Daily Adherence to Nicotine Replacement Therapy in Low-Income Smokers: The Role of Gender, Negative Mood, Motivation, and Self-Efficacy

Abstract

Background:

Low-income smokers may benefit from interventions promoting nicotine replacement therapies (NRTs), particularly for women, as women have worse smoking cessation outcomes than men. Little is known about factors that affect daily NRT adherence, such as negative mood, motivation, and self-efficacy, and whether gender moderates these associations.

Methods:

Secondary analysis of a randomized controlled trial comparing a novel, in session sampling of short-acting combination NRT intervention (In Vivo) to standard care smoking cessation behavioral counseling was performed. Low-income smokers (n = 62; M_age = 47.4; 42% female; 61% non-White) completed daily surveys using the Insight^™ mHealth platform for 55 consecutive days. Three types of NRT adherence were examined: dose amount of short-acting NRT, nicotine patch wear time, and combination NRT (cNRT) (combined nicotine patch and short-acting NRT).

Results:

Generalized multilevel models report same-day negative mood was associated with greater likelihood of nicotine patch adherence in men, but unrelated to women’s nicotine patch adherence. Same-day negative mood was associated with greater likelihood of cNRT adherence in men, but less likelihood in women. The relationship between same-day motivation and use of short-acting NRT varied by gender, but simple slope analyses revealed that motivation was unrelated to short-acting NRT use within either group. Same-day self-efficacy was related to an increased likelihood of nicotine patch adherence and cNRT adherence in women but unrelated to men’s adherence of either type.

Conclusions:

Future research should focus on gender differences in low-income smokers’ same-day negative mood, motivation, and self-efficacy processes on NRT adherence during smoking cessation interventions.

Introduction

Smoking remains the leading cause of preventable death and disability in the U.S. with about 12.5% of the population identifying as current smokers in 2020.¹ Additionally, smoking rates are approximately 50% higher in low income populations, such that the prevalence of cigarette use is 21.4% among individuals earning < $35,000 per year compared to those with household incomes of $35,000 −74,999 (15.7%), $75,000–99,999 (11.4%), and ≥ $100,000 (7.1%).² While there are several FDA-approved medications for smoking cessation, nicotine replacement therapy (NRT) remains one of the most widely available cessation products.³ NRT is available in multiple forms, including both long-acting and short-acting products, with many available over-the-counter (OTC).⁴ Combination NRT (cNRT) is often suggested to supply a steady release of nicotine in a 24-hr period (with patch) while also providing a fast-acting product to address cravings (e.g., lozenge).^5,6 Non-prescription NRTs are available in most drugstores and retail/grocery stores and are relatively inexpensive.⁷ Therefore, individuals without a primary physician or health insurance may still obtain the most popular forms of NRT. These factors have made NRT a more accessible resource for low-income smokers.⁸

Despite these benefits of NRT, low adherence rates are a significant barrier to successful smoking cessation with NRT. Specifically, one large-scale international survey reported about 70% of users prematurely quit use of NRTs.⁹ There are multiple cognitive factors unrelated to NRT products that affect adherence to NRT such as forgetfulness, self-efficacy for adherence, and motivation to quit, among others.^10–13 Specifically, higher motivation and self-efficacy for adherence are associated with a greater likelihood of adherence.¹¹ In one study examining adherence to the nicotine patch, motivation to quit smoking was associated with increased adherence and subsequently greater adherence also encouraged motivation.¹⁴ While interventions to increase medication adherence for smoking pharmacotherapies have had mixed success, building adherence self-efficacy is one way that has been suggested to address these issues.^15,16 Additionally, depressive symptoms and negative mood are important factors that negatively impact smoking cessation success^17,18 as well as adherence to treatment.^12,19 These variables and fluctuating factors could potentially impact NRT usage on a day-to-day basis, which may be further influenced by other variables such as income and gender.

