Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Nov 7.
Published in final edited form as: Health Promot Pract. 2020 Jul 23;22(6):850–862. doi: 10.1177/1524839920942214

A Pilot Study and Ecological Model of Smoking Cues to Inform Mobile Health Strategies for Quitting Among Low-Income Smokers

Shuo Zhou 1,2, Arnold H Levinson 1,2, Xuhong Zhang 1, Jennifer D Portz 2, Susan L Moore 1,2, M Odette Gore 2,3, Kelsey L Ford 1,2, Qing Li 4, Sheana Bull 1,2
PMCID: PMC10630930  NIHMSID: NIHMS1935889  PMID: 32698702

Abstract

One crucial factor that leads to disparities in smoking cessation between groups with higher and lower socioeconomic status is more prevalent socioenvironmental smoking cues in low-income communities. Little is known about how these cues influence socioeconomically disadvantaged smokers in real-world scenarios and how to design interventions, especially mobile phone–based interventions, to counteract the impacts of various types of smoking cues. We interviewed 15 current smokers living in low-income communities and scanned their neighborhoods to explore smoking-related experiences and identify multilevel cues that may trigger them to smoke. Findings suggest four major types of smoking cues influence low-income smokers—internal, habitual, social, and environmental. We propose an ecological model of smoking cues to inform the design of mobile health (mHealth) interventions for smoking cessation. We suggest that user-triggered strategies will be most useful to address internal cues; server-triggered strategies will be most suitable in changing perceived social norms of smoking and routine smoking activities to address social and habitual cues; and context-triggered strategies will be most effective for counteracting environmental cues. The pros and cons of each approach are discussed regarding their cost-effectiveness, the potential to provide personalized assistance, and scale.

Keywords: mHealth, smoking cessation, smoking cues, low-SES population, ecological model, just-in-time adaptive intervention

INTRODUCTION

In the United States, smoking prevalence is nearly three times higher among low-income adults (21.4% with annual household income <$35,000) than among higher income adults (7.6% with annual household income >$100,000; Centers for Disease Control and Prevention, 2018). Nearly three fourths of U.S. cigarette smokers are characterized as having low socioeconomic status (SES). Although there are no differences in quit attempts by income (Babb et al., 2017; Reid et al., 2010), the quit rate is much lower among low-SES smokers (Levinson, 2017). Thus, the key issue is not the differential motivation to quit smoking, but the differential success of quitting between high- and low-SES groups.

These disparities may be partly attributable to SES differences in exposure to cues to smoke, with lower SES populations more exposed to various types of smoking cues. First, environmental cues to smoke are more common in lower SES communities. The tobacco industry targets low-income neighborhoods with more cigarette advertising (Brown-Johnson et al., 2014; Cantrell et al., 2015), and tobacco outlet density is higher in communities with lower household income, larger proportions of Blacks, Hispanics, and women with low levels of education (Rodriguez et al., 2013; Yu et al., 2010). Tobacco outlet density is directly associated with smoking initiation (Morgenstern et al., 2013) and inversely associated with cessation (Kirchner, 2013).

Second, higher smoking prevalence in lower SES communities means more exposure to other smokers, including friends and family members. Cue-reactivity studies have demonstrated that nicotine-dependent people who are exposed to somebody smoking, see a lighted cigarette, or smell cigarette smoke are likely to experience physiological arousal, craving, and smoking behaviors (Conklin & Tiffany, 2001; Shiffman et al., 2002). Smoking-related environments (e.g., bars, restaurants) and people around whom smokers regularly smoke also elicit strong reactivity from smokers (Conklin, 2006; Conklin et al., 2013).

Finally, stress and negative moods are common cues to smoke across SES categories, but low-SES lifestyles often involve more frequent or chronic stress.

Current approaches to smoking cues have classified them into proximal cues (directly linked to tobacco use, e.g., cigarettes, ashtrays) versus distal cues (cars, people; Conklin et al., 2008), or cues that are internal (e.g., negative affect, urge to smoke) versus external (e.g., packs of cigarettes in a room; Herman, 1974; Litvin & Brandon, 2010). While these classifications are useful, a more theoretically informed and nuanced framework could help frame intervention research toward disrupting the link between cue exposure and smoking behavior among low-SES populations.

MOBILE HEALTH (MHEALTH) TECHNOLOGY FOR SMOKING CESSATION

Mobile technology offers smokers who want to quit a convenient way to track their behavior, more accurately identify smoking cues, and deliver automated support in real time to resist the temptations to smoke (Naughton et al., 2016). mHealth interventions have been widely used to educate, motivate, and empower people for health behavioral changes (James & Harville, 2017). Meta-analyses and systematic reviews of current literature show that mHealth interventions can significantly improve medication adherence, access to and delivery of health services, self-management of chronic diseases, physical activity, eating habits, and smoking quit rates; and reduce deaths and hospitalization, asthma symptoms, and depressive symptoms (Free et al., 2013; Krishna et al., 2009; Marcolino et al., 2018).

The low cost and high portability make mobile devices highly accessible across SES groups. In 2018, 95% of Americans owned a cellphone (Pew Research Center, 2019). One fifth of adults in households with less than $30,000 annual income were “smartphone-only” Internet users (Pew Research Center, 2019), and reliance on smartphones for information access is especially common among younger adults, non-Whites, and lower income Americans. Despite the high accessibility of mobile phone technology among disadvantaged groups and the potential of using mobile health technologies to reduce health disparities (Gurman et al., 2012; Muñoz, 2010), few mHealth interventions for smoking cessation have been specifically designed for low-SES groups (Cox et al., 2011). Among the limited studies that tested the efficacy of smoking cessation interventions in low-SES smokers, most of them focused only on text message–based interventions (Boland et al., 2018; Vidrine et al, 2019). How to effectively target disadvantaged populations and deliver tailored, timely support to increase successful quitting among low-SES smokers is largely unknown.

The developments of smartphones and mobile technologies enable an emerging mHealth strategy—just-in-time adaptive interventions (JITAIs), which prompt users with appropriate amount of personalized content, in the right format, and at their most needed times (Nahum-Shani et al., 2017). JITAIs can deliver support through three main mechanisms: user-triggered, server-triggered, and context-triggered interventions (Naughton, 2017). User-triggered interventions respond to a smoker’s decision to request or access support. For example, people could send certain messages (e.g., QUIT, RELAPSE, or HELP) and immediately receive messages directing them to appropriate resources. Server-triggered interventions use software systems to deliver support, “usually based on fixed schedules, random timing, a combination of the two, or schedules tailored to the individual’s self-reported predicted future behavior or actual past behavior” (Naughton, 2017, p. 380). Context-triggered interventions use sensors (e.g., motion sensors to detect physical activities or global positioning system to detect geographic locations) on or connected to the user’s smartphone to identify an individual’s immediate environment and, if predetermined to be high risk for smoking or relapses, deliver support. These interventions can provide behavioral support to address a need in real-time and tailored to each user’s unique situations. Studies of all three strategies are in their early stages and have not been fully applied to smoking cessation or tested in low-SES smokers. Systematic reviews suggest server-triggered interventions are often used to promote physical activities (Hardeman et al., 2019), whereas user-triggered interventions are frequently used to address mental health issues (Wang & Miller, 2019). Context-triggered interventions have been successfully used to predict emotional status and advise physical activity and healthy eating (Burns et al., 2011; Rabbi et al., 2015).

