Smoking Cessation in Black Adults: Differences by Sex and Age

Abstract

Introduction

Females have poorer smoking cessation outcomes than males, especially with nicotine replacement therapy (NRT)-aided attempts, perhaps due to ovarian hormones. In addition, Black adults experience higher rates of tobacco-related mortality and also experience important ovarian hormone differences. This is the first study to examine sex, age, and menopause differences in NRT-aided smoking cessation outcomes in an all-Black sample.

Aims and Methods

A secondary data analysis was conducted on non-Hispanic Black adults aged 18 + who smoked 5–30 cigarettes/day for at least 6 months. Participants were categorized by sex assigned at birth (female versus male) and age (< 45 versus ≥ 45 years) with females in the ≥ 45 age group restricted to those who were postmenopausal (defined as 1 year post menses per self-report). Smoking status 2 weeks post-quit date was assessed by cessation attempts, expired carbon monoxide-verified abstinence, and reduction in daily cigarettes. Statistical analyses utilized Mann–Whitney rank-sum tests, chi-square tests, and regression models controlling for baseline cigarettes/day.

Results

Participants (n = 340; 55.9% female) were, on average 53.4 (standard deviation: ± 11.6) years old. At baseline, males smoked significantly more cigarettes/day than females (13.9 ± 7.7 versus 11.9 ± 6.3, respectively; p-value = .031). Among those < 45 years of age, males (n = 30) had a greater reduction in cigarettes/day 2 weeks post cessation than females (n = 52; 11.3 ± 6.1 versus 7.9 ± 5.3, respectively; p-value = .008), but this difference became null after adjusting for baseline cigarettes/day (p-value = .989). No other significant differences were identified.

Conclusions

After adjusting for baseline differences in smoking rate, no significant sex differences in smoking cessation outcomes were observed in this all-Black sample. Additional research is needed to replicate this observation.

Implications

Despite known racial disparities and sex-based differences in NRT efficacy, little is known about sex differences in cessation among Black people who smoke. While we did not observe any differences by sex in smoking cessation outcomes within this all-Black sample, our results suggest that sex differences in the baseline level of cigarettes/day may play a critical role in cessation outcomes. Further research should replicate these observations and examine potential implications for cessation among Black people who smoke.

Introduction

Smoking combustible cigarettes persists as a public health challenge. Compared to males, females exhibit lower rates of successful long-term abstinence—attributed to multifaceted biological, psychological, and social factors.¹ Nicotine replacement therapy (NRT), a prominent, evidence-based treatment for cessation, has lower effectiveness in females compared to males.^2,3 In addition to sex differences, racial disparities in long-term abstinence outcomes exist. Smoking prevalence is not significantly different between White adults and Black adults, yet a larger proportion of Black adults are unsuccessful in long-term smoking cessation.^4,5 Moreover, Black adults are more likely to die from tobacco-related diseases compared to White adults.⁶ Currently, research on sex differences in Black people who smoke is lacking. Understanding these sex differences and racial disparities is necessary to close the gap in access to effective cessation interventions.

Smoking cessation varies across ages as older individuals tend to have greater nicotine dependence and poorer cessation outcomes.^7,8 This age disparity may be amplified in women,⁹ perhaps due to the unique challenges associated with the menopausal transition such as changes in mood, stress, and ovarian hormones—all of which are known to influence smoking cessation outcomes.^10,9 Interestingly, the role of ovarian hormones in smoking cessation outcomes in Black populations is also unexplored. Black females reach menopause at an earlier age compared to their White counterparts, which may further indicate racial differences in hormonal profiles and, thus, cessation outcomes.¹¹ Identification of sex-, race-, and age-specific smoking cessation outcomes will inform future personalized medicine approaches for smoking cessation.

This present study is a secondary data analysis utilizing data from a recently completed smoking cessation study among individuals who self-identified as Black or African American.¹² We sought to examine the relationship of sex, age, and self-reported menopausal status in this Black-only sample during NRT-aided smoking cessation. Due to differences in ovarian hormones during reproductive years versus later in life, we expected sex differences to be greater among those who are of reproductive age versus those who are older. Specifically, we hypothesized that females of reproductive age (i.e., < 45 years old) will show less favorable cessation-related outcomes as compared to age-comparable males. Conversely, we expected postmenopausal females and age-comparable males (i.e., ≥ 45 years old) to have similar cessation outcomes.

Methods

Study Sample

The parent study¹³ investigated an adaptive-based pharmacotherapy intervention for Black participants who were interested in quitting smoking. Recruitment occurred from May 2019 through June 2021. Participants were recruited through clinic- and community-based efforts, as well as mass and social media advertising. Eligibility criteria included non-Hispanic African American or Black adults over the age of 18 who smoked 5–30 cigarettes/day for longer than 6 months at the time of recruitment. Exclusion criteria included contraindications to smoking cessation pharmacotherapies. Additional details on recruitment and eligibility can be found elsewhere.¹³ For the current analyses, participants who were missing data on sex assigned at birth, age, or menopausal status (females only) were excluded.

