Jahnel et al.

BMC Public Health (2019) 19:1284

https://doi.org/10.1186/s12889-019-7631-2

RESEARCH ARTICLE Open Access

Daily stress as link between disadvantage

and smoking: an ecological momentary
assessment study
Tina Jahnel1, Stuart G. Ferguson1, Saul Shiffman2 and Benjamin Schüz1,3*

Abstract
Background: There is a well-established social gradient in smoking, but little is known about the underlying
behavioral mechanisms. Here, we take a social-ecological perspective by examining daily stress experience as a
process linking social disadvantage to smoking behavior.
Method: A sample of 194 daily smokers, who were not attempting to quit, recorded their smoking and information
about situational and contextual factors for three weeks using an electronic diary. We tested whether
socioeconomic disadvantage (indicated by educational attainment, income and race) exerts indirect effects on
smoking (cigarettes per day) via daily stress. Stress experience was assessed at the end of each day using Ecological
Momentary Assessment methods. Data were analyzed using random effects regression with a lower-level (2-1-1)
mediation model.
Results: On the within-person level lower educated and African American smokers reported significantly more daily
stress across the monitoring period, which in turn was associated with more smoking. This resulted in a small
significant indirect effect of daily stress experience on social disadvantage and smoking when using education and
race as indicator for social disadvantage. No such effects were found when for income as indicator for social
disadvantage.
Conclusion: These findings highlight the potential for future studies investigating behavioral mechanisms
underlying smoking disparities. Such information would aid in the development and improvement of interventions
to reduce social inequality in smoking rates and smoking rates in general.
Keywords: Social disadvantage, Smoking, EMA, Daily stress

Background smoking [7]. However, to date, little is known about the

It is well documented that there is a substantial social processes through which social disadvantage translates
gradient in smoking behavior: among smokers, those into smoking behavior.
who are more socially disadvantaged also smoke more Harwood and colleagues [8] posited that psychosocial
heavily [1–3] and suffer worse health implications as a factors, such as perceived stress mediate the relationship
result [4, 5]. In addition, the knowledge of, access to, between socioeconomic status (SES) and smoking. The
and use of healthcare resources vary widely across popu- unequal distribution of socioeconomic resources and the
lation subgroups [6]. The concentration of smoking resulting socioeconomic disadvantages imply greater
among socially disadvantaged groups is based on com- stress experience for those with less access to resources
plex individual and social processes and are fundamental [9, 10]. In addition, people who experience socioeco-
to understanding the persistent unequal distribution of nomic disadvantage also experience more stressful
events and have fewer resources to successfully cope
* Correspondence: Benjamin.schuez@uni-bremen.de with this stress [10–13]. This, in turn, might lead to in-
1
College of Health & Medicine, University of Tasmania, Hobart, Australia
3
Institute for Public Health and Nursing Science, University of Bremen,
creased smoking as a maladaptive way of coping with
Bremen, Germany
Full list of author information is available at the end of the article

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Jahnel et al. BMC Public Health (2019) 19:1284 Page 2 of 8