To our knowledge, with very few exceptions, limited intervention programs are designed to help low-SES smokers cope with real-time exposure to multiple levels of cues to smoke. In the current article, we report results of an exploratory smoking cue exposure study among low-income smokers; based on the findings, we propose an ecological model of smoking cues, map the types of cues to mHealth support mechanisms, and provide guidance for smoking cessation practices among low-SES smokers.

METHOD

Design

We conducted baseline surveys, structured interviews, and participant-guided tours with 15 current smokers who live in low-income communities. Interview data were content-analyzed to identify major categories of smoking triggers among low-income smokers. Data from guided tours were analyzed to explore the relationship between smoking cues and participant responses to cues.

Participant Recruitment and Setting

Low-income current smokers who intend to quit were recruited through flyers in both English and Spanish, a survey registry, and promotional events at community meetings in North Aurora, Colorado, where communities have higher health disparities, lower income, and educational levels than other parts of the Denver metropolitan area. The median household income in North Aurora was $49,793 in 2017 (www.city-data.com), much lower than the national median household income of $61,372 (U.S. Census Bureau, 2018). About 29.4% of people living in this area were in poverty—nearly three times of the poverty rate (10.6%) in the Denver metropolitan area, and only 18.4% of adults older than 25 years received college degree or above (denvermetrodata.org). Participants met the following inclusion criteria: (1) between 18 and 89 years old, (2) current smokers who have an intention to quit, (3) at or below 200% of the federal poverty level (e.g., a family of four earning less than $47,248), (4) English or Spanish speakers. Potentially interested smokers were contacted through email or phone to determine eligibility and, if eligible, were invited to complete a baseline survey of demographics, addresses, community environment, current smoking behaviors, intention to quit, and smoking history. Our goal was to recruit 10 to 20 participants to complete three study phases—baseline survey, structured interview, and participant-guided community tour. Following the principle of “data saturation,” it is usually suggested that 12 to 20 interviews of a homogenous group are enough to reach saturation (Green & Thorogood, 2018; Guest et al., 2006). Twenty-five potential participants completed the baseline survey; four did not provide consistent information regarding their household income, and six declined to participate further, so they were removed from the study. We stopped recruiting more participants when 15 participants completed all phases of the study since it becomes less likely to collect additional data to form new conceptual categories related to smoking cues.

Procedures

Each of the 15 retained participants scheduled an interview with the same researcher, who was fluent in both English and Spanish, to make sure the interviews are consistent and comparable. Since we only had one Hispanic participant who spoke Spanish and English equally well, the interviews were all conducted in English. A standard interview guide was developed based on the theoretical framework of social cognitive theory (Bandura, 2002), classical conditioning of smoking cues (Lazev et al., 1999), and ecological models of health behavior (Emmons, 2000; Hoffman & Driscoll, 2000; Hovell et al., 2009; McLeroy et al., 1988). The guide included open-ended questions about reasons for smoking; frequency of smoking behavior; locations, time, and specific triggers of smoking; peer and family opinions about smoking; challenges for quitting smoking; and ways to facilitate quitting smoking. The interview was used to characterize smoking patterns and have participants enumerate their smoking triggers. Interviews were conducted at participants’ homes, lasted between 20 and 40 minutes, and were audio recorded and transcribed verbatim.

Participants were then facilitated to develop and conduct guided tours of their communities for research staff. They first listed their four most frequently visited places on a weekly basis (e.g., workplace, convenient stores, supermarkets, friend’s home) and drew their routes from home to these places on a map. Guided by participants, the researcher visited these places on foot or by car and asked participants along the way to identify cues in their community that made them want to smoke. At each identified smoking cue, participants completed a survey that assessed the impact of the cue (e.g., craving level). The researcher also documented features of the smoking cue and participant’s smoking-related behaviors (i.e., purchased or smoked a cigarette). Photos were taken to record each environmental smoking cue that participants identified during the guided tours. The guided tours lasted between 1.5 and 2.5 hours. As an incentive and aid to participation, each participant received a $30 gift card for participating in the baseline survey and an additional $100 gift card for participating in the in-depth interview and the guided tour. The study protocol was reviewed and approved by the Colorado Multiple Institutional Review Board.

Analysis

Content Analysis of Interviews.

The interviewer first reviewed transcripts to verify their accuracy. Answers to open-ended interview questions were examined line-by-line, and two coders separately read the same four transcripts and highlighted texts that described typical smoking scenarios. During this open-coding process, each coder independently defined a keyword in participants’ words to capture concepts related to smoking cues. The two coders compared initial codes, discussed and resolved discrepancies, and agreed on the code list. One coder then analyzed the remaining transcripts using the code list, creating new codes when data did not fit existing codes. Coding was manual without the use of a computer program.

After transcripts were coded, the coder examined all data within each particular code to determine whether some codes needed to be combined (i.e., “seeing cigarettes” and “seeing cigarette packages” to “exposure to tobacco-related products”) or split into subcategories (i.e., “drinking” to “drinking coffee” and “drinking alcohol”). Final codes were categorized to identify latent themes.

Analysis of Cue Exposure Surveys During Guided Tours.

We calculated frequencies of each type of cue mentioned in guided tours, and we used bivariate regressions to associate each type of cue with concurrent self-reported craving levels. The 12-item short form of the Tobacco Craving Questionnaire (Heishman et al., 2008) was used to measure craving levels after exposure to each type of cue. Responses were averaged into a craving scale ranging from 1 to 5.

RESULTS

Participant Characteristics

Participants (n = 15) included eight females, and seven Whites/Caucasians, five African Americans/Blacks, two Asians, and one Hispanic/Latino. The average age was 39.7 years (ranging from 21 years to 64 years). Roughly three fourths had more education than high school. Six were unemployed, retired, unable to work, or students. Three had a smoking history of 1 to 5 years, three 6 to 10 years, four 11 to 20 years, and five more than 20 years. Fourteen smoked cigarettes (six “light” users ≤10 cigarettes/day, seven “moderate” = 11–19, one “heavy” ≥ 20), and one used vape pens (one cartridge/day). Thirteen participants had tried to quit smoking more than once, and seven had been advised by a health care professional to quit smoking.

