LJAS_A_ XXXXXXXXXX
Full Terms & Conditions of access and use can be found at
https:
www.tandfonline.com/action/journalInformation?journalCode=ijas20
Journal of Asthma
ISSN: XXXXXXXXXXPrint XXXXXXXXXXOnline) Journal homepage: https:
www.tandfonline.com/loi/ijas20
Co
elates of Smoking During an Economic
Recession Among Parents of Children with Asthma
Tracy L. Jackson, Annie Gjelsvik, Aris Ga
o & Deborah N. Pearlman
To cite this article: Tracy L. Jackson, Annie Gjelsvik, Aris Ga
o & Deborah N. Pearlman (2013)
Co
elates of Smoking During an Economic Recession Among Parents of Children with Asthma,
Journal of Asthma, 50:5, XXXXXXXXXX, DOI: XXXXXXXXXX/ XXXXXXXXXX
To link to this article: https:
doi.org/10.3109/ XXXXXXXXXX
Published online: 01 May 2013.
Submit your article to this journal
Article views: 156
View related articles
Citing articles: 1 View citing articles
https:
www.tandfonline.com/action/journalInformation?journalCode=ijas20
https:
www.tandfonline.com/loi/ijas20
https:
www.tandfonline.com/action/showCitFormats?doi=10.3109/ XXXXXXXXXX
https:
doi.org/10.3109/ XXXXXXXXXX
https:
www.tandfonline.com/action/authorSubmission?journalCode=ijas20&show=instructions
https:
www.tandfonline.com/action/authorSubmission?journalCode=ijas20&show=instructions
https:
www.tandfonline.com/doi/mlt/10.3109/ XXXXXXXXXX
https:
www.tandfonline.com/doi/mlt/10.3109/ XXXXXXXXXX
https:
www.tandfonline.com/doi/citedby/10.3109/ XXXXXXXXXX#tabModule
https:
www.tandfonline.com/doi/citedby/10.3109/ XXXXXXXXXX#tabModule
Journal of Asthma, 2013; 50(5): 457–462
Copyright © 2013 Informa Healthcare USA, Inc.
ISSN: XXXXXXXXXXprint/ XXXXXXXXXXonline
DOI: XXXXXXXXXX/ XXXXXXXXXX
ENVIRONMENTAL DETERMINANTS
Co
elates of Smoking During an Economic Recession Among Parents of
Children with Asthma
TRACY L. JACKSON, MPH,1,* ANNIE GJELSVIK, PH.D.,1 ARIS GARRO, MD, MPH,2
AND DEBORAH N. PEARLMAN, PH.D.1
1Department of Epidemiology, Brown University Wa
en Alpert Medical School, 121 South Main Street S 121-2, 02912 Providence,
RI, USA.
2Department of Pediatrics and Emergency Medicine, Brown University Wa
en Alpert Medical School, 125 Whipple Street, 3rd floor,
UEMF Suite, 02908 Providence, RI, USA.
Objective.We describe the co
elates of smoking among parents who have a child with asthma and examine whether the co
elates changed from
2008 to 2010, when the United States experienced a severe recession and a sharp increase in unemployment, a stressor that could influence smoking
ehavior.Methods. Data are from the 2008 and 2010 Behavioral Risk Factor Surveillance System (BRFSS), a nationally representative survey of
U.S. adults age 18 and older. Separate logistic regressions estimated the association between unemployment and smoking in 2008 and 2010,
adjusting for sociodemographic and other characteristics of parents of a child with asthma. Results. Being unemployedwas a significant predictor of
smoking in 2010 (AOR ¼ 1.80; 95% CI: 1.24–2.61), but was not a significant predictor in 2008 (AOR ¼ 1.26, 95% CI: 0.82–1.95). One central
component of well-being, as measured by being dissatisfied with one’s life, was significantly associated with parental smoking in 2010
(AOR ¼ 2.06, 95% CI: 1.00–4.27), but not in 2008 (AOR ¼ 1.62, 95% CI: 0.85–3.11). Several covariates had similar associations with parental
smoking in both survey years, including low education, not being cu
ently ma
ied, not having health insurance, and binge drinking. Conclusions.
Our results support the idea that during hard economic times unemployment and related stressors may be strong determinants of parental smoking
when a child in the home has asthma. Given that the BFRSS is a cross-sectional survey, definitive conclusions cannot be drawn regarding the causal
pathway connecting unemployment, global well-being, and parental smoking.