Previous research suggests that women have worse smoking cessation outcomes than men, including higher risk of smoking-related morbidity and mortality.^20,21 Numerous factors have been suggested to explain the differential smoking behaviors observed for women and men. For example, depressive mood, anxiety, anger, and perceived stress are predictors of time to relapse for women, but not men.²¹ Additionally, women smoke to cope with negative mood more often than men.²² Moreover, men consistently report better adherence to smoking cessation medications.¹¹ In a meta-analysis of gender differences in cessation using nicotine patches, women had poorer responses to nicotine patches than men.²³ Nonetheless, women are more likely to seek assistance with quitting and use smoking cessation services,²⁴ so it is important to understand what specific factors impact their medication adherence in order to optimize treatment for them.^25,26

While extensive literature supports the broad impact of mood, motivation, and self-efficacy on NRT adherence, limited research has evaluated how these constructs differ by gender and how they change on a daily basis. Moreover, limited research has used daily assessments to measure NRT use, although daily assessments of NRT use are likely more accurate than standard measures of use because they limit recall bias associated with reporting on daily behaviors like nicotine patch use and short-acting NRT use.^27,28 This study is the first to examine the relationship between same-day negative mood, motivation, and self-efficacy in relation to daily NRT adherence, while also evaluating gender differences in these relationships. This study utilized data from a randomized controlled trial comparing NRT to standard care counseling for low-income smokers who completed daily surveys on a smartphone app assessing daily NRT use and other variables (i.e., mood, motivation, self-efficacy). The current study will examine the relationships between three types of daily NRT adherence (i.e., dose amount of short-acting NRT use, nicotine patch wear time [16 hours or more], and combination NRT (cNRT) use of nicotine patch [16 hours or more] and short-acting NRT [8 doses or more]) and same-day negative mood, motivation, and self-efficacy, as well as whether these relationships vary by gender. Three research questions were examined:

• R1: Is same-day negative mood related to lower NRT adherence and is this relationship present only in women?

• R2: Is same-day motivation related to greater NRT adherence and is this relationship present only in women?

• R3: Is same-day self-efficacy related to greater NRT adherence and is this relationship present only in women?

Materials and Methods

Participants

Low-income smokers (i.e., annual household income of approximately <150% above the poverty line in Birmingham, AL or <$22,260 as a single or <$45,570 for a family of four) were enrolled in a pilot randomized controlled trial that compared a novel, in person, NRT sampling intervention (In Vivo) and standard care (SC) smoking cessation behavioral counseling (see Cropsey et al.²⁹ for additional information on the study procedures). Participants in the In Vivo group completed four weekly in person sessions of sampling each of four different short-acting NRT products (i.e., gum, lozenge, inhaler, and nasal spray) while at the same time wearing the nicotine patch in weeks 1–4 (e.g., sampling gum while wearing the nicotine patch). A standard dosing titration schedule was used for the nicotine patch (i.e., participants who smoked ≥ 11 cigarettes per day received 20 mg of nicotine and those who smoked ≤ 10 cigarettes per day received 14 mg of nicotine). Participants then selected their short-acting NRT product of choice in week 5 to use in conjunction with the nicotine patch. The SC participants completed four in person sessions of behavioral smoking cessation counseling followed by a short-acting NRT selection to use in conjunction with the patch in week 5 based on written descriptions and pictures of the products.

Procedure

Daily Diary Procedures

During weeks 5–13 of the trial, participants (n = 62) were prompted to complete a daily survey on a study-provided smartphone using the established Insight^™ mHealth platform application.³⁰ Participants were trained on the smartphone and daily diary procedures during the week 5 visit and were asked to schedule a quit day during that week (i.e., before the start of week 6). The daily surveys were initiated on the study smartphones during the morning (within thirty minutes of waking), which aligns with sampling procedures employed in prior research to assess smoking cessation outcomes with daily diary methods.^31,32 Approximately 65% of the daily surveys were answered over the 55 study days (n = 2,199).

Measures

Negative Mood

In the daily diary assessments, participants were presented with the following prompt: “Mark the response that most applies to you RIGHT NOW. I feel….”. Eight items followed to assess negative mood: stressed, irritable, frustrated, sad, worried, miserable, depressed, and anxious. Participants were asked to rate the extent to which they agreed/disagreed with each of the eight items using a 5-point scale (1 = Strongly Disagree, 5 = Strongly Agree). The eight items were averaged into a mean score to represent daily negative mood, where higher scores correspond to greater negative mood ratings (α = 0.95). Intraclass correlation coefficients (ICCs) were calculated using linear multilevel models with no predictors in order to separate the variance of the daily diary variables into between- and within-person components. The ICC for average daily negative mood was 0.74, indicating that 74% of the variance in day-level negative mood is between individuals, whereas the remaining 26% of the variance is explained by within-individual variation over time.