Interview Reports of Types of Smoking Cues

Four common types of cues emerged that would trigger people to smoke: (1) internal cues, (2) habitual cues, (3) social cues, and (4) environmental cues (Table 1). The most frequently mentioned cues were social, habitual, and internal (Table 2).

TABLE 1.

Meanings and Examples of Four Types of Smoking cues

Type of cues Meanings and examples
Internal cues Physiological and psychological, particularly emotional states that may induce people to smoke; for example, nicotine urges, feeling stressed, boredom
Habitual cues Routine or habitual actions that frequently co-occur with smoking; for example, driving, after eating, drinking coffee, drinking alcohol, before bed
Social cues Factors related to social others in the social contexts; for example, seeing others smoking on street, joining a social event, family arguments
Environmental cues Physically present factors in one’s environment. The encounter with environmental cues can be passive, for example, exposure to tobacco-related products, tobacco advertisements, and tobacco outlet stores
Open in a new tab

TABLE 2.

Top Smoking Cues Identified in Interviews and Guided Tours

Smoking cues Type of cues No. of participants identified this cue (%) in interviews No. of participants mentioned this cue at least once (%) in guided tours No. of times this cue was mentioned in guided tours (%)
Other people smoking Social 11 (74) 4 (27)   7 (11)
Joining social events Social   8 (53) — —
Food/after eating Habitual   8 (53) — —
Driving/in a car Habitual   7 (47) — —
Drinking alcohol Habitual   6 (40) — —
Feeling stressed Internal   6 (40) — —
Boredom Internal   5 (33) — —
Anxiety Internal   5 (33) — —
Waiting/between activities Habitual   4 (27) — —
Seeing tobacco-related products Environmental   4 (27) 6 (40) 11 (17)
Outside grocery stores Environmental   4 (27) 6 (40)   8 (13)
Seeing tobacco-related ads Environmental   4 (27) — —
At home Environmental — 7 (47) 11 (17)
At gas stations Environmental — 6 (40) 10 (16)
Open in a new tab

Internal Cues.

Participants frequently mentioned physiological, mental, and emotional reactions due to nicotine withdrawal, and nonnicotine-related negative emotions as major cues for craving and smoking. Seven participants expressed that nicotine or internal behavioral urges—not having nicotine for a while, or not having the physical action of putting a cigarette in the mouth, made them want to smoke. Other frequently mentioned internal states included feeling stressed (six participants), bored (five), anxious (five), frustrated (three), irritated (two), unfocused (one), reminded of the past (one), angry (one), sad (one), nervous (one), moody (one), pained (one), and having a migraine (one).

The irritation that comes along with not having the cigarettes, the nicotine. Easily frustrated. There’s a time where I had quit for about two weeks and it was—something had happened with the vehicle—I got a flat tire—and it was a trigger for me to go and buy a pack of cigarettes ‘cause it was very stressful. (Female, smoked for 1–5 years)

When I feel sad, I want to have a smoke to distract it … Just feel like I have something to smoke so I can forget about something else. (Male, smoked for 5–10 years)

Habitual Cues.

Smoking was also discussed as a habitual or routine activity. Habitual cues to smoke included food or after eating (eight), driving or being in the car (seven), drinking alcohol (six), waiting (four), drinking coffee (two), completing a task (two), working (two), waking up (one), before going to bed (one), and after sex (one).

A lot of the times I smoke after I eat. That’s one of the main things. And I will, on a routine basis, smoke first thing in the morning and before bed. Those are my specific routines. (Female, smoked for 1–5 years)

I smoked prior to leaving in Utah before we were to start our travels, and then, we stopped for gas in Rock Springs, Wyoming—also got out of the vehicle at a gas station off to the side and smoked at that time and then, again, in Laramie, Wyoming. Same thing. Got out of the vehicle and smoked. (Female, smoked for 1–5 years)

… but it’s also like a reward when I finish something. Finish an errand, finish a task, I’ll smoke then. (Female, smoked for more than 20 years)

Social Cues.

The majority of participants mentioned others’ smoking behavior or being in social settings usually triggers them to smoke. Social cues for smoking included seeing others smoking (11), joining social events (eight), and having conflicts/arguments with family members (two).

When I go out with friends—social gatherings—I have a tendency to want to join them when they’re out smoking. (Male, smoked for 10–20 years)

When I see someone smoking or light up a cigarette right in front of me, it’s just like an impulse in me … that I really want to smoke. It just comes out of nowhere. (Male, smoked for 1–5 years)

Social settings are hard. Typically, I didn’t use to smoke a lot. I was like, “It’s not a big deal. I’m not going to worry about it.” Then I stopped thinking about trying to quit. (Female, smoked for more than 20 years)

Environmental Cues.

Participants did not mention environmental cues until they were primed with a question that describes a person who wants to smoke when seeing a lighted cigarette or a smoking-related ad and asks whether they ever had similar experiences. In response, participants named exposure to tobacco-related advertisements (four), exposure to tobacco-related products (four), outside grocery/liquor stores (four), tobacco outlet stores (three), seeing people smoking on TV (three), and smoking cessation ads (one).

Well, when I see a Newport commercial, you know, that’s a trigger for me. Then I want a cigarette, I want a Newport. … I probably don’t even really want the cigarette, but you know, in my mind, it’s like you’re going to want a cigarette sooner or later, so you’ve got a chance to get a Newport. (Male, smoked for 10–20 years)

While seeing movies where let’s say a woman is smoking, it just looks almost like fancy. I don’t know how to explain it, but they make it look cool and I realize that I’ve been smoking since about 14 and that’s probably why. (It) is because I saw things around me and saw people around me smoke. That’s why I thought it was kind of normal to smoke. (Female, smoked for 5–10 years)

Guided Tours

Smoking Behaviors.

Seven participants smoked (half a cigarette to two cigarettes) during the guided tour of their community, and two participants purchased a pack of cigarettes during the tour. Smoking events took place when participants were outside a grocery store, a gas station, a convenience store, a liquor store, a coffee shop, a place of work, or seeing other people smoking.

Types of Smoking Cues.

Each participant identified three to six smoking cues along routes to their most commonly visited places. Sixty four cue instances were mentioned, comprising 12 distinct smoking cues. The top five were “seeing tobacco-related products” (11 times, 17%), “being at home” (11 times, 17%), “at gas stations” (10 times, 16%), “outside grocery stores” (eight times, 13%), and “seeing others smoking” (seven times, 11%; Table 2).

Craving Levels After Cue Exposure.

The unit of analysis is each instance of cue exposure. Overall, participants had an average craving level of 3.47 (SD = 1.16). The types of cues that resulted in relatively higher levels of craving included “coffee shop/restaurant” (M = 4.25), “shopping center” (M = 4.22), “workplace” (M = 4.04), “liquor store” (M = 3.94), “home” (M = 3.92), “convenience store” (M = 3.92), and “seeing others smoking” (M = 3.79).