the stress [14]; essentially, that some people may engage the potential of daily stress experience as a factor linking
in smoking as a means of coping with stress [15, 16]. socioeconomic disadvantage and smoking has not been
Providing enhanced cessation support to smokers ex- formally tested yet.
periencing stress [17, 18] could therefore help to reduce Examining the role of daily stress on the social gradi-
smoking prevalence, particularly among socially disad- ent in smoking has both theoretical implications and
vantaged smokers [10]. This idea implies a mediational may guide and inform the design of interventions tar-
relationship: Smoking is linked indirectly (mediation) to geted in particular at smokers with low socioeconomic
socioeconomic disadvantage through the experience of status through reducing individual stress experience.
everyday stress, i.e., people would be assumed to smoke The goals of these secondary analyses are to investigate
more as a response to experiencing higher levels of whether social disadvantage, assessed through the un-
stress in their everyday lives through socioeconomic dis- equal distribution of socioeconomic indicators, induces
advantage. While this assumption is both plausible and different levels of daily stress smokers experience and
supported by research on either ends (disadvantage leads whether this is associated with smoking. It is hypothe-
to higher levels of daily stress [19]; and higher levels of sized that perceived daily stress indirectly links indica-
daily stress are associated with higher levels of smoking tors of social disadvantage and smoking in a way that
[20], relatively few studies have investigated the role of more socially disadvantaged smokers perceive more daily
daily stress in the social gradient in smoking. stress, and that greater stress perception is in turn asso-
Most previous studies examining the role of stress on ciated with more cigarettes smoked.
the social gradient in smoking have operationalized
stress as major live events or chronic stress [12, 21]. For Methods
example, Mulder and colleagues reported that multiple Overview
chronic stressors (e.g., financial problems, low perceived To explore our research question, we conducted second-
life control and lack social support) were found to be ary analyses using data from a larger real-world study
directly related to smoking and partially mediated the looking at smoking patterns [25, 26]. The study used
relationship between educational level and smoking be- Ecological Momentary Assessment (EMA) methods to
havior [10]. In essence: Socially disadvantaged individ- assess smoker’s daily stress experiences and smoking as
uals experience more stressful situations which is they went about their day-to-day lives. This data collec-
associated with higher smoking rates, compared to less tion strategy allows us to examine fluctuations in daily
disadvantaged individuals [10, 13]. stress and smoking, while taking individual differences
This view, however, is based on the idea that there are (e.g. education and income) into account. The study was
differences between individuals in how much chronic or approved by the Institutional Review Board at the Uni-
acute stress they experience, and that these differences versity of Pittsburgh (REN16120111 / IRB0606147).
translate into stable differences in smoking rates be-
tween individuals (between-subject levels of analysis [22, Participants
23];). Such analyses cannot explore whether smokers Details on the recruitment methods have been reported
smoke more cigarettes in situations or on days when elsewhere [25, 26]. Briefly, between November 2007 and
they are more stressed (within-subject level of analysis January 2010, 194 daily smokers were recruited in the
[24];). Where studies using between-subject level of ana- Pittsburgh area. The study was introduced to the partici-
lysis aim to compare groups of people, ascertaining pants as a naturalistic study of smoking patterns. Eligi-
whether situations in which individuals are stressed cues bility criteria included being a smoker for at least three
smoking calls for within-subject level of analysis. A pre- years, report smoking every day (defined as smoking 5 to
vious study examining the link between the experience 30 cigarettes per day [CPD]) for at least the last 3
of daily stress and smoking for example showed that in- months, not planning to quit within the next month,
dividuals reported smoking significantly more in situa- and be at least 21 years old. The presence of diagnosed
tions when they experienced more stress [20]. In mental health conditions were not listed in the exclusion
addition, daily stress captures day-to-day experiences criteria. Written informed consent was obtained, using a
that potentially offer insight into the circumstances that consent form approved by the University of Pittsburgh
may facilitate or preserve social inequalities in health Institutional Review Board (IRB).
[19] and therefore might capture some of the variance in
the association between aspects of social disadvantage Procedure
and smoking. Grzywacz et al. [19], for example, found After reviewing and signing a consent form, participants
that exposure to daily stress was status related: severity completed a baseline questionnaire including informa-
of stressful situations for lower SES individuals was tion about basic demographic characteristics (age, gen-
higher than for individuals with higher SES. However, der, educational attainment, income, racial background).
Jahnel et al. BMC Public Health (2019) 19:1284 Page 3 of 8