Motivation

During the daily assessments, participants rated their same-day motivation to avoid smoking via the following item: “I am motivated to AVOID smoking” using a 5-point scale (1 = Strongly Disagree, 5 = Strongly Agree). The ICC for average daily motivation was 0.47, indicating that 47% of the variance in day-level motivation is between individuals and 53% is within individuals over time.

Self-Efficacy

In the daily surveys, participants also reported on their same-day levels of self-efficacy to quit smoking. Participants were asked: “How confident are you that you will be able to fully quit smoking?” using a 5-point scale (1 = Strongly Disagree, 5 = Strongly Agree). The ICC for average daily self-efficacy is 0.62, indicating that 62% of the variance in day-level self-efficacy is between individuals, whereas 38% is within individuals over time.

Daily NRT Adherence

During the daily assessments, participants self-reported their use of the short-acting NRT, in addition to their patch use on the previous day. Three types of NRT adherence were assessed in the current study because they each address adherence behavior differently. Moreover, prior research suggests that daily assessments of several types of NRT use (e.g., nicotine patch wear time and combination use of nicotine patch and gum) predict smoking abstinence.^27,28 First, the use of short-acting NRT (i.e., nicotine gum, lozenge, inhaler, and nasal spray) was measured as the sum doses (e.g., pieces of gum) of short-acting NRT use (range 0–30). Amount of short-acting NRT use was assessed with the item, “How much of (gum, lozenge, etc.) did you use yesterday?”. The ICC for average daily amount short-acting NRT use is 0.57, indicating that 57% of the variance in day-level amount of short-acting NRT use is between individuals and 43% is within individuals over time. Second, adherent patch use was measured as the number of hours the nicotine patch was worn (i.e., 16 hours or more, yes = 1, no = 0), assessed with the item, “How many hours did you wear your patch?”. The cutoff of ≥ 16 hours was chosen to represent daily adherence because it aligns with protocols regarding the hours of wear typical for nicotine patch dosing³³. The ICC for average daily adherent patch use is 0.55, indicating that 55% of the variance in day-level adherent patch use is between individuals and 45% is within individuals over time. Finally, the third daily NRT adherence variable measured adherence to combination NRT (cNRT) use. The hours/doses of nicotine patch use and short-acting NRT (respectively) were combined into a single measure to represent cNRT use during the previous day (i.e., 16 hours or more of patch use and 8 or more doses of short-acting NRT, yes = 1, no = 0). The measure of cNRT adherence was dichotomized in this way to reflect combined use that would be expected to produce a clinical effect. The ICC for average daily adherence to cNRT is 0.48, indicating that 48% of the variance in day-level adherence to cNRT is between individuals and 52% is within individuals over time.

Covariates

The Fagerström Test for Nicotine Dependence (FTND)³⁴ was used to assess subjective nicotine dependence at baseline. The 6-item scale was summed to reflect a FTND total score (range 2–10). Daily cigarette use was self-reported in the daily surveys with the item, “How many cigarettes did you smoke yesterday?”. Responses ranged from 0–41 cigarettes. Both the between-person version of daily cigarette use (i.e., each participant’s mean across the study days) and the within-person version of daily cigarette use (i.e., subtracting each participant’s mean from each day’s score to represent deviations from one’s usual level of daily cigarette use) were included as covariates in the models. Intervention status (0 = control group, 1 = treatment group), self-reported gender (0 = male, 1 = female), race/ethnicity (0 = White, 1 = non-White), current marital status (0=not married, 1=married), having any children (0=no, 1=yes), and highest level of education (1=less than high school, 2=high school graduate or GED, 3=more than high school) were also included in all models.

Statistical Analyses

We used generalized multilevel models to examine the associations between same-day negative mood, motivation, and self-efficacy and NRT use/adherence, as well as whether these relationships varied by gender. The equation below provides an example model, using same-day negative mood as a predictor of daily amount of short-acting NRT use.

$${\mathit{\text{ShortNRT}}}_{it}={\beta }_0+{\beta }_1\left({\mathit{\text{Female}}}_i\right)+{\beta }_2\left({\mathit{\text{NegativeMood}}}_{it}\right)+{\beta }_3\left({\mathit{\text{Female}}}_i*{\mathit{\text{NegativeMood}}}_{it}\right)+{\beta }_4\left({\overline{\mathit{\text{NegativeMood}}}}_i\right)+u_{0i}+u_{1i}\left({\mathit{\text{Time}}}_{it}\right)+e_{it}$$

