Comparing Homeless Smokers to Economically Disadvantaged Domiciled Smokers

Abstract

We compared characteristics of homeless smokers and economically disadvantaged domiciled smokers (Dallas, TX; August 2011–November 2012). Although findings indicated similar smoking characteristics across samples, homeless smokers (n = 57) were exposed to more smokers and reported lower motivation to quit, lower self-efficacy for quitting, more days with mental health problems, and greater exposure to numerous stressors than domiciled smokers (n = 110). The sample groups reported similar scores on measures of affect, perceived stress, and interpersonal resources. Results may inform novel cessation interventions for homeless smokers.

Homeless individuals in the United States¹ have higher rates of disease, shorter life expectancy, and disproportionately higher health care costs than domiciled, socioeconomically disadvantaged individuals.^2–5 A primary cause of these disparities is that smoking prevalence among homeless individuals (70% of whom smoke)^6–8 is twice as high as that among those living in poverty (34.7% of whom smoke⁹). Numerous studies have indicated that many variables typical of low socioeconomic status (SES) and homelessness (e.g., low education, low income, high financial strain, unemployment) are associated with a reduced likelihood of smoking cessation.^10–13 However, few studies have specifically examined psychosocial and smoking characteristics of homeless smokers. The purpose of the current study was to compare homeless smokers with domiciled, socioeconomically disadvantaged smokers to highlight additional obstacles specific to homeless smokers that may need to be addressed during smoking cessation interventions.

METHODS

Participants included in the current analyses were recruited into 1 of 2 studies at tobacco cessation clinics in the Dallas, Texas, metropolitan area between August 2011 and November 2012. Inclusion criteria were being aged 18 years or older, a reading level higher than 6th grade (assessed via the Rapid Estimate of Adult Literacy in Medicine),¹⁴ smoking 5 cigarettes or more per day, carbon monoxide level of 8 parts per million or more at baseline, willing to quit smoking within 7 days, and ability to attend 6 weekly assessment sessions. The domiciled sample was recruited from a Dallas safety-net hospital smoking cessation clinic and the homeless sample was recruited from the smoking cessation clinic at a Dallas homeless shelter. In the homeless sample, only those who resided in the transitional shelter were eligible.

All participants completed measures of sociodemographic and smoking characteristics (Table 1). In addition, participants completed measures of subjective social status,¹⁵ nicotine dependence,¹⁶ affect and perceived stress,^17,18 mental health,^19–21 negative experiences or exposure to threat or harm,^22–25 interpersonal resources,^26–30 and self-efficacy or motivation for smoking cessation (Table 2; Castro et al., unpublished data, 2012).³¹ We conducted analyses of group differences (i.e., homeless vs domiciled samples) using χ² or analysis of variance.

TABLE 1—

Demographic and Smoking Characteristics of the Sample Participants: Dallas, TX, August 2011–November 2012

Characteristic	Homeless Smokers (n = 57), Mean (SD) or %	Domiciled Smokers (n = 110), Mean (SD) or %	P
Demographic
Age, y	50.0 (7.7)	52.6 (7.2)	.03
Gender, male	66.7	43.6	.005
Race, Black	55.4	65.5	.205
Married or partnered	35.1	55.5	.013
Education, y	12.4 (2.0)	12.1 (1.9)	.258
Reading level14	61.9 (4.9)	60.8 (5.8)	.229
Employed at least part time	5.3	17.3	.03
Family income < $12 000/y	96.3	58.3	< .001
Not insured, % yes	87.7	55.5	< .001
Community social status ladder15	4.3 (2.5)	5.6 (2.2)	.001
US social status ladder15	3.3 (2.3)	4.3 (2.0)	.005
Smoking
Cigarettes/d	18.3 (10.5)	17.0 (8.5)	.375
Years smoking	29.3 (10.7)	31.6 (9.5)	.161
Lifetime quit attempts lasting at least 24 h	4.2 (3.3)	4.1 (3.3)	.772
No. of smokers exposed to each d	42.9 (29.1)	3.5 (4.1)	< .001
Heaviness of Smoking Index16	2.9 (1.5)	3.1 (1.2)	.401

TABLE 2—

Comparison of Homeless and Domiciled Smoker Samples: Dallas, TX; August 2011–November 2012