DISCUSSION

This study expands on previous studies of smoking cues and suggests that they occur at multiple levels—intrapersonal (internal cues; habitual cues), interpersonal (social cues), and impersonal/community–institutional (environmental cues). Smokers were most aware of internal and habitual smoking cues, but may neglect environmental cues at the community level that also significantly influenced craving and smoking behaviors. These multiple levels closely align with ecological models of health behavior, which emphasize the complex determinants of human health and the dynamic interactions of biological, psychological, social, and environmental factors. Based on these findings, we propose an ecological model of smoking cues.

The proposed model (Figure 1) suggests that craving and smoking behavior are influenced by an individual’s physiological and psychological status, habitual behaviors, social contexts, and physical environments. The model takes a holistic view of cues to smoke that considers an individual’s social and physical environments. In theory, it suggests that smoking cessation practitioners should consider an individual’s risks for smoking or relapse at all levels and tailor smoking cessation interventions to each individual’s constantly changing physiopsychological and contextual states.

FIGURE 1.

FIGURE 1

Open in a new tab

An Ecological Model of Smoking Cues

Implications for mHealth Strategies to Counteract Smoking Cues

If a smoker who intends to quit could receive just-in-time support whenever a risk situation occurs, the danger of relapse can be greatly reduced. Although existing mHealth interventions have shown efficacy in using text messages or mobile apps to deliver educational materials, track smoking behavior and quit progress, and induce motivations to quit smoking (Buller et al., 2014; Ubhi et al., 2015), few intervention programs are designed to help low-SES users cope with various smoking cues or provide tailored real-time cessation support (Hébert et al., 2018; Hoeppner et al, 2019; Naughton et al., 2016). We propose that mHealth support can be aligned to different types of smoking cues. User-triggered strategies, which deliver support on user-initiated requests for help and self-report status change, will be most useful to address internal cues. Server-triggered strategies, which deliver preprogrammed system alerts to users based on a fixed schedule and frequency, will be most suitable to address social and habitual cues. Context-triggered strategies, which deliver support messages when the user nears a predefined context or location, will be most effective for counteracting environmental cues. In the next section, we further consider the use of each mHealth strategy for its corresponding cue type.

User-Triggered mHealth Interventions to Counteract Internal Cues for Smoking

In our pilot study, internal cues that originated from physiopsychological states were usually among the first cues enumerated. Self-reporting is the fastest and the most accurate way to identify one’s own changing physiological or mental states and needs for support. To prevent smokers from smoking or relapses triggered by internal cues (such as nicotine urge, feeling stressed, depressed, or bored), we can rely on user-triggered methods to deliver support through mobile devices. Smokers have to initiate the dialogue and report their current internal status, and then the intervention program will send customized messages directly to help with emotional management and craving reduction.

As suggested by mood management theory (Zillmann, 1988), people strive to get rid of negative emotions and to maintain positive emotions. Different types of messages can be used to facilitate mood management. Positive, pleasant messages and messages with great absorption potential can effectively intervene in bad moods. Messages with high levels of excitation can help individuals who suffer from boredom. People who are experiencing stressful situations can benefit from consuming relaxing content. When a user signals a need to address internal cues for smoking, intervention should immediately deliver coping strategies and personalized messages to help users get out of the negative state. In addition, according to social cognitive theory, individual cognitive factors (e.g., goals, self-efficacy, beliefs) are critical in influencing health behaviors, including smoking. Support messages should reinforce smokers’ cessation goals and enhance perceived efficacy in controlling their internal states and quitting smoking.

For example, the “Smiling Instead of Smoking” (SiS) app encouraged users to engage daily happiness exercises for 3 weeks and learn how to maintain or increase positive affect. The program was feasible and favorably rated by the users. Users reported increased confidence and decreased urge to smoke after use. Self-report abstinence rate was 53% (16/30) at 6-month post quit day (Hoeppner et al., 2019). In a text message-based program called “txt2stop,” 38.5% of the users self-initiated support request and texted “crave” or “lapse” to a central automated system. Users who texted “crave” in the first month were more likely to report abstinence at 1 month (Devries et al., 2012). Although using user-triggered interventions to address internal cues for smoking is promising and well received by users, there is a lack of controlled trials to examine effectiveness.

One advantage of user-triggered interventions for smoking cessation is that they can offer instant, personalized help tailored to smokers’ immediate physiopsychological status. As long as algorithms are built into the program, users will receive automated messages to address various kinds of internal needs (e.g., smoking urges, coping with stress, lack of motivation). Therefore, user-triggered interventions are easily sustained, scalable to large populations, and potentially cost-effective as they can achieve desirable health outcomes at a relatively low cost. Potential limitations include a need for high motivations from smokers to engage and initiate the program whenever support is needed, and difficulties in recognizing nuanced or unconscious mental states, both of which may lead to underreport of internal needs.

Server-Triggered mHealth Interventions to Counteract Habitual and Social Cues for Smoking

Many habitual cues are largely predictable, as they happen at certain times in certain locations. Server-triggered interventions, which deliver support based on prespecified rules, can help smokers resist habitual cues for smoking. Reactivity to habitual cues reflects prior classical conditioning. For example, if drinking is repeatedly paired with smoking behavior, an individual learns to associate the drinking stimulus with smoking and experiences craving whenever he or she drinks. One way to decouple habitual activity and smoking behavior is to reduce the expectancy of smoking after the activity. Intervention messages can be timed to an individual’s schedule of activities that are habituated to smoking (e.g., after waking up, during coffee breaks, before going to bed) and can deliver attention-diverting activities (e.g., small games) or promote new contingencies (i.e., drinking coffee with reading, after eating/drinking with walking or talking to a friend).

As with habitual activities, perceived social norms and reactions to social cues for smoking form gradually over time. Persistent support may be useful to break the links between social cues and smoking behavior. For example, regular reminders about smoke-free social norms might lower expectancy of smoking in social scenarios. Self-regulation theory suggests smokers need either to avoid risk situations (action planning) or to develop cognitive or behavioral coping responses (coping planning) aimed at inhibiting the undesired behavior of smoking (Sniehotta et al., 2005). Avoidance of every social cue to smoke is impractical, so mHealth interventions should provide tips about how to cope with risky social situations.

For example, the “Smart-T” app randomly assessed low-SES users’ risk level for smoking lapse five times a day and sent support messages tailored to their risk level and self-reported presence of smoking cues (e.g., cigarette availability, other people smoking) for 3 weeks. Server-triggered tailored messages reduced lapse risk factors (Hébert et al., 2018). Similarly, the MobileCoach Tobacco program automatically sent messages to help users change their habitual behaviors and manage social cues for smoking (e.g., self-regulatory skills, norms, social support, how to cope with situational cues) based on their quit stage for three months. The self-report 7-day abstinence rate at follow-up was 13.9% (Haug et al., 2017). These server-triggered interventions were feasible and well accepted. However, limited evidence is available to show efficacy.