The EMA monitoring procedures have been reported services that reduce stress or minimize the sources of
previously [25, 26]. Data collection took place using stress [14]. Lower income therefore can also represent a
hand-held electronic diaries (EDs). Participants were proxy for social disadvantage. Income was operationalized
supplied with an ED running data collection software as the annual household income in US dollar. To ensure
programmed specifically for the study. Participants re- balanced groups sizes, income was dichotomized in low (≤
ceived hands-on individual training to monitor their $14,999; coded as higher social disadvantage = 1) and high
smoking, activities, feelings and social setting in real- income (≥ $15,000; coded as lower social disadvantage = 0).
time. Participants were asked to carry the ED at all times Finally, we also included racial background as another
during the waking day. Every night, participants were indicator for social disadvantage. Racial background was
also asked to complete an evening report and state the dichotomized in “African American” (AA; coded as
severity of their perceived stress on any given day. Add- higher social disadvantage = 1) and “Other” (“Caucasian”
itionally, participants were given the opportunity to use and “other”) as most previous studies that address the
the evening report to record extra cigarettes they had relationship between stress and smoking in various racial
missed to report in real-time. Participants monitored groups have been conducted among African Americans
their smoking over 21 days (M = 22.5, SD = 4.12). [30, 31]. In particular in the US, individuals with African
American background experience substantially more
Measures everyday discrimination, which in turn results in higher
To identify those who are most and least socially disad- stress levels, both subjectively experienced as well as
vantaged, individuals were classified according to their verified by biomarkers [32].
educational attainment, income, and racial background The primary outcome—CPD—was operationalized by
(all indicators were assessed during the baseline visit). daily cigarette counts assessed through ED logs and daily
The statuses people occupy as adults and the resulting retrospective reports assessed at the end of the day [33].
social disadvantages/ advantages can be the consequence The variable assumed to mediate between social disad-
of many factors. Although these factors are correlated vantage and smoking (daily stress) was operationalized
with one another, they are only partially overlapping and as the perceived daily stress and was reported by the
exert their effects on behavior differentially and there- participant during the evening reports (“Since last Even-
fore can provide opportunities or barriers towards be- ing Report: Felt nervous/stressed?” with possible answers
havior [27]. Hence, we need to examine different ranging from 0 to 100). This item has reasonable face
socioeconomic indicators to determine their association validity as it asks directly about stress perception.
with the experience of stress and smoking.
Higher education, for example, indicates easier access Analysis
to and better processing capability of health-related in- The aim of this project was the examination of indirect
formation [28]. Education is usually established early in effects of social disadvantage indicated by education, in-
life, is generally stable throughout the life course and is come and racial background (independent variable) on
in particular important as it is associated with future eco- smoking (dependent variable) via perceived daily stress
nomic resources and knowledge about health-related (mediator [34];). Because of the hierarchical structure of
strategies (e.g. strategies to cope with stress [29];). More- EMA data with multiple assessments nested under par-
over, educational attainment has been a widely used proxy ticipants, the non-independence of observations was
for social disadvantage in previous studies and it is less taken into account using multilevel analysis. For the ana-
prone to endogeneity bias from reverse causality (smoking lysis, 2–1-1 mediation models with random intercepts
effecting educational attainment) than measure such as in- and fixed slopes were used [35]. This means that the in-
come. Those with lower educational attainment, therefore, dependent variable (social disadvantage) was measured
are considered as more socially disadvantaged. Education on the between participant level, whereas the dependent
was dichotomized as lower education (“8th grade or less”, variable (CPD) and the mediator variable (perceived
some high school, no graduation/ GED”, “high school daily stress) were assessed on the within participant
graduate/ GED”; coded as higher social disadvantage = 1) level, i.e. via repeated measurement occasions (Fig. 1).
and higher education (“some college”, “college graduate/ The intercepts of the dependent variable (CPD) and of
degree”, “some graduate work”, “graduate degree”; coded the mediator variable (perceived daily stress) were
as lower social disadvantage = 0). allowed to vary on the between participant level, and
Income, in contrast, captures the resources individuals their residual variances to correlate. In the analysis we
can access to purchase goods and support in times of controlled for gender, age and day in study. To estimate
need and defines the possibility someone has to afford these models, MPlus 7 was used (TYPE = TWOLEVEL
prestige in society [11]. For example, individuals with RANDOM [36];). All analyses controlled for day in the
low income may lack the ability to purchase goods and study, baseline CPD and gender.
Jahnel et al. BMC Public Health (2019) 19:1284 Page 4 of 8

Social disadvantage

Level-2
(participant-level)
Level-1
(day-level)

Stress Stress Stress Stress

Experience Experience Experience … Experience

N Cigarettes N Cigarettes N Cigarettes … N Cigarettes

Day 1 Day 2 Day 3 … Day n

Time
Fig. 1 Conceptual diagram of the 2–1-1 mediation model

Results period compared to non-AA individuals. Higher levels of

Information on participant characteristics in regards to average perceived stress within individuals in turn was
educational attainment, income and race are shown in related to greater increases in CPD when controlling for
Table 1. Lower educated participants were significantly baseline CPD. This resulted in a significant indirect ef-
older compared to higher educated participants but did fect of racial background via perceived daily stress on
not differ in regards to gender or baseline CPD. Partici- CPD while controlling for the between-person direct ef-
pants who reported higher income did not differ in fect of racial background on CPD. This indirect effect is
regards to demographic characteristics when compared the result of higher intercepts of daily stress and less
with participants with lower income. African American CPD for AA smokers compared to non-AA smokers
smokers were significantly older and reported smoking (Table 3). When separating the sample by race, the re-
fewer cigarettes at baseline compared to smokers with sults show a non-significant negative effect of daily stress
different ethnic background. on CPD among AA smokers and a non-significant posi-
The results for the multilevel mediation analysis are tive effect of daily stress on CPD among non-AA
shown in Table 2. Participants with lower educational smokers. No indirect or direct effects were found for
attainment perceived significantly more daily stress than income as indicator of social disadvantage.
participants with higher education. On days during
which participants perceived more daily stress a higher Discussion
number of cigarettes were smoked. This resulted in a Using a within-subjects design, the aim of this project was
small significant indirect effect of education via per- to examine the role of the experience of daily stress as an
ceived daily stress on CPD while controlling for the dir- intermediary process linking social disadvantage to cigarette
ect effect of education on CPD, baseline CPD and consumption in an US sample of smokers. The results sug-
gender. The findings show that perceived daily stress gest that those individuals who are socially disadvantaged in
partially mediates the effect of education on CPD. terms of education and racial background experienced more
When using race as indicator for social disadvantage, daily stress. Those who experienced more daily stress in turn
the results show that AA smokers reported smoking smoked more cigarettes, resulting in significant indirect ef-
fewer cigarettes than non-African American smokers. fects of education and racial background on smoking, medi-
However, AA smokers perceived significantly greater ated by stress. Although the within-person positive indirect
daily stress averaged across days of the monitoring effect indicates that AA smokers are more affected by daily

