Type of Social Support for Smoking Cessation among Low-Income Postpartum Women

Abstract

Background:

Efficacious programs to sustain smoking cessation are limited for postpartum women, particularly for those who are low-income. Social support may help enhance cessation efforts. However, the specific types of support potentially associated with cessation success are not well evaluated. The purpose of this study was to assess the relationship between different types of social support and smoking relapse among low-income women in the postpartum phase.

Methods:

A secondary data analysis was completed using data from a single-arm pilot study assessing a smoking cessation intervention for low-income postpartum women. Women were recruited from Philadelphia Women, Infants, and Children clinics (N=106) and completed a baseline and 1-month follow-up survey. Multivariable logistic regression analyses assessed the relationship between types of social support and smoking relapse, specifically cessation-specific support (Ways of Quitting social support subscale) as well as general (non-smoking focused) social support and its subdomains [i.e., emotional/informational, tangible, affectionate, social positive support (Medical Outcomes Study Social Support Survey)].

Results:

Participants were predominantly non-Hispanic Black (64%), 33% lived with their significant other, and 71% had a household income below $15,000. Most participants had relapsed by the one-month postpartum assessment (77%). In the final regression models, cessation-specific support was significantly associated with staying smoke free at one month postpartum. However, neither general social support, nor its subdomains, were associated with staying smoke free.

Discussion:

Cessation-specific support may bolster women’s cessation attempts during the stressful postpartum period. Smoking cessation interventions should consider integrating cessation-specific support to decrease relapse rates among low-income postpartum women.

INTRODUCTION

Smoking during pregnancy is associated with adverse morbidity and mortality outcomes for both mother and child (Mund et al., 2013). Consequently, many women quit smoking during pregnancy to protect the health of their fetus or to avoid social stigma (Bauld et al., 2017). However, most women relapse after delivery (Tong et al., 2013) with low-income women experiencing the highest rates of relapse (50–90% by 1 year postpartum; Kia et al., 2018). This is likely due to increased stresses on low-income women after giving birth (e.g., increased financial constraints, stress of caring for a newborn) (Bauld et al., 2017). With limited safety evidence for pharmacologic treatment among pregnant and breastfeeding individuals (Kreyberg et al., 2019; McLafferty et al., 2016), postpartum women are often recommended to use cognitive-behavioral interventions or counseling to sustain smoking cessation efforts (McLafferty et al., 2016). Unfortunately, efficacious cognitive-behavioral smoking cessation interventions are limited for low-income postpartum women (Brown et al., 2019).

Social support from either informal (e.g., family, friends) or formal (e.g., healthcare providers) sources has demonstrated success in helping individuals stay smoke free (Westmaas et al., 2010) and may be a promising strategy for cognitive-behavioral interventions targeting postpartum women (Brown et al., 2019). Social support can be classified into different subdomains including emotional (e.g., expressing emotions or venting), appraisal (e.g., constructive feedback), informational (e.g., advice, information), and instrumental support (e.g., money, transportation) (Cohen, 2004; House, 1981). Typically, these support functions are general and non-specific to smoking cessation (e.g., emotional support to vent about the difficulties of caring for a newborn) (Westmaas et al., 2010). However, evidence suggests that social support may be more efficacious if social support strategies are tailored to be cessation specific in order to specifically support quit attempts (e.g., emotional support to vent about the difficulty of abstaining) (Westmaas et al., 2010). Quantitative studies that directly compare the effects of different domains of social support during the postpartum period are lacking, particularly for low-income women who experience numerous cessation-interfering stressors (Brown et al., 2019; Notley et al., 2015). This study completed a secondary data analysis to assess the relationship between (a) cessation-specific support and (b) general social support with smoking relapse among low-income postpartum women.

METHODS

Study population

The current study used survey data from a single arm pilot study that evaluated a text messaging-based smoking cessation intervention (Tagai et al., 2020). Women were recruited through flyers posted in Women, Infants, and Children (WIC) clinics in Philadelphia, Pennsylvania from 2011–2013. Women were eligible if they were (a) able to communicate in English, (b) 18 years or older, (c) in their third trimester or within ten days postpartum, and (d) had made at least one quit attempt of all forms of tobacco during their current pregnancy. Interested eligible participants (N=106) provided informed consent and completed a baseline survey. Participants received a cell phone with texting capability and instructions for using the phone. Participants received a $15 gift card at the time of enrollment and a $20 gift card after completion of the one-month survey. The study was approved by the Fox Chase Cancer Center Institutional Review Board.