The advantages of server-triggered strategies include (1) fixed schedules and preprogrammed settings that are easy to maintain and scalable, (2) support based on each individual’s personal schedule, and (3) capacity for innovative forms of support (e.g., interactive games, multimedia training materials, social support networks/peer support communities). The challenges include (1) finding a sweet spot between enough messages to achieve desired outcomes and not so many messages that psychological reactance arises and (2) expecting a relatively long time before the intervention begins to overcome long-established habitual behaviors and perceived social norms.

Context-Triggered mHealth Interventions to Counteract Environmental Cues for Smoking

Results from our study suggest that, although smoking behaviors are often linked with environmental risk factors, smokers may not be aware of environmental cues to smoke, and these cues thus influence smokers subconsciously. Context-triggered mHealth strategies can help bring these cues into consciousness. The strategy relies on sensors (i.e., motion sensors as well as geographic sensors) that monitor smokers’ dynamic environments and deliver support when they detect a predefined risky situation.

Previous literature and our current findings provide clear evidence that smoking episodes are context-driven (Cerrada et al., 2017), thus successful interventions need cessation strategies tailored to different contexts. Context-triggered JITAIs can potentially effectively disrupt links between smoking and environmental cues (e.g., tobacco stores, common smoking locations, and exposure to advertisements) by recognizing and adapting to the changing contexts and providing just-in-time support. mHealth smoking cessation programs can be programmed with universally potent cues (e.g., geolocations of stores that sell tobacco) as well as individual environmental cues (e.g., workplace, coffee shops) and deliver intervention messages appropriate to each type of location. Context-triggered interventions only send messages to smokers when they are closer or exposed to environmental cues to smoke. Although a few studies have used this innovative method to design interventions for smoking cessation, more evidence is needed to demonstrate its feasibility and efficacy.

There is a lack of theoretical frameworks to guide the design of context-triggered smoking cessation interventions, partially because few theories can capture the dynamic behavioral change processes and momentarily changing contextual factors (Klasnja et al., 2015; Riley et al., 2011). Ecological models and social cognitive theory generally point out the importance of environmental factors in influencing behavior but do not suggest specific ways to reduce such impacts. Self-regulation theory and self-determination theory may be more applicable. The former suggests making cessation goals more salient and helping smokers develop cognitive or behavioral coping strategies for environmental cues. The latter suggests enhancing perceived autonomy in controlling one’s behaviors and making choices of the environments, perceived competence in exerting necessary skills for behavior change, and perceived relatedness can increase smokers’ intrinsic motivations for smoking cessation and resist the adverse environmental cues for smoking (Deci & Ryan, 1985). However, more dynamic models are needed to capture the continuously changing contexts and make more precise predictions of behaviors in real time (Spruijt-Metz & Nilsen, 2014).

Mobile health interventions using context-triggered strategies for smoking cessation are still in the pilot testing phase. An example is the Q Sense app, which recorded users’ smoking locations and automatically sent support messages when it detected that the user had stayed in these locations for more than 5 minutes. Support messages triggered by the momentarily changing context were found to be a feasible strategy and evaluated positively by the users (Naughton et al., 2016). Since current studies are limited and have only been tested with a small sample size, it is hard to make any conclusions about the efficacy of context-triggered interventions for counteracting environmental cues of smoking.

Advantages of context-triggered strategies include helping smokers recognize environmental risk factors for smoking and relapse that they may have overlooked and responding automatically in real time to real-world situations. These advantages may be particularly helpful for populations that are more exposed to environmental cues to smoke and cannot easily avoid the exposure. Potential disadvantages include privacy concerns about collecting user data, over- or desensitizing smokers to environmental cues to smoke, and high cost in maintaining the database of environmental cues.

The pros and cons and theoretical constructs related to each strategy are summarized in Table 3.

TABLE 3.

Summary of mHealth Strategies for Addressing Multilevel Smoking Cues

mHealth strategies Smoking cues to counteract Related theoretical frameworks Examples Empirical evidence Pros and cons
User-triggered Internal cues • Mood management theory
• Social cognitive theory (to increase participants’ motivations and self-efficacy in managing one’s internal states and resisting the temptation of smoking)		 Users open an app or website; call the Quitline; text a chatbot for meditation, mindfulness, positive psychology exercises, coping strategies, information, advice, and other support • The Smiling Instead of Smoking app has demonstrated feasibility and acceptability. Users completed positive psychology exercises for 3 weeks. (Hoeppner et al., 2019).
• In the txt2stop program, 38.5% of users texted “crave” or “lapse” to request for support. Users who texted “crave” in the first month were more likely to report abstinent at 1 month (Devries et al., 2012).		 Pros:
• Provide instant help
• Provide tailored support
• Scalable to large population
• Easy to maintain
• Cost-effective
Cons:
• Need motivations to initiate the program
• Hard to accurately detect the implicit feelings
Server-triggered Habitual cues and social cues • Classic conditioning (to break the link between a habit and smoking behavior)
• Social norms theory
• Self-regulation theory (action planning and coping planning)		 Text messages, notifications, or alerts that are automatically sent to users for smoking status monitoring, risk assessments, behavioral change guidance, coping strategies, education about the non-smoking norm, and cessation support • The Smart-T app assessed low–socioeconomic status users’ risk level for smoking lapse 5 times/day and sent support messages tailored to their risk level and self-reported presence of triggers for 3 weeks. It demonstrated feasibility and reduced risk factors (Hébert et al., 2018).
• The MobileCoach Tobacco program automatically sent cessation messages to users tailored to their quit stage for 3 months. The self-report 7-day abstinence rate at follow-up was 13.9% (Haug et al., 2017).		 Pros:
• Preprogrammed automatic delivery of support
• Provide personalized assistance based on users’ habits, past behaviors, and stage of change
• Scalable to large population
• Cost-effective
Cons:
• Psychological resistance
• Need long-term support
Context-triggered Environmental cues • Social cognitive theory (raise awareness of environmental risk factors)
• Self-regulation theory (coping planning)
• Self-determination theory (autonomy, competence, relatedness)
• Dynamic models		 The app tracks users’ geo-locations that are flagged as risky (e.g., high density of tobacco stores, presence of tobacco products, or usual smoking places) and automatically sends support messages when the user enters or is close to these locations. • The Q Sense app identified users’ most
frequent smoking locations and automatically sent support messages when the user stayed in these locations for more than 5 minutes. Support messages triggered by geo-locations were feasible and evaluated positively by the users (Naughton et al., 2016).		 Pros:
• Adaptive to each user’s real-world contexts with momentary assessments
• Particularly helpful for underrepresented populations who face greater environmental risks of smoking
Cons:
• Privacy concerns
• May prime people to think about smoking-related concepts and induce craving
• Need to constantly monitor and update the cues
• High maintenance cost
• Scalability dependent on databases of environmental cues
Open in a new tab

Limitations and Future Directions

The small sample size in our pilot study limits the generalizability of findings. Additionally, there is a geographical limitation on the sample. Because our study was contextualized in low-income urban communities in North Aurora, Colorado, findings may not be generalizable to other settings (e.g., suburban or rural areas). However, the advantage of focusing on a specific disadvantaged area is to have an in-depth understanding of and quickly identify socioenvironmental challenges faced by low-SES smokers living in this neighborhood. This formative research was intended as a first step to inform design and development of effective mHealth interventions for smoking cessation. Future work can design user-informed mHealth interventions that integrate user-triggered, server-triggered, and context-triggered strategies to deliver real-time support; test the feasibility of these interventions, and conduct randomized controlled trails to determine efficacy, especially among socioeconomically disadvantaged smokers.