Table 1 Participant Characteristics

Overall Low education High education Low income High income African American Other Ethnicity
(n = 194) (n = 81) (n = 113) (n = 91) (n = 103) (n = 73) (n = 121)
Age 41.2 (11.2) 42.6 (10.6)** 40.0 (11.5) 40.9 (11.2) 41.5 (11.2) 43.6* (9.8) 39.7 (11.7)
Gender (Male) 55.2% (n = 107) 56.8% (n = 46) 54.0% (n = 61) 54.9% (n = 50) 55.3 (n = 57) 46.6% (n = 34) 60.3% (n = 73)
CPD 16.1 (6.5)*** 16.5 (6.4) 15.7 (6.6) 15.9 (7.0) 16.2 (6.0) 14.2** (6.0) 17.2 (6.6)
Entries are M (SD), unless % is specified. M = mean, SD = standard deviation, CPD = Cigarettes per day. Comparisons were made using t-tests (continuous variables)
or chi-square tests (categorical variables). * p < .05, ** p < .01, *** p < .001
Jahnel et al. BMC Public Health (2019) 19:1284 Page 5 of 8

Table 2 Coefficient Estimates for parameters in 2–1-1 mediation model predicting CPD with education, income and race as
indicators for social disadvantage
Coefficient Estimates Education Race Income
Fixed Effects (Measurement occasion level)
Intercept CPD 1.99 (1.13) 2.20 (1.10)* 1.74 (1.13)
Intercept Daily Stress (DS) 71.52 (2.56)*** 76.00 (1.41)*** 78.98 (1.45)***
Social Disadvantage (D) → DS 8.42 (2.81)** 6.95 (2.10)*** −0.82 (2.18)
DS → CPD 0.01 (0.01)** 0.01 (0.01)** 0.01 (0.01)**
Direct Effect SD → CPD −0.53 (0.66) −1.39 (0.55)* −0.35 (0.61)
Indirect Effect 0.12 (0.04)* 0.10 (0.05)* −0.01 (0.03)
Total Effect −0.42 (0.65) −1.29 (0.54)* −0.37 (0.61)
Random Effects (Participant level)
Residual Variance Intercept DS 203.07 (19.18)*** 202.08 (18.92)*** 212.89 (19.21)***
Residual Variance Intercept CPD 16.16 (1.97)*** 15.79 (1.91)*** 16.17 (1.96)***
Covariance CPD DS 12.61 (0.81)*** 12.61 (0.81)*** 12.61 (0.81)***
Entries are B (SD), B = coefficient estimates, SD = standard deviation, CPD = Cigarettes per day, DS = Daily stress, D = Social Disadvantage. *p < .05,
**p < .01, ***p < .001