Keywords economic recession, education, employment, mental health, smoking, socioeconomic status, stresso
The adverse health effects of exposure to environmental
tobacco smoke (ETS) for children with asthma are well
known (1–9). Exposure to ETS in children with asthma has
een associated with poorer quality of life, reduced lung
function, increased asthma-related emergency room visits
and hospitalizations, and prolonged recovery after an
asthma attack (2, 4, 8, 9). Because children with asthma
are adversely affected by ETS exposure, one might expect
that parents of children with asthma would be less likely to
smoke, yet this is not the case. Parents of children with
asthma continue to smoke at levels comparable to the
general population of smokers (2, 8). To date, no studies
have focused on whether the factors associated with cigar-
ette smoking are unique to parents of a child with asthma as
compared to the general U.S. adult population.
Studies examining risk factors for smoking in the gen-
eral adult U.S. population have found that smoking is
associated with low socioeconomic status (SES), depres-
sive symptoms, and parental and peer smoking (10–16).
Low SES, either measured through educational attainment
or income, is the strongest predictor (10–16). Anothe
strong predictor of smoking is poor mental health, as
smoking has been found to occur at higher rates among
individuals with anxiety and depression (17–24).
Unemployment has also been associated with increased
smoking rates in both cross-sectional (25) and longitudinal
studies (26, 27). This is consistent with previous literature
suggesting that unemployment itself can be a stressor and
may serve as a proxy measure for otherwise unmeasured
social and economic strain. However, unemployment also
has been associated with decreased cigarette consumption
as smokers seek to balance their budgets when faced with
job loss (28).
In the present study, we examined mental health mea-
sures, employment status, and SES in relation to parental
smoking. We also were interested in investigating whethe
these factors differed at two points in time—in 2008 as the
U.S. economy entered a recession, and in 2010 after the U.S.
went through one of its longest and most severe recessions
since the Great Depression of the 1930s (29).
METHODS
Study Design
Data are from the 2008 and 2010 Behavioral Risk Facto
Surveillance System (BRFSS). The BRFSS is a state-based
andom-digit-dialed telephone survey of the noninstitutiona-
lized U.S. civilian population aged �18 years (30). The
*Co
esponding author: Tracy L. Jackson, MPH, Department of Epidemiology,
Brown University Wa
en Alpert Medical School, 121 South Main Street S
121-2, 02912 Providence, RI, USA; Tel: XXXXXXXXXX; E-mail:
Tracy_Jackson@
own.edu
457
sample obtained in 2010 was independent of the sample in
2008 as the BRFSS are serial cross-sectional surveys.
The BRFSS Childhood Asthma optional module was
utilized by 37 states/te
itories in 2008 and 35 states/te
i-
tories in 2010. Two asthma prevalence measures are asked
about a randomly selected child in the household aged 17
years or younger. Cu
ent asthma is determined from the
esponse, “Does the child still have [diagnosed] asthma?”
Eligibility study criteria for the study included the fol-
lowing: respondent was a parent/guardian of a child who
lived in their home, the child in the home (one child was
andomly selected if the parent reported having more than
one) had cu
ent doctor-diagnosed asthma and was
etween 2 and 17 years of age, and the parent had valid
information on smoking and employment status.
The total number of BRFSS respondents in 2008 was
414,509, of which 69,814 were parents/guardians of a child
who lived in their home. Of this sample of parents/guardians,
5811 reported that their randomly selected child had cu
ent
doctor-diagnosed asthma (8.3%), 5426 of which were chil-
dren 2–17 years of age. After respondents with missing data
on smoking status and employment status were excluded, the
final analytic sample for 2008 was 4244.
The total number of BRFSS respondents in 2010 was
451,075, of which 78,619 were parents/guardians of a
child who lived in their home. Of this sample of parents
guardians, 5155 reported that their randomly selected child
had cu
ent doctor-diagnosed asthma (6.6%), 4818 of
which were children 2–17 years of age. The final analytic
sample for 2010 was 3601, after excluding respondents
with missing data on smoking and employment status.
Measures
Cu
ent smoking was measured by whether the parent o
guardian of a child in the home with asthma cu
ently
smoked (yes/no). SES was measured according to parents’
educational level (<12th grade, high school graduate,
some college or college graduate), rather than household
income. Income may be problematic as an indicator of
SES, since income can fluctuate over the life course, unlike
education, which is more stable beyond early adulthood
(31). Furthermore, BRFSS response rates tend to be highe
for educational attainment (,99%), unlike household
income, which typically has lower response rates
(,93%) and consequently higher response bias. In addi-
tion to educational status, health insurance was included as
a proxy for access to care and potential tobacco cessation
counseling. Marital status was defined as either ma
ied o
unma
ied. We also included a measure of binge drinking,
given that binge drinking is the most common pattern of
excessive alcohol use in the United States (32), and alcohol
and tobacco use often co-occur in the general adult
U.S. population (33). Binge drinkers were defined as
men who reported consuming five or more drinks o
females who reported consuming four or more drinks in
one sitting on one or more occasions in the past 30 days.
