Cessation and reduction in smoking behavior: impact of creating a smoke-free home on smokers

Abstract

The aim of this study was to assess the effect of a creating a smoke-free home (SFH) on cessation and reduction of cigarette smoking on low-income smokers. This secondary data analysis uses data from study participants who were originally recruited through 2-1-1 information and referral call centers in Atlanta (Georgia, 2013), North Carolina (2014) and the Texas Gulf Coast (2015) across three randomized controlled trials testing an intervention aimed at creating SFHs, pooling data from 941 smokers. Participants who reported adopting a SFH were more likely to report quitting smoking than those who did not adopt a SFH. This was true at 3-month follow-up and even more pronounced at 6-month follow-up and persisted when considering only those who consistently reported no smoking at 3 and 6 months. Among those who did not stop smoking, the number of cigarettes per day declined significantly more and quit attempts were more frequent for those who created a SFH compared with those who did not. Findings suggest that creating a SFH facilitates cessation, reduces cigarette consumption and increases quit attempts. Future studies should assess the long-term impact of SFHs on sustaining cessation.

Introduction

Associations between smoke-free homes (SFHs) and smoking behaviors have been well established in cross-sectional and observational longitudinal studies [1, 2]. However, no studies have investigated the impact of a household changing its rules to ban smoking in the home.

A 2009 review by Mills et al. [1] showed fairly consistently that SFHs were associated with quit attempts, increased quit duration, and reduced relapse in both cross-sectional and longitudinal studies. In longitudinal studies those with a SFH were 50% more likely to quit smoking compared with those who allow smoking in the home. A more recent study [2] found that smokers who lived in homes with a smoking ban were 4.8 times more likely to quit smoking successfully than those living in homes without restrictions. Smokers living in a SFH also smoked fewer CPD (10.8 versus 16.5) [2]. Beyond cessation, Mills et al. [1] found a positive association between SFHs and reduction in smoking behavior in 13 out of 14 studies (both cross-sectional and longitudinal). Furthermore, studies a larger decrease in cigarettes smoked per day (CPD) in smokers living in SFHs over time [3–6].

Thus, previous observational research suggests that household smoking bans are related to (i) cessation and (ii) reduction in CPD. This study pools data from low-income participants of three randomized controlled trials (RCTs) to assess the impact of a newly created SFH on smokers.

Materials and methods

Participants

Participants who allowed some smoking in their home were identified through 2-1-1 referral call centers in Atlanta (GA) [7], North Carolina [8] and Texas Gulf Coast [9] and recruited into an RCT testing a brief intervention—consisting of three mailings and a 20-min coaching call—aimed at creating a SFH. Further eligibility criteria were (i) having at least one smoker and one non-smoker in the household, (ii) speaking English and (iii) being at least 18-years old. Across trials, the intervention was successful in creating SFHs [7–9]. Notably, the intervention did not emphasize smoking cessation and only provided quit line contact information in one of the mailings and during the coaching call when relevant and requested. Many participants in the control groups also created a SFH, which we attribute to heightened consciousness of SFHs resulting from the studies’ repeated extensive assessments.

This study pools data from the 941 smokers (62.5% of all participants) from intervention and control groups across the three RCTs who provided data on their SFH status and their smoking behaviors 3 and 6 months post-baseline.

Measures

SFH status was assessed through self-report at all three time-points (and validated through nicotine monitors at 3 months follow-up in the Atlanta trial) [7]. Smoking behaviors included smoking status, CPD and quit attempts in the past 3 months, and nicotine dependence [6, 10]. Demographic information on the participant’s ethnicity/race, age, gender, educational level, marital status, household income, household composition (including number of adult smokers and non-smokers in the household), and employment status were collected at baseline. Poverty level was calculated from household income and composition data using procedures from the Centers for Disease Control and Prevention, accounting for year of reporting and dichotomized into below/above federal poverty level.

Data analysis

After assessing univariate distributions and bivariate associations between outcomes and SFH ban status for all participants, logistic regression models were used to assess effects of creating a SFH on smoking status (cessation among baseline smokers) and smoking behaviors of those who continued to smoke. Regression analyses were first conducted with only the SFH ban status at follow-up as a predictor (unadjusted results) and then adjusted for the following covariates: group assignment (intervention or control group), trial site, age, gender, race (African American or not), household composition (being the only smoker in the house or not, living with a significant other, work status, having children under 18, under 5 and under 1 year of age in the household), and smoker characteristics (CPD at baseline, confidence to quit at baseline, and level of nicotine addiction). All analyses were conducted using SAS 9.4.

Results

Demographic characteristics, household composition, and smoking behaviors have been reported for each of the trials [7–9] and compared across trials. Overall, the smokers in this study were mostly female (85.0%), African American (73.2%), and living below the federal poverty level (80.9%). A large majority reported daily smoking (85.9%), and the mean number of CPD was 12.9 (SD = 8.70). At baseline, 56.5% had a partial smoking ban in the home, with the remainder having no rules or restrictions regarding smoking in the home.