stress experience, the direct effect showed that AA partici- stressors are more severe and are experienced as more dis-
pants smoked less overall compared to smokers from other ruptive to their daily lives through a lack of material and
racial backgrounds. This is in line with previous research psychological coping skills [40]. Individuals with lower in-
suggesting that differences in CPD for AA and non-AA come may experience rather long-term stress in the form
smokers may be explained through biological markers as of financial strain, which may not necessarily translate into
AA smokers have a slower rate of nicotine metabolism [37] daily stress experience, but rather chronic stress.
and that smokers who have a faster rate of nicotine metabol- Some limitations need to be noted when interpreting
ism smoke more cigarettes per day [38]. the findings. Firstly, we used a crude single-item measure
The findings suggest that the experience of daily stress of daily stress experience which may not adequately reflect
might function as one possible factor linking social disad- how much stress participants experienced on a day-to-day
vantage and smoking [14, 21]. Further, the results support basis. However, other studies have found that single-item
previous findings that social disadvantage is associated with stress measures are both reliable and valid, perform com-
perceived stress [39]. In addition, it could be shown that parably to longer stress scales [41] and have been success-
the association between social disadvantage and the experi- fully tested in real-world studies [42]. Future research on
ence of daily stress differed according to the indicators of the association between daily stress and smoking may
social disadvantage used in this study. This suggests that consider collecting data on stress experience not just at
the mechanisms of how social disadvantage translates into the end of each day, but at multiple time points per day in
daily stress experience differ, depending on which indicator order to get a more accurate measure of the intensity of
for social disadvantage is used. Each indicator of social dis- daily stress. Assessing the different types of stressors
advantage highlights important differences in terms of linked to the stress experience may also yield deeper in-
scope and implications on smoking, and although they side into the effects of stress on smoking. In addition, for
overlap in how they affect smoking behavior, they should social disadvantage, we used proxy measures such as edu-
not be used interchangeably. We can only speculate, but cation, income and racial background. Other indicators
for example, individuals with lower educational attainment such as occupation, that tap explicitly into a person’s
may be more vulnerable to daily stress because their standing within a society may yield further inside into the
effects of social disadvantage and smoking. Nevertheless,
the indicators used in this study have been widely used in
Table 3 Estimates for daily stress and CPD for AA and non-AA
smokers
health research [43]. Future studies might consider more
comprehensive measures of social disadvantage in order
AA Non-AA
to better capture differences in stress and smoking related
Intercept CPD (SD) 9.952 (1.630) 10.263 (1.213)
to socioeconomic status. In order to gain sufficient and
Intercept Daily Stress (SD) 82.542 (1.633) 76.00 (1.41) comparable sample sizes we dichotomized the indicators
Daily stress → CPD B (SD) −0.013 (0.018) 0.012 (0.015) which may have minimized their effects on daily stress ex-
SD = standard deviation, CPD = Cigarettes per day perience and smoking. Larger samples would yield the
Jahnel et al. BMC Public Health (2019) 19:1284 Page 6 of 8

opportunity to explore effects of social disadvantage on study with more recent data is needed to confirm the re-
daily stress experience and smoking more decidedly. For sults from our study.
an EMA study however, the sample size was relatively In summary, the findings highlight the need for more
large considering the vast amount of data collected for focused research, with more diverse samples and better
each individual participant. developed theory, in order to better understand the role
Additionally, we have explored the effects of social dis- of daily stress on social inequalities in smoking behavior.
advantage and daily stress experience on CPD, as the From an intervention perspective, examining the link be-
most important driver of nicotine intake. Other factors, tween stress and smoking is important as smoking is
such as puffing topography, or nicotine extraction per suggested to be a maladaptive coping response to stress
cigarette might also be important indicators of smoking [16]. Some social support interventions target stress
as a means to relieving stress. In the context of an EMA management and coping skills [52, 53]. As noted, the lim-
study however, these factors would be difficult to meas- ited number of studies conducted to date broadly supports
ure and potentially add substantial participant burden. the premise that social support interventions targeting
However, further research may explore ways to passively better management of daily stress may improve cessation
detect smoking instances, including puffing topography outcomes, especially when focusing on socially disad-
and how they relate to social disadvantage and daily vantaged smokers. The findings from this work, how-
stress experience. ever, suggest that the direct effect of stress on day-to-
Further, EMA data is correlational in nature which day smoking is likely minimal, suggesting that add-
make causal interpretations difficult. Thus, reverse caus- itional work is required to understand how such inter-
ation explanations (i.e. smoking causing stress) are also ventions should be utilized.
plausible. Some models such as the stress induction model
posits that smoking induces stress [44]. It has also been
suggested that smoking may induce (financial) stress and Conclusion
that stress may lead to smoking induced deprivation [45]. The current study extends previous work on stress re-
However, the findings of Study 1 are supported by experi- lated effects on smoking by taking a within-participant
mental studies, showing that the experience of stress re- and time-varying perspective. Using real-time, ecologic-
duces the ability to resist smoking and increases both ally valid assessments of daily stress experience and
smoking intensity and reward [46, 47]. Further, self- smoking, we examined whether people experience, as a
reports and situational characteristics have the potential function of different indicators of social disadvantage,
to induce reactivity, which means that the monitoring it- different levels of daily stress, and whether this affected
self might change participant’s behavior. To date, research smoking differentially. The mixed findings and small ef-
on the potential that EMA methods might have on re- fect sizes highlight the need for future studies investigat-
activity are mixed [48, 49]. Nevertheless, EMA methods ing whether and how the experience of daily stress
represent a substantial improvement over more common influences smoking and smoking cessation. Such infor-
retrospective methods, as they maximize ecological valid- mation would aid in the development and improvement
ity, while avoiding recall bias [50]. of social support interventions specifically tailored to so-
It must also be noted that the significant effects we cially disadvantaged populations that focus on healthier
found were only small (see Table 2). Hence, perceived coping skills and daily stress management. In order to
stress does not offer a comprehensive explanation of better understand the social gradient in smoking behav-
how educational attainment and racial background ior future research may focus on investigating the effects
might translate into smoking behavior. Other variables, of different indicators of social disadvantage on stress
such as the environment, social norms and networks experience and the types of stress (e.g. daily hassles) and
may represent other processes linking SES and smoking how this affect daily smoking.
[51]. However, as we conducted secondary analysis for
this study, socioeconomic variables were not the main Abbreviations
focus when data was originally collected. AA: African American; CPD: Cigarettes smoked per day; DS: Daily stress;
ED: Electronic diary; EMA: Ecological Momentary Assessment; SD: Social
Lastly, the data collection ended about 9 years ago. A Disadvantage; SES: Socioeconomic status
substantial amount of time has passed, in which smoking
related public health interventions, changes in the costs
Acknowledgements
of living etc. may have had significant effects on the rela- The authors acknowledge that this research was conducted with the support
tionship between social disadvantage and smoking. With of an Australian Government Research Training Program Scholarship. The
rising inequalities however, we would expect the effects authors are grateful to Thomas Kirchner, Deborah Scharf and Sarah M. Scholl
for help launching the study and for input on study design; and to Anna
of social disadvantage on daily stress experience and Tsivina, Joe Stafura, Rachelle Gish, and Aileen Butera for their work
smoking to be even stronger today. Replication of this conducting research sessions.
Jahnel et al. BMC Public Health (2019) 19:1284 Page 7 of 8