In the equation above, the amount of short-acting NRT use (ShortNRT_it) for individual i at time t is predicted by gender (Female_i), negative mood reported during the same-day (NegativeMood_it), and the cross-level interaction between gender and negative mood (Female_i * NegativeMood_it), which tests whether the within-person association between same-day negative mood and short-acting NRT use is stronger or weaker based on gender. The model includes a random intercept (u_0i) (which captures the between-person variability in the amount of short-acting NRT use across all days) and a random slope for time [u_1i(Time_it)], the variance of which gives an estimate of how much the within-person association between time and short-acting NRT use varies between individuals. The within-individual residual e_it captures the time-specific deviation in amount short-acting NRT use from the predicted mean for individual i. The model also adjusts for the mean of negative mood experienced by each individual ($\overline{{\mathit{\text{NegativeMood}}}_i}$), which removes all between-person variance in the negative mood predictor and ensures that model parameters associated with NegativeMood_it describe within-person associations between same-day negative mood and short-acting NRT use. Linear models were used for predicting amount short-acting NRT use; logistic models were used for predicting binary adherent patch use and adherent cNRT use. Similar models were utilized for same-day motivation and self-efficacy. Significant interactions were followed by estimating simple effects of same-day negative mood (as well as motivation and self-efficacy) directly from the model for women and men.

Results

The descriptive statistics of the study population are presented in Table 1. Approximately 42% identified as female and the racial/ethnicity of the sample included 61% non-White and 39% White participants. Approximately 10% of the sample reported being married and 71% reported having children. Moreover, 37% of the sample completed a high school degree or GED. The mean age of the sample was approximately 47 years old.

Table 1:

Descriptive Statistics

Variables	Full sample	Women	Men
	Mean (SD)	Mean (SD)	Mean (SD)
Gender	--	41.94%	58.06%
Race
White	38.71%	42.31%	36.11%
Black	58.06%	50.00%	63.89%
American Indian	1.61%	3.85%	0.00%
Bi-racial	1.61%	3.85%	0.00%
Marital status (% married)	9.68%	11.54%	8.33%
Children (% yes)	70.97%	84.62%	61.11%
Education
Less than high school	17.74%	30.77%	8.33%
High school or GED	37.10%	19.23%	50.00%
More than high school	45.16%	50.00%	41.67%
Age	47.40 (10.50)	48.89 (10.16)	46.33 (10.62)
FTND total	6.12 (1.80)	5.48 (1.58)	6.57 (1.81)
Intervention group	43.55%	53.85%	36.11%
Daily cigarette use (# smoked)	6.10 (6.24)	5.89 (6.93)	6.26 (5.72)
Daily negative mood	2.18 (0.92)	2.23 (0.99)	2.15 (0.87)
Daily motivation	4.07 (0.89)	4.15 (0.93)	4.03 (0.87)
Daily self-efficacy	4.17 (0.91)	4.24 (0.93)	4.13 (0.89)
Daily amount short-acting NRT	8.21 (4.88)	7.44 (4.92)	8.66 (4.80)
Daily patch adherence (16 hrs)	70.15%	77.02%	65.67%
Daily cNRT adherence (16 hrs; 8 doses)	38.03%	36.37%	39.05%

Note: SD = standard deviation; FTND = Fagerström test for nicotine dependence.

• R1: Is same-day negative mood related to lower NRT adherence and is this relationship present only in women?

  Table 2 shows the associations between same-day negative mood on daily NRT use across the three types of NRT use/adherence (i.e., amount short-acting NRT use, adherent patch use, and adherent cNRT use). The conditional main effect of gender was significant for amount short-acting NRT use and for cNRT adherence. Women reported less short-acting NRT use (b = −2.51, p = 0.03) and were also 82% less likely to adhere to cNRT use for 16 hours or more (OR = 0.18, p = 0.04). Same-day negative mood was significantly associated with a greater likelihood of adherent patch use (OR = 2.23, p < 0.01) and adherent cNRT use (OR = 2.23, p = 0.02). However, these conditional main effects were qualified by significant negative mood by gender interactions for adherent patch use (OR = 0.30, p < 0.01) and adherent cNRT use (OR = 0.27, p < 0.01). Simple slope analyses revealed that same-day negative mood was significantly associated with a greater likelihood of nicotine patch adherence in men (OR = 2.23, p < 0.01), but not in women (OR = 0.66, p = 0.29). Moreover, same-day negative mood was associated with a greater likelihood of cNRT adherence in men (OR = 2.23, p = 0.02), but less likelihood in women (OR = 0.60, p = 0.02).