Psychosocial stress was assessed through self-reported
global life satisfaction, receipt of emotional and social
support, and frequency of mental distress. The Centers fo
Disease Control and Prevention (CDC) assesses global well-
eing with two valid measures—life satisfaction and emo-
tional support received. The CDC also supports frequency of
mental distress as a measure of health-related quality of life
with acceptable criterion validity and test–retest reliability
(34–36). Questions measuring psychosocial stress in the
2008 and 2010 BRFSS included the following: “In general,
how satisfied are you with your life?” (very satisfied, satis-
fied, dissatisfied/very dissatisfied), “Howoften do you get the
social and emotional support you need?” (always/often,
sometimes
arely/never), and “Now thinking about you
mental health, which includes stress, depression, and pro-
lems with emotions, for how many days during the past
30 days was your mental health not good?” (0 days, 1–13
days, 14 or more days). CDC measures the prevalence of
poor mental health based on number of mentally unhealthy
days. Respondents who report�14 mentally unhealthy days
over the past month are defined as having frequent mental
distress (34, 35, 37). The 14-day minimum period also was
used because clinicians and researchers often use this period
as a marker for clinical depression. Adopting this precedent
ensured comparability with other studies looking at poo
mental health based on frequent mental distress (�14 days)
in relation to health outcomes.
Unemployment can be a significant source of stress.
Financial strain, stigma of job loss, and uncertainty about
the future may be particularly prevalent during an eco-
nomic recession. In the cu
ent study, employment status
was defined as either employed or unemployed.
Respondents who were homemakers, retired, students, o
unable to work were excluded from both sample years.
Statistical Analysis
Sampling weights that co
ect for unequal probabilities of
sample selection and adjust for nonresponse and telephone
noncoverage were applied to the BRFSS adult samples to
obtain a nationally representative sample of parents with a
child in the home with asthma (38).
Separate logistic regressions estimated the associations
etween unemployment and smoking in 2008 and 2010,
adjusting for the same covariates in each model. An inter-
action term between cu
ent smoking and survey yea
assessed statistically significant differences in unemploy-
ment by time period. Since the interaction term was not
significant in the logistic regressions in either survey year,
it was dropped from each model.
RESULTS
The rates of smoking among parents of children with
asthma were similar across the two study years. In 2008,
20.0% of parents were smokers, compared to 17.8% in
2010 (Table 1). There were no significant differences
etween the two sample populations by socio-
demographic and health characteristics or in rates of unem-
ployment. The only significant difference between the two
periods was the substantial increase in the proportion of
458 T. L. JACKSON ET AL.
parents without health insurance, from 12.0% in 2008 to
16.6% in 2010 (p < .01; Table 1). There was no significant
difference in the proportion of children with cu
ent
asthma by age, sex, or race/ethnicity across the two study
periods (data not shown). The average age of children in
2008 was 10.3 years (95% confidence interval (CI): 10.1–
10.5) and 10.2 years in XXXXXXXXXX% CI: 9.9–10.5).
In 2008, unemployment was not significantly asso-
ciated with parental smoking [adjusted odds ratios
(AOR) ¼ 1.26, 95% CI: 0.82–1.95], but in 2010 this
association was significant (AOR ¼ 1.80, 95% CI: 1.24–
2.61). Dissatisfaction with one’s life was not a significant
predictor of smoking in 2008. In 2010, parents who
eported being either satisfied with life or dissatisfied
with life were more likely to smoke compared to those
who reported being very satisfied with life (Table 2).
Several covariates had similar associations with paren-
tal smoking in both survey years. The AOR for smoking
among those with less than a high school level of education
were XXXXXXXXXX% CI: 2.46–6.84) in 2008 and XXXXXXXXXX% CI:
2.82–8.22) in 2010. Reporting frequent mental distress (14
or more days of poor mental health in the last month), not
eing cu
ently ma
ied, and being a binge drinker also
were positively associated with smoking in both study
years. Hispanic ethnicity was negatively associated with
parental smoking in both 2008 and 2010. There were no
significant associations between smoking and not getting
social and emotional support needed in either study year.
DISCUSSION
The results of this study confirm previous findings in the
literature regarding the relationship between unemployment
and smoking in cross-sectional and longitudinal studies
where the unemployed were more likely to smoke than the
employed or the general population of adults (25–27, 39–42).
The study adds to the previous literature by showing that a
TABLE 2.—Adjusted odds ratio in logistic regressionmodel for predictors of
cu
ent smoking