Table I shows that at 3-month follow-up, more participants who reported a smoking ban had stopped smoking [15.6 versus 4.5%, aOR = 3.90, 95% CI = (2.074, 7.329), P < 0.0001]. At 6-month follow-up, the difference was more pronounced, with 26.1% of those with a household smoking ban reporting they had ceased smoking compared with 5.1% of those who did not create a SFH [aOR = 6.56, 95% CI = (4.123, 10.422), P < 0.0001]. When considering only those who consistently reported no smoking at 3 and 6 months, 9.6% (compared with 1.9%) of those who had a SFH had stopped smoking [aOR = 5.43, 95% CI = (2.547, 11.562), P < 0.0001].

Table I.

Impact of change in SFH ban status on smoking status and smoking behaviors in continued smokers from pooled trial data: Atlanta 2013, North Carolina 2014 and Texas Gulf Region 2015

	Three-month follow-up					Six-month follow-up
	Full ban		No full ban			Full ban		No full ban
Cessation
Smoking status	n	%	n	%	P-value	n	%	n	%	P-value
Former smokers	36	15.58	30	4.48		80	26.14	27	5.12
Continued smokers	195	84.42	639	95.52	<0.0001	226	73.86	500	94.88	<0.0001
	aOR	95% CI				aOR	95% CI
Multivariable logistic regression	3.9	2.074	7.329		<.0001	6.02	3.47	10.47		<0.0001
Smoking behaviors in continued smokers
Cigarettes per day	Mean	SD	Mean	SD	P-value	Mean	SD	Mean	SD	P-value
Number of cigarettes per day	7.9	6.91	11.7	8	<0.0001	7.3	6.06	11.1	7.94	<0.0001
Change in number of cigs per day	−4.2	7.52	−1.5	6.59	<0.0001	−4.9	7.71	−2.5	6.59	<0.0001
	Mean	SE				adjusted Mean	SE
Multivariable regression	−2.0	0.61			0.001	−2.5	0.63			<0.0001
Quit attempts (past 3 months)
‘All continued smokers’	n	%	n	%		n	%	n	%
Any quit attempt (Yes reported)	50	56.18	189	42.57	0.02	75	61.98	163	44.54	0.001
	aOR	95% CI				aOR	95% CI
Multivariable logistic regression	2.86	1.457	5.626		0.002	2.33	1.269;	4.266		0.006
‘Smokers without quit attempts at baseline’	n	%	n	%		n	%	n	%
Any quit attempt (Yes reported)—‘activation’	20	40	52	22.22	0.009	30	48.39	55	26.96	0.002
	aOR	95% CI				aOR	95% CI
Multivariable logistic regression	3.09	1.394;	6.848		0.006	3.00	1.426	6.322		0.004

Among participants who did not stop smoking, the number of CPD was significantly reduced among those who had created a SFH at 3 months (M = −4.2, SD = 7.52 versus M = −1.5, SD = 6.59 P < 0.0001) and 6 months (M = −4.9, SD = 7.71 versus M = −2.5, SD = 6.59, P < 0.0001) compared with baseline. Furthermore, creating a SFH was related to an increased number of quit attempts and ‘activated’ some smokers who had indicated no quit attempts at baseline to attempt quitting.

Discussion

Findings from this study advance the growing body of evidence linking SFHs to cessation by providing the first data from prospective longitudinal studies of SFH adoption. Although observational research has suggested a link between household smoking bans and cessation and smoking behaviors [2–4, 6], this study allows us to draw stronger conclusions. Our data show that smokers who create a SFH are more likely to quit smoking. This is in line with research that has consistently shown that smoke-free environments lead to reduction in smoking prevalence, e.g. in workplaces [11]. Furthermore, our results also indicate a reduction in smoking for those with a SFH who continue to smoke. Similar effects have been reported after implementing other smoke-free policies and so it is not surprising that reducing the spaces in which smoking is allowed reduces cigarette consumption.

This study has multiple strengths including the rigor of the studies, replication of results across three contexts, and its statistical power due to pooled data. Limitations are the self-report of smoking behaviors and the home-smoking ban. However, the self-report of home-smoking bans was validated using a bogus pipeline and nicotine monitors [7].

Public health implications

Our findings provide yet another reason for tobacco control efforts to emphasize SFHs. It is possible that SFH interventions are viewed by smokers as less threatening than cessation interventions, yet with collateral benefits on cessation that accomplish the same goal. For example, some smokers receiving an overtly cessation-focused intervention might marshal their cognitive resources against messages that are clearly intended to get them to quit smoking. In contrast, SFH interventions might not evoke the same defensive reaction or counter-arguing [12]. Rather, they could help create conditions that make cessation easier or lead smokers who were not about to quit to consider it.

Funding

This work was supported by the National Cancer Institute’s State and Community Tobacco Control Research Initiative (grant number U01CA154282). The trials are registered with ClinicalTrials.gov number: NCT01625468.

Conflict of interest statement

None declared.