• R2: Is same-day motivation related to greater NRT adherence and is this relationship present only in women?

  The associations between same-day motivation with daily NRT use/adherence are reported in Table 3. The results report a conditional main effect of gender on daily amount short-acting NRT use (b = −2.74, p = 0.03) and for cNRT adherence (OR = 0.16, p = 0.04), where women were less likely to use both types of NRT. The only significant interaction to emerge between gender and same-day motivation was for amount short-acting NRT use (b = 0.48, p = 0.05). In decomposing the interaction, the relationship between same-day motivation and short-acting NRT use did not reach significance for women (b = 0.14, p = 0.38), nor for men (b = −0.34, p = 0.06).

• R3: Is same-day self-efficacy related to greater NRT adherence and is this relationship present only in women?

  Table 4 shows the relationships between same-day self-efficacy and NRT use/adherence. The conditional main effect of gender was significant for amount short-acting NRT use and for cNRT adherence. Women reported less short-acting NRT use (b = −2.59, p = 0.04) and were also less likely to be adherent to cNRT use (OR = 0.19, p = 0.04). However, the interaction between gender and same-day nicotine patch adherence was significant (OR = 2.12, p = 0.02). The decomposed interaction suggests that higher levels of self-efficacy were associated with a greater likelihood of same-day nicotine patch adherence in women (OR = 1.66, p = 0.02) but not in men (OR = 0.78, p = 0.27). Moreover, a significant interaction was reported for gender and same-day levels of self-efficacy on cNRT adherence (OR = 2.23, p < 0.01), where the decomposed interaction suggests that higher levels of self-efficacy were associated with a greater likelihood of same-day cNRT adherence in women (OR = 2.20, p < 0.01) but not in men (OR = 0.98, p = 0.92).

Table 2:

The Interaction between Gender and Same-Day Negative Mood and NRT Adherence

	Amount short-acting NRT	Adherent patch (16 hrs)	Adherent cNRT (16 hrs; 8 doses)
Fixed effects (Same-day)	b (SE)	OR (95% CI)	OR (95% CI)
Intercept	4.49 (3.19)	2.33 (0.01, 629.71)	0.03 (0.00, 3.27)
Female	−2.51* (1.14)	1.24 (0.27, 5.60)	0.18* (0.03, 0.90)
Negative mood (day-level)	0.37 (0.51)	2.23** (1.60, 3.12)	2.23* (1.14, 4.38)
Female × Negative mood (day-level)	−0.63 (0.55)	0.30** (0.13, 0.69)	0.27** (0.12, 0.59)
Negative mood (person-level)	−1.50* (0.62)	1.36 (0.45, 4.09)	0.79 (0.31, 2.03)
Day	−0.04** (0.01)	1.00 (0.98, 1.02)	0.96** (0.94, 0.99)
Random effects (Same-day)	Est. (SE)	Est. (SE)	Est. (SE)
Intercept	15.49** (3.62)	7.80** (2.14)	6.47** (1.93)
Day	0.01** (0.00)	0.00** (0.00)	0.00** (0.00)

Note: OR = Odds ratio; CI = Confidence interval; Est. = estimate; Significant estimates are in bold.

^**p < 0.01;

^*p < 0.05.

All models adjust for FTND total scores (Fagerström test for nicotine dependence), same-day # cigarettes smoked, intervention status, race, marital status, children, and education; Having children was significantly associated with daily adherence to cNRT (16 hrs; 8 doses). Sensitivity analyses including random effects for same-day negative mood did not change the reported findings.

Table 3:

The Interaction between Gender and Same-Day Motivation and NRT Adherence

	Amount short-acting NRT	Adherent patch (16 hrs)	Adherent cNRT (16 hrs; 8 doses)
Fixed effects (Same-day)	b (SE)	OR (95% CI)	OR (95% CI)
Intercept	4.25 (3.42)	2.90 (0.01, 737.91)	0.03 (0.00, 3.45)
Female	−2.74* (1.23)	1.23 (0.26, 5.84)	0.16* (0.03, 0.92)
Motivation (day-level)	−0.34 (0.18)	0.77 (0.51, 1.17)	0.96 (0.75, 1.22)
Female × Motivation (day-level)	0.48* (0.24)	1.42 (0.91, 2.21)	1.36 (0.89, 2.06)
Motivation (person-level)	−0.74 (0.93)	1.68 (0.42, 6.80)	0.69 (0.17, 2.86)
Day	−0.05** (0.01)	1.00 (0.98, 1.02)	0.96** (0.94, 0.99)
Random effects (Same-day)	Est. (SE)	Est. (SE)	Est. (SE)
Intercept	17.09** (3.97)	7.50** (2.06)	6.76** (1.99)
Day	0.01** (0.00)	0.00* (0.00)	0.00** (0.00)

Note: OR = Odds ratio; CI = Confidence interval; Est. = estimate; Significant estimates are in bold.