Authors’ contributions 11. Adler NE, Snibbe AC. The role of psychosocial processes in explaining the
SGF and SS designed the original study, wrote the protocol, contributed to gradient between socioeconomic status and health. Curr Dir Psychol Sci.
study concept, and edited the manuscript. BS contributed to the study 2003;12:119–23.
concept, supervised writing, and contributed to the statistical analysis. TJ 12. Carr D, Umberson D. The social psychology of stress, health, and coping. In:
developed the study concept, conducted the statistical analysis and wrote DeLamater J, Ward A, editors. Handbooks of sociology and social research.
the first draft of the manuscript and all authors contributed to and have Handbook of social psychology (pp. 465-487). New York: Springer Science +
approved the final manuscript. Business Media; 2013.
13. De Vogli R, Santinello M. Unemployment and smoking: does psychosocial
Funding stress matter? Tob Control. 2005;14:389–95.
Data collection for this study was supported by Grant R01-DA020742 (PI: 14. Kaplan SA, Madden VP, Mijanovich T, Purcaro E. The perception of stress and
Shiffman) from the U.S. National Institutes of Health, National Institute on its impact on health in poor communities. J Community Health. 2013;38:142–9.
Drug Abuse. The National Institute on Drug Abuse had no role in the study 15. Hajek P, Taylor T, McRobbie H. The effect of stopping smoking on perceived
design, collection, analysis or interpretation of the data, writing the manu- stress levels. Addiction. 2010;105:1466–71.
script, or the decision to submit the paper for publication. 16. Park CL, Iacocca MO. A stress and coping perspective on health behaviors:
theoretical and methodological considerations. Anxiety Stress Coping. 2014;
27:123–37.
Availability of data and materials 17. Bandiera FC, Atem F, Ma P, Businelle MS, Kendzor DE. Post-quit stress mediates the
The dataset used and analysed during the current study are available from relation between social support and smoking cessation among socioeconomically
the corresponding author on reasonable request. disadvantaged adults. Drug Alcohol Depend. 2016;163:71–6.
18. Pearlin LI, Schieman S, Fazio EM, Meersman SC. Stress, health, and the life
Ethics approval and consent to participate course: some conceptual perspectives. J Health Soc Behav. 2005;46:205–19.
The study was approved by the Institutional Review Board at the University 19. Grzywacz JG, Almeida DM, Neupert SD, Ettner SL. Socioeconomic status and
of Pittsburgh (REN16120111/IRB0606147). Written informed consent was health: a micro-level analysis of exposure and vulnerability to daily stressors.
obtained, using a consent form approved by the University of Pittsburgh IRB. J Health Soc Behav. 2004;45:1–16.
20. Todd M. Daily processes in stress and smoking: effects of negative events,
nicotine dependence, and gender. Psychol Addict Behav. 2004;18:31–9.
Consent for publication
21. Slopen N, Kontos EZ, Ryff CD, et al. Psychosocial stress and cigarette
Not applicable.
smoking persistence, cessation, and relapse over 9-10 years: a prospective
study of middle-aged adults in the United States. Cancer Causes Control.
Competing interests 2013;24:1849–63.
SS provides consulting services on tobacco harm minimization (including 22. Jorm AF, Rodgers PA, Christensen H, Henderson S, Korten AE. Smoking and
nicotine replacement therapy and digital vapor products) to subsidiaries of mental health: results from a community survey. Med J Aust. 1999;170:74–7.
Reynolds American Inc. SGF has worked as a consultant for pharmaceutical 23. Vollrath M. Smoking, coping and health behavior among university
companies on matters relating to smoking cessation. All other authors students. Psychol Health. 1998;13:431–41.
declare that they have no conflicts of interest. 24. Kassel JD, Stroud LR, Paronis CA. Smoking, stress, and negative affect:
correlation, causation, and context across stages of smoking. Psychol Bull.
Author details 2003;129:270–304.
1
College of Health & Medicine, University of Tasmania, Hobart, Australia. 25. Shiffman S, Dunbar MS, Li X, Scholl SM, Tindle HA, Anderson SJ, Ferguson