Study design

Guided by the Cognitive-Social Health Information-Processing (C-SHIP) model (Miller & Diefenbach, 1998), a text messaging-based intervention was developed to address five psychosocial relapse factors: knowledge (i.e., provision of knowledge about the harmful effects of smoking); beliefs (i.e., increasing self-efficacy to avoid temptation); distress (i.e., decreasing distress related to new motherhood); decisional balance (i.e., decreasing the cons and increasing the pros of staying smoke free); and self-regulation (i.e., promoting smoking cessation strategies) (Wen et al., 2014). Participants received one text message per week prior to giving birth and then three system-initiated text messages per day for one month after giving birth—the total message library is 183 messages (26% knowledge, 23% distress, 20% decisional balance, 15% beliefs, 37% self-regulation). A total of 13 messages (7%) had content about seeking social support from others, including family, friends, and their doctor. Participants completed an in-person baseline survey via paper-and-pencil at the time of enrollment and a one-month follow-up survey via phone.

Measures

Demographic information, pregnancy experience, age of starting smoking, and social support variables were collected on the baseline survey. Cessation-specific social support was assessed using the 4-item social support subscale from the Ways of Quitting questionnaire (e.g., I spent more time with my family so I wouldn’t smoke, I tried to stop smoking with a friend) (Myers et al., 2007). General social support was assessed using the 19-item Medical Outcomes Study Social Support Survey total score and four subscales: General social support was subdivided as emotional/informational (e.g., someone whose advice you desire; 8 items), tangible (e.g., someone who takes you to the doctor; 4 items), affectionate (e.g., someone who shows you love and affection; 3 items), and positive social interaction (e.g., someone to do something enjoyable with; 3 items) (Ware & Sherbourne, 1992). Smoking status was assessed at each time point and participants were asked if they smoked every day, some days, or not at all. Participants were considered to have relapsed if they selected “every day” or “some days.” Following an intent-to-treat approach, those who did not respond to the item were considered relapsers. In addition, at the one-month assessment, participants were asked if they had “slipped” since giving birth (e.g., smoked one cigarette, even one puff). This was not considered a relapse if they were not smoking at the time of the one-month follow-up.

Data analysis

Descriptive statistics were completed to summarize the demographics of the study population. Bivariate analyses were completed using chi-square tests, Fisher’s exact test, and independent t-tests to identify sociodemographic and smoking variables associated with smoking status at one-month follow-up. Variables with a p-value of .20 or less were included as covariates in the multivariable logistic regression analyses. Logistic regression analyses were completed to assess the relationship between social support and relapse at one month, controlling for race/ethnicity and pregnancy was planned (ps < .20 in bivariate analyses). The models first included types of social support individually (i.e., only one social support variable included in the model along with the covariates). Then, final models were completed with all social support variables included: one model with cessation-specific support and the total general social support score and one model with cessation-specific support and the general social support subscales. Repeated measures ANOVA was also completed to evaluate social support from baseline to 1-month between those who relapsed and stayed smoke free. Finally, bivariate analyses were completed to assess a smoking lapse (e.g., “slipped” and smoked) with relapse at one month as well as the types of smoking support. Analyses were completed using SPSS version 26.0.

RESULTS

A total of 106 women were enrolled in the study and completed the baseline survey. Participants were predominantly low income (96% reported < $30,000/year in annual household income), Non-Hispanic Black (68%), and single (88%; Table 1). At baseline, participants were on average 26.84 years old and had an average of 1.27 children prior to their current pregnancy. Of the 106 women who enrolled in the study, 56 women completed the one-month smoking status item on the follow-up survey (53%). Non-responders to the one-month survey (47%) were considered to have relapsed. In the total sample, 23% (24 of 106) of women reported smoking abstinence at one month postpartum while 15% reported smoking every day (N = 16) and 15% reported smoking some days (N = 16). About half of the women (N = 11, 46%) who reported smoking abstinence had slipped at least once in the month since giving birth and was significantly associated with relapse at 1 month (p < .001) as well as cessation-specific support [t(104) = 1.954, p = .033].

Table 1.