^**p < 0.01;

^*p < 0.05.

All models adjust for FTND total scores (Fagerström test for nicotine dependence), same-day # cigarettes smoked, intervention status, race, marital status, children, and education; Having children was significantly associated with daily use of amount short-acting NRT and daily adherence to cNRT (16 hrs; 8 doses). Sensitivity analyses including random effects for same-day motivation did not change the reported findings.

Table 4:

The Interaction between Gender and Same-Day Self-Efficacy and NRT Adherence

	Amount short-acting NRT	Adherent patch (16 hrs)	Adherent cNRT (16 hrs; 8 doses)
Fixed effects (Same-day)	b (SE)	OR (95% CI)	OR (95% CI)
Intercept	5.04 (3.38)	2.87 (0.01, 879.85)	0.04 (0.00, 4.65)
Female	−2.59* (1.20)	1.28 (0.26, 6.27)	0.19* (0.04, 0.95)
Self-efficacy (day-level)	−0.16 (0.23)	0.78 (0.51, 1.21)	0.98 (0.72, 1.34)
Female × Self-efficacy (day-level)	0.51 (0.30)	2.12* (1.14, 3.95)	2.23** (1.34, 3.72)
Self-efficacy (person-level)	0.23 (0.78)	1.36 (0.39, 4.83)	1.35 (0.40, 4.60)
Day	−0.04** (0.01)	1.01 (0.99, 1.03)	0.97** (0.95, 0.99)
Random effects (Same-day)	Est. (SE)	Est. (SE)	Est. (SE)
Intercept	15.66** (3.66)	8.06** (2.21)	6.53** (1.94)
Day	0.01** (0.00)	0.00* (0.00)	0.00** (0.00)

Note: OR = Odds ratio; CI = Confidence interval; Est. = estimate; Significant estimates are in bold.

^**p < 0.01;

^*p < 0.05.

All models adjust for FTND total scores (Fagerström test for nicotine dependence), same-day # cigarettes smoked, intervention status, race, marital status, children, and education; Having children was significantly associated with daily use of amount short-acting NRT and daily adherence to cNRT (16 hrs; 8 doses); Treatment group was significantly associated with daily use of amount short-acting NRT. Sensitivity analyses including random effects for same-day self-efficacy did not change the reported findings. However, the interaction between gender and self-efficacy (day-level) became significant in the model predicting amount short-acting NRT (b=0.88, p < 0.01).

Discussion

The current study used daily diary data to shed light on the relationship between daily cognitive factors that may impact NRT adherence—such as same-day negative mood, motivation, and self-efficacy—and whether these relationships are moderated by gender. The study findings support that these three cognitive factors are associated with NRT adherence differentially for women compared to men.

Most notably, the moderation of gender on the relationship between same-day negative mood and NRT adherence was evident for nicotine patch adherence and cNRT adherence. Upon decomposing the interactions, same-day negative mood was significantly associated with a greater likelihood of nicotine patch adherence in men but was unrelated to nicotine patch adherence in women. Moreover, same-day negative mood was associated with a greater likelihood of cNRT adherence in men, but less likelihood in women. These findings from the simple slope tests regarding the relationship between same-day negative mood and greater NRT adherence in men are interesting and warrant further discussion. Perhaps negative mood is related to withdrawal symptoms in men, which might also be associated with greater success with smoking abstinence among men, as supported in previous literature.^11,20,21,23 In turn, greater success with smoking abstinence may be a linked to greater NRT adherence in men.²³ Future research may wish to examine these associations more closely with daily reports of mood/NRT adherence in men. However, these findings are also consistent with previous literature indicating that negative mood in women might lead to reduced NRT adherence,³⁵ which could in turn contribute to women having more difficulty quitting. These findings suggest that just-in-time adaptive interventions (JITAIs) targeting negative mood could be especially useful to women, such as a behavioral activation treatments paired with NRT.³⁶