Our ecological model of smoking cues does not yet articulate the role of policy factors, and we are uncertain whether or how mHealth interventions designed for individual smokers would respond to policy changes, although mHealth interventions could collect data that could inform regulation of tobacco-related risk factors in our environments.

Mobile health interventions have great promise to deliver support to low-SES populations in resource-limited settings as it can provide autonomous, sustainable, and cost-effective health interventions (Wallis et al., 2017). Studies have shown that low-SES smokers were receptive to tech-based smoking cessation support in the form of text messaging (Boland et al., 2017), indicating that this population is motivated and interested in adopting mobile-based health support. However, whether other forms of mHealth interventions are acceptable by low-SES users is less investigated. The potential barriers to utilizing mHealth interventions in low-SES populations should be considered from both resource and user perspectives. From the resource and infrastructure perspective, low-SES populations might not be able to afford high-quality network services or the high volume of cellular data, which are essential to enable advanced mobile technologies, such as real-time monitoring of health status. This is a concern especially when the intervention is designed to automatically deliver server- or context-triggered support. From the users’ perspective, barriers may include lack of engagement, low mobile literacy, and cultural differences. Socioeconomically disadvantaged users, especially those with lower literacy or older age, are less likely than higher SES users to engage with mobile devices for life management or to obtain resources (Lee & Kim, 2014). Therefore, low-SES populations may need more instructions and take more time to fully utilize the various functions of mobile-based health applications. Moreover, as SES is closely intertwined with ethnicity, ethnic minority users may need culturally tailored prevention interventions. These call for the more targeted design of mHealth interventions for disadvantaged populations with appropriate literacy level, engaging features, and cultural adaptations to improve their usability and effectiveness. Specific to the topic of smoking cessation, future research should investigate the influence of different cue types on smoking behaviors by age and ethnicity, to inform tailored design of mHealth interventions targeting specific populations. In addition, most existing mHealth interventions for smoking cessation were designed to address one or two types of smoking cue. Research is encouraged to design and test integrated interventions based on the ecological model of smoking cues proposed in this study, providing real-time support to help smokers cope with all four types of smoking cues.

CONCLUSIONS

Results of our pilot study suggest a multilevel ecological model best describes cues to smoke, and that the levels include (1) internal cues, which reflect individual’s physiological and psychological status; (2) habitual cues, which reflect conditioned smoking responses with certain activities; (3) social cues, which come from social interactions, perceived social norms, and cultural influences; and (4) environmental cues in the physical environment. This ecological model can guide the design of effective mHealth smoking cessation interventions that integrate user-triggered, server-triggered, and context-triggered strategies to address smoking cues from all four levels.

Authors’ Note:

We are grateful to Veronica Parra Mendoza for her assistance with conducting interviews, guided tours, and content analysis coding. The authors have no conflicts of interest to report. This research was funded by American Cancer Society Institutional Research Grant 16-184-56. MOG was supported by NIH grant K23-HL131939 and by a Pilot Study Research Grant from Denver Health and Hospital Authority.

