondy S, Bercovitz K. Non-smoking worksites in the residential construction sector: using an online forum to study perspectives and practices. Tob Control. 2011;20:189. doi: 10.1136/tc.2010.038398. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Borland R, Yong HH, Cummings KM, Hyland A, Anderson S, Fong GT. Determinants and consequences of smoke-free homes: findings from the International Tobacco Control (ITC) Four Country Survey. Tob Control. 2006;15:42–50. doi: 10.1136/tc.2005.012492. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brownson RC, Hopkins DP, Wakefield MA. Effects of smoking restrictions in the workplace. Annu Rev of Public Health. 2002;23:333–348. doi: 10.1146/annurev.publhealth.23.100901.140551. [DOI] [PubMed] [Google Scholar]
- CDC. Cigarette smoking among adults--United States : Changes in the definition of Current Smokers. MMWR. 1994;43:342–346. [PubMed] [Google Scholar]
- Clark D. In: Algorithm for imputing missing income in the Current Population Survey. Ayadi AA, editor. Washington DC: 2011. [Google Scholar]
- CPS. Current Population Survey: Design and Methodology. U.S. Census Bureau; 2006. http://www.census.gov/prod/2006pubs/tp-2066.pdf. [Google Scholar]
- Dollar K, Homish GG, Kozlowski L, Leonard KE. Spousal and alcohol-related predictors of smoking cessation: A longitudinal study in a community sample of married couples. Am J Public Health. 2009;99:231. doi: 10.2105/AJPH.2008.140459. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Farkas AJ, Gilpin EA, Distefan JM, Pierce JP. The effects of household and workplace smoking restrictions on quitting behaviors. Tob Control. 1999;8:261–265. doi: 10.1136/tc.8.3.261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Franks MM, Pienta AM, Wray LA. It takes two: Marriage and smoking cessation in the middle years. J Aging Health. 2002;14:336–354. doi: 10.1177/08964302014003002. [DOI] [PubMed] [Google Scholar]
- Giovino G, Pederson L, Trosclair A. NIOSH. Work, Smoking and Health. Center for Disease Control and Prevention; Washington DC: 2000. The prevalence of selected cigarette smoking behaviors by occupational class in the United States; pp. 22–31. [Google Scholar]
- Gorber S, Schofield-Hurwitz S, Hardt J, Levasseur G, Tremblay M. The accuracy of self-reported smoking: A systematic review of the relationship between self-reported and cotinine-assessed smoking status. Nicotine Tob Res. 2009;11:12–24. doi: 10.1093/ntr/ntn010. [DOI] [PubMed] [Google Scholar]
- Ham DC, Przybeck T, Strickland JR, Luke DA, Bierut LJ, Evanoff BA. Occupation and workplace policies predict smoking behaviors: Analysis of national data from the Current Population Survey. J Occup Environ Med. 2011;11:1337–1345. doi: 10.1097/JOM.0b013e3182337778. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hawkins S, Berkman L. Parental home smoking policies: The protective effect of having a young child in the household. Prev Med. 2011;53:61–63. doi: 10.1016/j.ypmed.2011.05.016. [DOI] [PubMed] [Google Scholar]
- Hyland A, Higbee C, Travers M, Van Deusen A, Bansal-Travers M, King B, Cummings K. Smoke-free homes and smoking cessation and relapse in a longitudinal population of adults. Nicotine Tob Res. 2009;11:614–618. doi: 10.1093/ntr/ntp022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Manchon Walsh P, Carrillo P, Flores G, Masuet C, Morchon S, Ramon J. Effects of partner smoking status and gender on long term abstinence rates of patients receiving smoking cessation treatment. Addict Behav. 2007;32:128–136. doi: 10.1016/j.addbeh.2006.03.027. [DOI] [PubMed] [Google Scholar]
- Mills AL, Messer K, Gilpin EA, Pierce JP. The effect of smoke-free homes on adult smoking behavior: a review. Nicotine Tob Res. 2009;11:1131. doi: 10.1093/ntr/ntp122. [DOI] [PubMed] [Google Scholar]
- Moher M, Hey K, Lancaster T. Workplace interventions for smoking cessation. Cochrane Database Syst Rev. 2005:CD003440. doi: 10.1002/14651858.CD003440.pub2. [DOI] [PubMed] [Google Scholar]
- Monden CW, de Graaf ND, Kraaykamp G. How important are parents and partners for smoking cessation in adulthood? An event history analysis. Prev Med. 2003;36:197–203. doi: 10.1016/s0091-7435(02)00026-9. [DOI] [PubMed] [Google Scholar]
- Murray RP, Connett JE, Istvan JA, Nides MA, Rempel-Rossum S. Relations of cotinine and carbon monoxide to self-reported smoking in a cohort of smokers and ex-smokers followed over 5 years. Nicotine Tob Res. 2002;4:287. doi: 10.1080/14622200210141266. [DOI] [PubMed] [Google Scholar]
- Okechukwu CA, Krieger N, Sorensen G, Yi L, Barbeau EM. Massbuilt: Effectiveness of an Apprenticeship Site-Based Smoking Cessation Intervention for Unionized Building Trades Workers. Cancer Causes Control. 2009;20:887–894. doi: 10.1007/s10552-009-9324-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Okechukwu CA, Nguyen K, Hickman NJ. Partner smoking characteristics: Associations with smoking and quitting among blue-collar apprentices. Am J Ind Med. 2010;53:1102–1108. doi: 10.1002/ajim.20890. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schroeder S. Stranded in the periphery--the increasing marginalization of smokers. N Engl J Med. 2008;358:2284–2286. doi: 10.1056/NEJMe0802708. [DOI] [PubMed] [Google Scholar]
- Shopland DR, Anderson CM, Burns DM, Gerlach KK. Disparities in smoke-free workplace policies among food service workers. J Occup Environ Med. 2004;46:347. doi: 10.1097/01.jom.0000121129.78510.be. [DOI] [PubMed] [Google Scholar]
- Sorensen G, Barbeau E, Hunt M, Emmons K. Reducing social disparities in tobacco use: A social contextual model for reducing tobacco use among blue collar workers. Am J Public Health. 2004;94:230–239. doi: 10.2105/ajph.94.2.230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sorensen G, Emmons K, Stoddard AM, Linnan L, Avrunin J. Do social influences contribute to occupational differences in quitting smoking and attitudes toward quitting? Am J Health Promot. 2002;16:135–141. doi: 10.4278/0890-1171-16.3.135. [DOI] [PubMed] [Google Scholar]
- Soulakova J, Davis W, Hartman A, Gibson J. The Impact of Survey and Response Modes on Current Smoking Prevalence Estimates Using TUS-CPS: 1992–2003. Survey and Research Methods. 2009;3:123–137. [PMC free article] [PubMed] [Google Scholar]
- Syamlal G, Mazurek JM, Malarcher AM. Morbidity and Mortality Weekly Report (MMWR) Centers for Disease Control and Prevention; Atlanta: 2011. Current cigarette smoking prevalence among working adults---United States, 2004--2010; pp. 1305–1309. [Google Scholar]