Participant demographics and smoking history

Characteristic	All Participants (N=106) n(%) or M (SD)	Relapsed at 1 month (N=82) n(%) or M (SD)	Smoke free at 1 month (N=24) n(%) or M (SD)	p-value
Age	26.84 (5.53)	26.77 (5.62)	27.08 (5.34)	.808
Race/ethnicity				.038
Non-Hispanic Black	68 (64.2)	55 (67.1)	13 (54.2)
Non-Hispanic White	7 (16.0)	15 (18.3)	2 (8.3)
All other race/ethnicitiesa	21 (19.8)	12 (14.6)	9 (37.5)
Marital status				.758
Single	88 (83.0)	67 (81.7)	21 (87.5)
Married, separated, or divorced	18 (17.0)	15 (18.3)	3 (12.5)
Lives w/ significant other	35 (33.0)	27 (32.9)	8 (33.3)	.577
Significant other smokes	27 (36.5)	20 (35.7)	7 (38.9)	.510
Income				.740
$0–15,000	75 (70.8)	57 (69.5)	18 (75.0)
$15,001–30,000	27 (25.5)	21 (25.6)	6 (25.0)
$30,001–45,000	3 (2.8)	3 (3.7)	0 (0.0)
$45,001–75,000	0 (0.0)	0 (0.0)	0 (0.0)
$75,001+	1 (0.9)	1 (1.2)	0 (0.0)
Education				.754
Less than high school diploma	22 (20.8)	18 (22.0)	4 (16.7)
High school diploma	46 (43.4)	36 (43.9)	10 (41.7)
More than high school diploma	38 (35.8)	28 (34.1)	10 (41.7)
Number of children at baseline	1.27 (1.33)	1.26 (1.30)	1.33 (1.44)	.803
Pregnancy was planned	24 (22.6)	21 (25.6)	3 (12.5)	.141
Weeks free smoking at baseline	13.73 (10.75)	12.49 (11.21)	17.73 (8.10)	.010
Nicotine dependence	2.76 (1.95)	2.84 (1.98)	2.50 (1.84)	.453
Age started smoking	16.29 (3.93)	16.21 (3.90)	16.58 (4.10)	.682
Total years regularly smoking	8.55 (5.43)	8.85 (5.67)	7.50 (4.44)	.284
Cigarettes per day	9.31 (7.11)	9.65 (6.72)	8.12 (8.35)	.357
Used other tobacco products	20 (19.0)	14 (17.3)	6 (25.0)	.284

^aHispanic = 16, American Indian/Native American = 1, Asian American = 2

Participants who remained smoke free at one month postpartum reported receiving greater levels of all types of social support than those who relapsed (Supplemental Table 1). After controlling for race/ethnicity, pregnancy was planned, and weeks smoke free at baseline (ps < .20), cessation-specific social support was the only type of support significantly associated with remaining smoke free [OR = 1.65 (1.14 – 2.38)]. Similarly, after including all types of support in the final models, only cessation-specific support remained significant. In the model with cessation-specific support and the total general social support score, individuals with greater cessation-specific support had significantly greater likelihood of remaining smoke free [OR = 1.56 (1.07 – 2.27)] (Table 2). These results held in the model with cessation-specific support and the four general social support subscales with only cessation-specific support remaining significant [OR = 1.60 (1.10 – 2.31)] (Supplemental Table 2). Additional analyses were completed to confirm the univariate regression findings between type of support and remaining smoke free among those who completed the one-month survey (Supplemental Table 3). All types of support except for positive social interaction had similar findings to the univariate analyses (Supplemental Table 1) among the total sample. Positive social interaction was significantly associated with remaining smoke free in the total sample [OR = 1.91 (1.02 – 3.59)] but was no longer significant in the smaller sample of one-month respondents [OR = 1.99 (0.99 – 4.00)].

Table 2.

Logistic Regression Evaluating Cessation-specific Support and Total Functional Social Support for Smoking Relapse at 1 Month Postpartum

	OR	95% CI
Race/ethnicitya
Non-Hispanic Black	2.310	0.248–21.539
All other race ethnicities	6.398	0.616 – 66.416
Planned pregnancyb	0.497	0.113 – 2.177
Weeks smoke free at baseline	1.027	0.972 – 1.086
Cessation-specific support	1.560	1.071 – 2.272
Total general social support	1.753	0.739 – 4.158

^aReference group: Non-Hispanic White

^bReference group: No, did not plan pregnancy

The repeated measures analysis of social support at baseline and 1 month and relapse status at 1 month found significant differences by relapse status for cessation-specific support (F = 4.106, p = .048), total general social support (F = 6.54, p = .013), emotional/informational social support (F = 7.36, p = .009), affectionate social support (F = 4.26, p = .044), and positive social interaction support (F = 4.76, p = .034; Supplemental Table 4 and Supplemental Figure 1). Tangible social support was significantly decreased from baseline (M = 4.57) to 1 month (M = 4.27) for individuals who were still smoke free (F = 8.58, p = .008), but was still higher at 1 month than individuals who relapsed at 1 month (M = 4.06).