REFERENCES

  1. Babb S, Malarcher A, Schauer G, Asman K, & Jamal A (2017). Quitting smoking among adults—United States, 2000–2015. MMWR Morbidity and Mortality Weekly Report, 65, 1457–1464. 10.15585/mmwr.mm6552a1 [DOI] [PubMed] [Google Scholar]
  2. Bandura A. (2002). Social cognitive theory of mass communication. In Bryant J & Zillmann D (Eds.), LEA’s communication series. Media effects: Advances in theory and research (pp. 121–153). Lawrence Erlbaum. [Google Scholar]
  3. Boland VC, Mattick R, Mcrobbie H, Siahpush M, & Courtney R (2017). “I’m not strong enough; I’m not good enough. I can’t do this, I’m failing”: A qualitative study of low-socioeconomic status smokers’ experiences with accessing cessation support and the role for alternative technology-based support. International Journal for Equity in Health, 16, 196. 10.1186/s12939-017-0689-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Boland VC, Stockings EA, Mattick RP, McRobbie H, Brown J, & Courtney RJ (2018). The methodological quality and effectiveness of technology-based smoking cessation interventions for disadvantaged groups: A systematic review and meta-analysis. Nicotine & Tobacco Research, 20(3), 276–285. 10.1093/ntr/ntw391 [DOI] [PubMed] [Google Scholar]
  5. Brown-Johnson CG, England LJ, Glantz SA, & Ling PM (2014). Tobacco industry marketing to low socioeconomic status women in the USA. Tobacco Control, 23(e2), e139–e146. 10.1136/tobaccocontrol-2013-051224 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Buller DB, Borland R, Bettinghaus EP, Shane JH, & Zimmerman DE (2014). Randomized trial of a smartphone mobile application compared to text messaging to support smoking cessation. Telemedicine and e-Health, 20(3), 206–214. 10.1089/tmj.2013.0169 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Burns MN, Begale M, Duffecy J, Gergle D, Karr CJ, Giangrande E, & Mohr DC (2011). Harnessing context sensing to develop a mobile intervention for depression. Journal of Medical Internet Research, 13(3), e55. 10.2196/jmir.1838 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cantrell J, Pearson JL, Anesetti-Rothermel A, Xiao H, Kirchner TR, & Vallone D (2015). Tobacco retail outlet density and young adult tobacco initiation. Nicotine & Tobacco Research, 18(2), 130–137. 10.1093/ntr/ntv036 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Centers for Disease Control and Prevention. (2018). Current cigarette smoking among adults—United States, 2017. MMWR Morbidity and Mortality Weekly Report, 67(44), 1225–1232. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cerrada CJ, Dzubur E, Blackman KC, Mays V, Shoptaw S, & Huh J (2017). Development of a just-in-time adaptive intervention for smoking cessation among Korean American emerging adults. International Journal of Behavioral Medicine, 24(5), 665–672. 10.1007/s12529-016-9628-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Conklin CA (2006). Environments as cues to smoke: Implications for human extinction-based research and treatment. Experimental and Clinical Psychopharmacology, 14(1), 12–19. 10.1037/1064-1297.14.1.12 [DOI] [PubMed] [Google Scholar]
  12. Conklin CA, Robin N, Perkins KA, Salkeld RP, & McClernon FJ (2008). Proximal versus distal cues to smoke: The effects of environments on smokers’ cue-reactivity. Experimental and Clinical Psychopharmacology, 16(3), 207–214. 10.1037/1064-1297.16.3.207 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Conklin CA, Salkeld RP, Perkins KA, & Robin N (2013). Do people serve as cues to smoke? Nicotine & Tobacco Research, 15(12), 2081–2087. 10.1093/ntr/ntt104 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Conklin CA, & Tiffany ST (2001). The impact of imagining personalized versus standardized urge scenarios on cigarette craving and autonomic reactivity. Experimental and Clinical Psychopharmacology, 9(4), 399–408. 10.1037/1064-1297.9.4.399 [DOI] [PubMed] [Google Scholar]
  15. Cox LS, Okuyemi K, Choi WS, & Ahluwalia JS (2011). A review of tobacco use treatments in US ethnic minority populations. American Journal of Health Promotion, 25(5 Suppl.), S11–S30. 10.4278/ajhp.100610-LIT-177 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Deci EL, & Ryan RM (1985). Intrinsic motivation and self-determination in human behavior. Plenum. [Google Scholar]
  17. Devries KM, Kenward MG, & Free CJ (2012). Preventing smoking relapse using text messages: Analysis of data from the txt2stop trial. Nicotine & Tobacco Research, 15(1), 77–82. 10.1093/ntr/nts086 [DOI] [PubMed] [Google Scholar]
  18. Emmons K. (2000). Health behaviours in a social context. In Berkman L & Kawachi I (Eds.), Social Epidemiology (pp. 242–266). Oxford University Press. [Google Scholar]
  19. Free C, Phillips G, Watson L, Galli L, Felix L, Edwards P, Patel V, & Haines A (2013). The effectiveness of mobile-health technologies to improve health care service delivery processes: A systematic review and meta-analysis. PLoS Medicine, 10(1), Article e1001363. 10.1371/journal.pmed.1001363 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Green J, & Thorogood N (2018). Qualitative methods for health research. Sage. [Google Scholar]
  21. Guest G, Bunce A, & Johnson L (2006). How many interviews are enough? An experiment with data saturation and variability. Field Methods, 18(1), 59–82. 10.1177/1525822X05279903 [DOI] [Google Scholar]
  22. Gurman TA, Rubin SE, & Roess AA (2012). Effectiveness of mHealth behavior change communication interventions in developing countries: A systematic review of the literature. Journal of Health Communication, 17(Suppl. 1), 82–104. 10.1080/10810730.2011.649160 [DOI] [PubMed] [Google Scholar]
  23. Hardeman W, Houghton J, Lane K, Jones A, & Naughton F (2019). A systematic review of just-in-time adaptive interventions (JITAIs) to promote physical activity. International Journal of Behavioral Nutrition and Physical Activity, 16(1), Article 31. 10.1186/s12966-019-0792-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Haug S, Castro RP, Kowatsch T, Filler A, & Schaub MP (2017). Efficacy of a technology-based, integrated smoking cessation and alcohol intervention for smoking cessation in adolescents: Results of a cluster-randomised controlled trial. Journal of Substance Abuse Treatment, 82, 55–66. 10.1016/j.jsat.2017.09.008 [DOI] [PubMed] [Google Scholar]
  25. Hébert ET, Stevens EM, Frank SG, Kendzor DE, Wetter DW, Zvolensky MJ, Buckner JD, & Businelle MS (2018). An ecological momentary intervention for smoking cessation: the associations of just-in-time, tailored messages with lapse risk factors. Addictive Behaviors, 78, 30–35. 10.1016/jaddbeh.2017.10.026 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Heishman SJ, Singleton EG, & Pickworth WB (2008). Reliability and validity of a Short Form of the Tobacco Craving Questionnaire. Nicotine & Tobacco Research, 10(4), 643–651. 10.1080/14622200801908174 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Herman CP (1974). External and internal cues as determinants of the smoking behavior of light and heavy smokers. Journal of Personality and Social Psychology, 30(5), 664–672. 10.1037/h0037440 [DOI] [PubMed] [Google Scholar]
  28. Hoeppner BB, Hoeppner SS, Carlon HA, Perez GK, Helmuth E, Kahler CW, & Kelly JF (2019). Leveraging positive psychology to support smoking cessation in nondaily smokers using a smartphone app: Feasibility and acceptability study. JMIR mHealth and uHealth, 7(7), e13436. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Hoffman MA, & Driscoll JM (2000). Health promotion and disease prevention: A concentric biopsychosocial model of health status. In Brown SD & Lent RW (Eds.), Handbook of counseling psychology (pp. 532–567). Wiley. [Google Scholar]
  30. Hovell MF, Wahlgren DR, & Adams M (2009) The logical and empirical basis for the behavioral ecological model. In DiClemente RJ, Crosby R, & Kegler M (Eds.), Emerging theories and models in health promotion research and practice: Strategies for enhancing public health (2nd ed., pp. 415–449). Jossey-Bass [Google Scholar]