Variable	Homeless Smokers (n = 57), Mean (SD) or %	Domiciled Smokers (n = 110), Mean (SD) or %	P
Affect and perceived stress
PANAS—Negative Affect17	18.0 (6.6)	19.5 (8.0)	.233
PANAS—Positive Affect17	31.2 (10.5)	29.2 (9.4)	.22
Perceived Stress Scale18	6.1 (3.4)	6.3 (3.3)	.618
Mental health
PHQ Alcohol Dependence20	17.5	17.3	.965
Depression diagnosis history	80.7	50.9	<.001
BRFSS no. of days with mental health problems21	11.5 (10.5)	8.0 (9.8)	.035
CES–D,9	15.7 (10.3)	16.0 (11.1)	.879
Negative experiences and exposure to threat or harm
Detroit Discrimination Scale25	31.5 (13.4)	19.5 (10.0)	<.001
Urban Life Stress Scale24	48.4 (11.1)	43.2 (11.8)	.006
Social Cohesion and Trust Scale23	14.2 (2.3)	15.6 (2.7)	.001
Fear Scale22	1.7 (0.8)	1.5 (0.6)	.027
Mistrust Scale22	1.9 (0.7)	1.6 (0.6)	.002
Reserve capacity
Loneliness28	5.3 (2.0)	5.2 (1.9)	.713
General Self-Efficacy Scale26	34.1 (6.1)	33.8 (5.8)	.738
Revised Life Orientation Test30	13.8 (4.4)	14.5 (4.1)	.362
ISEL27
Appraisal scale	12.4 (3.1)	12.5 (2.9)	.87
Belonging scale	12.5 (2.9)	12.2 (3.1)	.523
Tangible support scale	12.3 (3.0)	12.3 (3.0)	.077
Lubben Social Network Scale29	12.1 (7.3)	13.7 (6.3)	.122
Self-efficacy/motivation for quitting
Self-efficacy and motivation for quitting: TSAMS Motivation for Quittinga	20.9 (4.4)	22.2 (3.9)	.047
Self-efficacy for quitting31
Positive affect and social situations	2.2 (0.9)	2.7 (0.8)	<.001
Negative affect situations	2.1 (0.9)	2.3 (0.9)	.069
Habit and craving situations	2.3 (0.9)	2.8 (0.8)	.001

Note. BRFSS = Behavioral Risk Factor Surveillance System; CES-D = Center for Epidemiological Studies—Depression scale; ISEL = Interpersonal Support Evaluation List; PANAS = Positive and Negative Affect Schedule; PHQ = Patient Health Questionnaire; TSAMS = Texas Smoking Abstinence Motivation Scale.

^aCastro et al., unpublished data, 2012.

RESULTS

Homeless participants (n = 57) were more likely to be male, younger, single, uninsured, and unemployed than domiciled participants (n = 110; Table 1). In addition, domiciled smokers placed themselves on higher rungs of the community and US subjective social status ladders.¹⁵ Although smoking characteristics were similar across samples, homeless smokers reported daily exposure to substantially more smokers than did domiciled smokers (Table 1).

The homeless and domiciled samples were similar on measures of recent affect, current symptoms of depression, perceived stress, and alcohol abuse (Table 2). However, the homeless sample reported more recent days with mental health problems, greater depression diagnosis prevalence, higher levels of discrimination, higher scores on the Urban Life Stress Scale, more fear, more mistrust of others, and lower social cohesion and trust than did the domiciled sample (Table 2). The sample groups scored similarly on measures of loneliness, general self-efficacy, dispositional optimism, social support, and social isolation (Table 2). Finally, the homeless sample was less motivated to quit smoking and reported lower confidence in maintaining abstinence than did the domiciled sample (Table 2).

DISCUSSION

Study results indicate that, compared with low-SES domiciled smokers, homeless smokers may have more mental health problems, be surrounded by more smokers, be exposed to substantially more stressors and discrimination, and have lower motivation and self-efficacy for quitting. Each of these variables may play a role in the extremely high prevalence of smoking among homeless individuals and the low smoking cessation rate in this population. These differences may suggest that homeless smokers seeking treatment may not respond to cessation interventions specifically developed for domiciled low-SES smokers. Study findings also demonstrate that homeless smokers possess psychosocial resources comparable to those of socioeconomically disadvantaged domiciled smokers. Thus, homeless individuals may have effective coping mechanisms that may be used to increase successful smoking cessation if tapped in novel smoking cessation interventions.

Findings highlight many variables that may be targeted in future cessation programs specifically tailored to the needs of homeless smokers, and results may be used to support changes in tobacco use policies at shelters. For example, creating smoke-free zones or disallowing smoking altogether on shelter grounds may reduce continued exposure to other smokers, thus addressing a known barrier to successful smoking cessation.^32,33 This policy is consistent with recommendations from the Break Free Alliance Expert Panel.³⁴

Study limitations include the use of small regional samples seeking cessation treatment, which may limit generalizability and analysis power, reliance on self-report, and our comparison of 2 different populations of smokers. Although these limitations are significant, we believe that this type of comparison is warranted because of the dearth of knowledge regarding the potential causes for the high smoking prevalence and the low smoking cessation rate among homeless individuals. Novel smoking cessation interventions that address specific barriers experienced by homeless smokers should be developed. These tailored interventions may have an enormous impact on the health and life expectancy of this underserved and vulnerable population.