2
Department of Psychology, University of Pittsburgh, Pittsburgh, USA. SG. Craving in intermittent and daily smokers during ad libitum smoking.
3
Institute for Public Health and Nursing Science, University of Bremen, Nicotine Tob Res. 2014;16(8):1063–9. https://doi.org/10.1093/ntr/ntu023.
Bremen, Germany. 26. Shiffman S, Dunbar MS, Li X, Scholl SM, Tindle HA, Anderson SJ, Ferguson SG.
Smoking patterns and stimulus control in intermittent and daily smokers. PLoS
Received: 5 June 2019 Accepted: 16 September 2019 One. 2014;9(3):e89911. https://doi.org/10.1371/journal.pone.0089911.
27. Schüz B. Socio-economic status and theories of health behaviour: time to
upgrade a control variable. Br J Health Psychol. 2017;22:1–7.
References 28. Elo IT, Preston SH. Educational differences in mortality: United States, 1979-
1. Fidler JA, Jarvis MJ, Mindell J, West R. Nicotine intake in cigarette smokers in 85. Soc Sci Med. 1996;42:47–57.
England: distribution and demographic correlates. Cancer Epidemiol 29. Ross CE, Wu CL. The links between education and health. Am Sociol Rev.
Biomark Prev. 2008;17:3331–6. 1995;60:719–45.
2. Hiscock R, Bauld L, Amos A, Fidler JA, Munafo M. Socioeconomic status and 30. Businelle MS, Kendzor DE, Costello TJ, et al. Light versus heavy smoking
smoking: a review. Ann N Y Acad Sci. 2012;1248:107–23. among African American men and women. Addict Behav. 2009;34:197–203.
3. Marmot M. Smoking and inequalities. Lancet. 2006;368:341–2. 31. Slopen N, Dutra LM, Williams DR, et al. Psychosocial stressors and cigarette
4. Gregoraci G, van Lenthe FJ, Artnik B, et al. Contribution of smoking to smoking among African American adults in midlife. Nicotine Tob Res. 2012;
socioeconomic inequalities in mortality: a study of 14 European countries, 14:1161–9.
1990–2004. Tob Control. 2017;26:260–8. 32. Hill LK, Hoggard LS, Richmond AS, et al. Examining the association between
5. Van der Heyden JHA, Schaap MM, Kunst AE, et al. Socioeconomic perceived discrimination and heart rate variability in African Americans.
inequalities in lung cancer mortality in 16 European populations. Lung Cultur Divers Ethnic Minor Psychol. 2017;23:5–14.
Cancer. 2009;63:322–30. 33. Shiffman S, Dunbar MS, Ferguson SG. Stimulus control in intermittent and
6. De Lew N, Weinick RM. An overview: eliminating racial, ethnic, and SES daily smokers. Psychol Addict Behav. 2015;29:847–55.
disparities in health care. Health Care Financ Rev. 2000;21:1–7. 34. Krull JL, MacKinnon DP. Multilevel modeling of individual and group level
7. Poland B, Frohlich K, Haines RJ, et al. The social context of smoking: the mediated effects. Multivar Behav Res. 2001;36:249–77.
next frontier in tobacco control? Tob Control. 2006;15:59–63. 35. Preacher KJ, Zyphur MJ, Zhang Z. A general multilevel SEM framework for
8. Harwood GA, Salsberry P, Ferketich AK, Wewers ME. Cigarette smoking, assessing multilevel mediation. Am Psychol Assoc. 2010;15:209–33.
socioeconomic status, and psychosocial factors: examining a conceptual 36. Muthén LK, Muthén BO. MPlus User’s Guide. 7th ed. Los Angeles: Muthén &
framework. Public Health Nurs. 2007;24:361–71. Muthén; 2012.
9. Cohen S, Janicki-Deverts D. Who's Stressed? Distributions of Psychological 37. Shiffman S, Dunbar MS, Benowitz NL. A comparison of nicotine biomarkers and
Stress in the United States in Probability Samples from 1983, 2006, and smoking patterns in daily and nondaily smokers. Cancer Epidemiol Biomark Prev.
20091. J Appl Soc Psychol. 2012;42:1320–34. 2014;23(7):1264–72. https://doi.org/10.1158/1055-9965.EPI-13-1014.
10. Mulder BC, de Bruin M, Schreurs H, van Ameijden EJ, van Woerkum CM. 38. Benowitz NL, Pomerleau OF, Pomerleau CS, Jacob P 3rd. Nicotine
Stressors and resources mediate the association of socioeconomic position metabolite ratio as a predictor of cigarette consumption. Nicotine Tob Res.
with health behaviours. BMC Public Health. 2011;11:798. 2003;5(5):621–4 6DCURVYMPL1C6UQ5.
Jahnel et al. BMC Public Health (2019) 19:1284 Page 8 of 8