  31. James DC, & Harville C (2017). Barriers and motivators to participating in mHealth research among African American men. American Journal of Men’s Health, 11(6), 1605–1613. 10.1177/1557988315620276 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kirchner TR, Cantrell J, Anesetti-Rothermel A, Ganz O, Vallone DM, & Abrams DB (2013). Geospatial exposure to point-of-sale tobacco: Real-time craving and smoking-cessation outcomes. American Journal of Preventive Medicine, 45(4), 379–385. 10.1016/j.amepre.2013.05.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Klasnja P, Hekler EB, Shiffman S, Boruvka A, Almirall D, Tewari A, & Murphy SA (2015). Microrandomized trials: An experimental design for developing just-in-time adaptive interventions. Health Psychology, 34(Suppl.), 1220–1228. 10.1037/hea0000305 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Krishna S, Boren SA, & Balas EA (2009). Healthcare via cell phones: a systematic review. Telemedicine and e-Health, 15(3), 231–240. 10.1089/tmj.2008.0099 [DOI] [PubMed] [Google Scholar]
  35. Lazev AB, Herzog TA, & Brandon TH (1999). Classical conditioning of environmental cues to cigarette smoking. Experimental and Clinical Psychopharmacology, 7(1), 56–63. 10.1037/1064-1297.7.1.56 [DOI] [PubMed] [Google Scholar]
  36. Lee JH, & Kim J (2014). Socio-demographic gaps in mobile use, causes, and consequences: A multi-group analysis of the mobile divide model. Information, Communication & Society, 17(8), 917–936. 10.1080/1369118X.2013.860182 [DOI] [Google Scholar]
  37. Levinson AH (2017). Where the US tobacco epidemic still rages: Most remaining smokers have lower socioeconomic status. Journal of Health Care for the Poor and Underserved, 28(1), 100–107. 10.1353/hpu.2017.0012 [DOI] [PubMed] [Google Scholar]
  38. Litvin EB, & Brandon TH (2010). Testing the influence of external and internal cues on smoking motivation using a community sample. Experimental and Clinical Psychopharmacology, 18(1), 61–70. 10.1037/a0017414 [DOI] [PubMed] [Google Scholar]
  39. Marcolino MS, Oliveira JAQ, D’Agostino M, Ribeiro AL, Alkmim MBM, & Novillo-Ortiz D (2018). The impact of mHealth interventions: systematic review of systematic reviews. JMIR mHealth and uHealth, 6(1), e23. 10.2196/mhealth.8873 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. McLeroy KR, Bibeau D, Steckler A, & Glanz K (1988). An ecological perspective on health promotion programs. Health Education Quarterly, 15(4), 351–377. 10.1177/109019818801500401 [DOI] [PubMed] [Google Scholar]
  41. Morgenstern M, Sargent JD, Isensee B, & Hanewinkel R (2013). From never to daily smoking in 30 months: The predictive value of tobacco and non-tobacco advertising exposure. BMJ Open, 3(6), e002907. 10.1136/bmjopen-2013-002907 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Muñoz RF (2010). Using evidence-based internet interventions to reduce health disparities worldwide. Journal of Medical Internet Research, 12(5), e60. 10.2196/jmir.1463 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Nahum-Shani I, Smith SN, Spring BJ, Collins LM, Witkiewitz K, Tewari A, & Murphy SA (2017). Just-in-time adaptive interventions (JITAIs) in mobile health: key components and design principles for ongoing health behavior support. Annals of Behavioral Medicine, 52(6), 446–462. 10.1007/s12160-016-9830-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Naughton F. (2017). Delivering “just-in-time” smoking cessation support via mobile phones: current knowledge and future directions. Nicotine & Tobacco Research, 19(3), 379–383. 10.1093/ntr/ntw143 [DOI] [PubMed] [Google Scholar]
  45. Naughton F, Hopewell S, Lathia N, Schalbroeck R, Brown C, Mascolo C, McEwen A, & Sutton S (2016). A context-sensing mobile phone app (Q Sense) for smoking cessation: a mixed-methods study. JMIR mHealth and uHealth, 4(3), e106. 10.2196/mhealth.5787 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Pew Research Center (2019). Mobile fact sheet. https://www.pewinternet.org/fact-sheet/mobile/ [Google Scholar]
  47. Rabbi M, Pfammatter A, Zhang M, Spring B, & Choudhury T (2015). Automated personalized feedback for physical activity and dietary behavior change with mobile phones: a randomized controlled trial on adults. JMIR mHealth and uHealth, 3(2), e42. 10.2196/mhealth.4160 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Reid JL, Hammond D, Boudreau C, Fong GT, & Siahpush M, & ITC Collaboration (2010). Socioeconomic disparities in quit intentions, quit attempts, and smoking abstinence among smokers in four Western countries: Findings from the International Tobacco Control Four Country Survey. Nicotine & Tobacco Research, 12(Suppl. 1), S20–S33. 10.1093/ntr/ntq051 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Riley WT, Rivera DE, Atienza AA, Nilsen W, Allison SM, & Mermelstein R (2011). Health behavior models in the age of mobile interventions: are our theories up to the task? Translational Behavioral Medicine, 1(1), 53–71. 10.1007/s13142-011-0021-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Rodriguez D, Carlos HA, Adachi-Mejia AM, Berke EM, & Sargent JD (2013). Predictors of tobacco outlet density nationwide: A geographic analysis. Tobacco Control, 22(5), 349–355. 10.1136/tobaccocontrol-2011-050120 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Shiffman S, Gwaltney C, Balabanis M, Liu K, Paty J, Kassel J, Hickcox M, & Gnys M (2002). Immediate antecedents of cigarette smoking: An analysis from ecological momentary assessment. Journal of Abnormal Psychology, 111(4), 531–545. 10.1037/0021-843X.111.4.531 [DOI] [PubMed] [Google Scholar]
  52. Sniehotta FF, Schwarzer R, Scholz U, & Schüz B (2005). Action planning and coping planning for long-term lifestyle change: theory and assessment. European Journal of Social Psychology, 35(4), 565–576. 10.1002/ejsp.258 [DOI] [Google Scholar]
  53. Spruijt-Metz D, & Nilsen W (2014). Dynamic models of behavior for just-in-time adaptive interventions. IEEE Pervasive Computing, 13(3), 13–17. 10.1109/MPRV.2014.46 [DOI] [Google Scholar]
  54. Ubhi HK, Michie S, Kotz D, Wong WC, & West R (2015). A mobile app to aid smoking cessation: Preliminary evaluation of SmokeFree28. Journal of Medical Internet Research, 17(1), e17. 10.2196/jmir.3479 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. U.S. Census Bureau. (2018). Income and poverty in the United States: 2017. https://www.census.gov/library/publications/2018/demo/p60-263.html
  56. Vidrine DJ, Frank-Pearce SG, Vidrine JI, Tahay PD, Marani SK, Chen S, Yuan Y, Cantor SB, & Prokhorov AV (2019). Efficacy of mobile phone-delivered smoking cessation interventions for socioeconomically disadvantage individuals: A randomized clinical trial. JAMA Internal Medicine, 179(2), 167–174. 10.1001/jamainternmed.2018.5713 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Wallis L, Blessing P, Dalwai M, & Shin SD (2017). Integrating mHealth at point of care in low-and middle-income settings: the system perspective. Global Health Action, 10 (sup3), 1327686. doi: 10.1080/16549716.2017.1327686 [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Wang L, & Miller LC (2019). Just-in-the-moment adaptive interventions (JITAI): A meta-analytical review. Health Communication. Advance online publication 10.1080/10410236.2019.1652388 [DOI] [PubMed] [Google Scholar]
  59. Yu D, Peterson NA, Sheffer MA, Reid RJ, & Schnieder JE (2010). Tobacco outlet density and demographics: Analysing the relationships with a spatial regression approach. Public Health, 124(7), 412–416. 10.1016/j.puhe.2010.03.024 [DOI] [PubMed] [Google Scholar]
  60. Zillmann D (1988). Mood management through communication choices. American Behavioral Scientist, 31(3), 327–340. 10.1177/000276488031003005 [DOI] [Google Scholar]

RESOURCES