The moderation of gender on same-day motivation and self-efficacy and NRT use was less prominent, although a few important associations emerged. The relationship between same-day motivation and use of short-acting NRT varied by gender, but simple slope analyses revealed that motivation was unrelated to the use of short-acting NRT within either gender group. Additionally, the relationship between same-day self-efficacy and nicotine patch adherence and cNRT adherence varied by gender, respectively. Simple slopes revealed that same-day self-efficacy was related to an increased likelihood of daily nicotine patch adherence and cNRT adherence in women, but was unrelated to adherence of either type in men. However, there may be additional factors impacting the relationship between gender and daily motivation and self-efficacy on NRT adherence. For example, in a bupropion trial, greater effectiveness expectancies for pharmacotherapy among women were associated with higher motivation to quit.³⁷ Thus, it is possible that the women in the current study had lower effectiveness expectancies of the NRT, which could affect the association between daily motivation and medication adherence. Moreover, assessing motivation and self-efficacy once per day may not be relevant for predicting daily NRT adherence. Indeed, previous research has highlighted the importance of examining within-day variation in self-efficacy and motivation during a smoking quit attempt.³⁸ To our knowledge, no prior research has examined the relationship between variability in motivation and/or self-efficacy throughout the day and NRT adherence; however, morning, afternoon, and evening assessments of these cognitive factors may be associated with NRT adherence differently. Future research may seek to develop JITAIs that focus on increasing motivation and/or self-efficacy (e.g., when morning levels of these cognitive factors are low) to improve daily NRT adherence, especially in women.

Limitations

The participants enrolled in the current study were low-income smokers. Therefore, the results may not generalize to other populations of smokers using NRT to reduce smoking behaviors. Relatedly, since NRT was distributed for free in this pilot study, the findings on NRT adherence might not generalize to real-life adherence rates when low-income smokers purchase NRT out of pocket. Second, a standardized measure of depression was not administered at baseline, so we were unable to consider the possible role of clinically diagnosed depressive symptoms while assessing daily negative mood. Moreover, the items used to assess daily negative mood were not part of a standardized diagnostic scale; therefore, we were unable to examine whether fluctuations in symptoms of clinical depression across days in the study might impact NRT adherence. Third, although the current study reports evidence for the relationship between same-day negative mood, motivation, self-efficacy and various types of NRT adherence to vary by gender, subsequent simple slope analyses had limited power to detect smaller effects (i.e., whether negative mood, motivation, and self-efficacy were related to NRT adherence within each gender group). We suggest replication of these analyses in larger samples of low-income smokers. Fourth, although the use of daily diary methods to measure same-day associations between negative mood, motivation, self-efficacy, and NRT adherence is a strength of the current study, the sampling of additional surveys throughout the day would help to illuminate how daily fluctuations in the constructs under study might impact NRT adherence (e.g., how within-day fluctuations in self-reported motivation impact NRT adherence). Fifth, participants completed approximately 65% of the possible daily diary surveys in the current study; however, this compliance rate is comparable to previous research with low-income samples.³⁹ Indeed, low-income populations are more likely to report experiencing daily stressors than are other populations of smokers,⁴⁰ which may impact the number of successfully completed daily surveys. Future research should examine the relationships between same-day negative mood, motivation, self-efficacy and NRT adherence in other samples of smokers and with multiple daily assessments of these cognitive factors. Compliance rates for daily assessments could also be improved in low-income smokers by administering daily assessments during fixed times throughout the day (based on the personal schedules of the participants) to avoid prompting participants during work, school, and/or during other daily obligations. Finally, the current study was underpowered to explore mediational relationships between the study variables. Future research may wish to explore whether daily negative mood, motivation, and self-efficacy may mediate the association between gender and daily NRT adherence with larger samples of low-income smokers.

Conclusion

The results illustrate that daily negative mood, motivation, and self-efficacy are important factors to consider when measuring adherence to NRT in low-income populations, but that they operate differently in men and women. Future interventions may wish to target daily negative mood, motivation, and self-efficacy during quit attempts in order to improve NRT adherence and smoking cessation outcomes, especially among low-income female smokers.

• Low-income smokers may benefit from nicotine replacement therapies (NRTs).

• Daily negative mood, motivation, and self-efficacy may impact NRT adherence

• Daily negative mood, motivation, and self-efficacy impacts NRT adherence differently in men and women