DISCUSSION

The majority of women in this study had relapsed at one month postpartum. However, those who reported cessation-specific support were more likely to have remained smoke free. This is consistent with research showing that smokers who receive cessation-specific support (e.g., expressing praise for the smoker’s efforts to quit) may be more likely to quit and sustain abstinence for one year and longer (Lawhon et al., 2009). Additional qualitative research shows that cessation-specific support from a partner or family (e.g., encouragement to not smoke) is linked to success in staying smoke free among postpartum women (Notley et al., 2015). Many women who quit smoking and receive cessation-specific support during pregnancy often lose this type of social support in the postpartum period (Notley et al., 2015). Further, the general social support typically received in this new phase (e.g., help with taking care of the newborn, financial assistance) may not be sufficient for women to overcome barriers to staying smoke free.

These barriers include social pressure to smoke, expecting to resume smoking postpartum (Notley et al., 2015), stress, and untreated craving (Notley et al., 2015; Tagai et al., 2020). Psychosocial factors that negatively influence smoking cessation, notably stress and other smokers in the home, may be amplified among low-income individuals (Garg et al., 2021), placing them at greater risk for smoking relapse (Allen et al., 2019). Indeed, Correa and colleagues found that low-income postpartum women more frequently cited stress as a factor for relapsing compared to women with higher incomes (Correa et al., 2015). On the positive side, our results suggest that cessation-specific support may help reduce these barriers through targeted social support such as reminders about their intentions to stay smoke free and the knowledge that others want them to succeed.

Cessation-specific support in a smoking cessation intervention can include assistance with problem solving, rewarding quitting, and helping the individual improve their cognitive and behavioral coping skills (Lawhon et al., 2009). Research has also demonstrated that low-income postpartum women reported decreased smoking relapse when counselors helped women identify specific actions that supported their efforts to quit (e.g., helping to get others to not smoke around the individual) (Hennrikus et al., 2010). Additionally, including partners who share the goal of remaining smoke free can be critical to the success of postpartum women’s smoke free attempts (Brown et al., 2019; Kedia et al., 2021; Notley et al., 2015; Phillips et al., 2021). Of interest, participants in the current study received some cessation-specific social support recommendations (7% of total text message content), such as seeking out support (e.g., “pick someone you trust to support you and help you stay smoke free”), rather than the provision of trained support or inclusion of partners or family in the intervention (Wen et al., 2014), which possibly reduced the intervention’s efficacy alongside additional psychosocial factors associated with relapse (e.g., increased smoking temptation among those who relapsed; Tagai et al., 2020). Overall, cessation-specific, actionable social support items may be a promising intervention for a population that may lack the targeted support needed during this stressful postpartum period. Interventions for low-income women should also consider the multilevel barriers to cessation including increased stress, higher rates of smoking among lower income individuals (i.e., potential for others in their life to smoke; Casetta et al., 2017), and barriers to cessation resources in healthcare settings (Piñeiro et al., 2019) when developing cessation-specific support interventions.

Implications and future research.

Our findings suggest that cessation-specific support may facilitate remaining smoke free after giving birth among low income women—a population generally lacking access to cessation-specific resources and support (Brown et al., 2019). The study is limited in its small sample size using convenience sampling for recruitment. Additionally, the study had high attrition at the one-month follow-up and participants may have underreported smoking relapse due to concerns about stigma and substance use in pregnancy. Similar telephone-based data collection methods in this study population have also experienced retention difficulties (Murphy et al., 2022). Alternate survey methods such as mailed paper-and-pencil surveys may reduce attrition given the high time demands participants most likely experience at one-month postpartum. However, the univariate analyses evaluating type of support among one-month survey responders, we found similar associations between type of support and smoking relapse and the difference in positive social support may be due to the small sample size. Overall, the results highlight the potential benefit of including cessation-specific support in smoking cessation interventions during a window of opportunity for this at risk population, before an initial relapse occurs. Future research should explore cessation-specific support as a feasible and useful strategy for helping low-income women maintain smoking abstinence, despite the new challenges of the postpartum period.

Abbreviations:

C-SHIP

Cognitive-Social Health Information-Processing

WIC

Women, Infants, and Children