39. Algren MH, Ekholm O, Nielsen L, et al. Associations between perceived

stress, socioeconomic status, and health-risk behaviour in deprived
neighbourhoods in Denmark: a cross-sectional study. BMC Public Health.
2018;18:250.
40. Almeida DM, Neupert SD, Banks SR, Serido J. Do daily stress processes account
for socioeconomic health disparities? J Gerontol Series B. 2005;60B:34–9.
41. Littman AJ, White E, Satia JA, Bowen DJ, Kristal AR. Reliability and validity of 2
single-item measures of psychosocial stress. Epidemiology. 2006;17:398–403.
42. Jones M, Taylor A, Liao Y, Intille SS, Dunton GF. Real-time subjective
assessment of psychological stress: associations with objectively-measured
physical activity levels. Psychol Sport Exerc. 2017;31:79–87.
43. Galobardes B, Shaw M, Lawlor DA, Lynch JW, Davey Smith G. Indicators of
socioeconomic position (part 1). J Epidemiol Community Health. 2006;60:7–12.
44. Parrott AC. Does cigarette smoking cause stress? Am Psychol. 1999;54:817–20.
45. Siahpush M, Borland R, Yong HH. Sociodemographic and psychosocial
correlates of smoking-induced deprivation and its effect on quitting:
findings from the International Tobacco Control Policy Evaluation Survey.
Tob Control. 2007;16:e2.
46. Childs E, de Wit H. Effects of acute psychosocial stress on cigarette craving
and smoking. Nicotine Tob Res. 2010;12:449–53.
47. McKee SA, Sinha R, Weinberger AH, et al. Stress decreases the ability to
resist smoking and potentiates smoking intensity and reward. J
Psychopharmacol. 2011;25:490–502.
48. Hufford MR, Shields AL, Shiffman S, Paty J, Balabanis M. Reactivity to
ecological momentary assessment: an example using undergraduate
problem drinkers. Psychol Addict Behav. 2002;16:205–11.
49. Rowan PJ, Cofta-Woerpel L, Mazas CA, et al. Evaluating reactivity to
ecological momentary assessment during smoking cessation. Exp Clin
Psychopathol. 2007;15:382–9.
50. Schwarz N. Retrospective and concurrent self-reports: the rationale for real-
time data capture. In: Stone AA, Shiffman S, Atienza A, Nebeling L, editors.
The science of real-time data capture: self-reports in health research. New
York: Oxford University Press; 2007. p. 11–26.
51. Jahnel T, Ferguson SG, Shiffman S, Thrul J, Schüz B. Momentary smoking
context as a mediator of the relationship between SES and smoking. Addict
Behav. 2018;83:136–41.
52. Burns RJ, Rothman AJ, Fu SS, Lindgren B, Joseph AM. The relation between
social support and smoking cessation: revisiting an established measure to
improve prediction. Ann Behav Med. 2014;47:369–75.
53. Creswell KG, Cheng Y, Levine MD. A test of the stress-buffering model of
social support in smoking cessation: is the relationship between social
support and time to relapse mediated by reduced withdrawal symptoms?
Nicotine Tob Res. 2015;17:566–71.

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.