Study Procedures

Participants who provided informed consent were enrolled and completed baseline. This included collection of demographics and smoking-related variables, as well as setting a quit date within 7 days, smoking cessation counseling, and initiation of a 24-hour 21 mg NRT patch. Two weeks later participants completed a follow-up visit including measurement of expired carbon monoxide (CO), self-report of number of cigarettes/day, psychosocial measures, and other smoking behaviors. Additional protocol details may be found elsewhere.¹³

Study Measures

Participants were classified into four groups based on sex assigned at birth (female versus male), age (18–44 years versus 45 years or older), and self-reported menopausal status within age groups. We selected an age cutoff of 45 given Black females and females who smoke enter menopause earlier than the typical age of 51.^14,11,15 These four groups included: premenopausal females (females < 45; ages 18–44 with menses and/or hormonal contraceptive use reported within the past year), age-comparable males (males < 45), as well as postmenopausal females (females ≥ 45; at least 45 years of age with onset of last menses at least 1 year since last menstrual period without the use of hormonal contraceptives per self-report at baseline) and age-comparable males (males ≥ 45).

Our outcome variables included three cessation-related outcomes at week 2. First, in line with the parent study,¹² abstinence was defined as those who have an expired carbon monoxide (CO) value of 5 ppm or less whereas those with a value of 6 ppm or greater were defined as smoking. Second, cessation attempts were classified by cessation counselors based on conversations with participants as either abstaining for at least 24 h in the last week or did not. Last, we evaluated the reduction in cigarettes/day in absolute terms by subtracting the self-reported number of cigarettes/day at the follow-up visit in week 2 from the reported number at baseline.

Additional descriptive variables included participants’ age when they began smoking regularly, time to first-morning cigarette, current use of menthol cigarettes (yes/no), cannabis use (yes/no), smoking urges (per the Questionnaire for Smoking Urges¹⁶), depressive symptoms (per the Patient Health Questionnaire—9¹⁷), anxiety (per the Generalized Anxiety Disorder—7¹⁸), and stress (per the Perceived Stress Scale¹⁹).

Statistical Analysis

Differences in smoking cessation by sex (males versus females) and by sex within each age group (females < 45 versus males < 45 and females ≥ 45 versus males ≥ 45) were assessed using Mann–Whitney rank-sum (Wilcoxon rank-sum) tests for continuous outcomes (reduction in cigarettes/day from baseline to week 2) and Pearson’s chi-square tests for categorical outcomes (CO-verified abstinence [yes/no] and quit attempt [yes/no]). Linear and logistic regression models were used to assess relationships between sex and cessation outcomes while controlling for baseline cigarettes/day. To test whether the effect of sex on smoking outcomes differed by age, we included sex-by-age interaction terms in regression models. To observe the effects of hormonal contraception on the results, a sensitivity analysis excluding individuals using hormonal contraceptives was also conducted. Analyses were performed in R version 4.4.0, and the significance level was defined as p-value < .05.

Results

Study Sample

Of 392 participants from the parent study, 52 were excluded for either menopausal status that did not align with age categories (n = 30) or missing data on cessation outcomes (n = 22). Overall, participants (n = 340) included 190 (55.9%) females and 258 (75.9%) were 45 years or older (Table 1). On average, males reported smoking more cigarettes/day than females (13.9 ± 7.7 versus 11.9 ± 6.3, respectively; p-value = .031). This sex difference was also observed in those under 45 (males < 45: 15.5 ± 7.1 versus females < 45: 10.6 ± 5.6, p-value = .002) but not in those aged 45 and older (males ≥ 45: 13.6 ± 7.9 versus females ≥ 45: 12.4 ± 6.5, p-value = .500). There were no other statistically significant differences by sex in demographic and baseline measures.

Table 1.

Participant Characteristic by Sex and Age (n = 340)

	Overall			Age < 45 years			Age ≥ 45 years
Characteristic mean (SD)	Female n = 1901	Male n = 1501	p-value	Female n = 521	Male n = 301	p-value	Female n = 1381	Male n = 1201	p-value
Current age	52.9 (12.5)	54.1 (10.4)	0.9002	35.2 (6.1)	36.9 (6.0)	0.2002	59.5 (6.1)	58.3 (5.7)	0.0812
Age started smoking	18.5 (5.5)	17.6 (5.3)	0.0622	17.3 (3.3)	16.1 (3.6)	0.1002	18.9 (6.1)	18.0 (5.6)	0.2002
Baseline cigarettes/day	11.9 (6.3)	13.9 (7.7)	0.0312	10.6 (5.6)	15.5 (7.1)	0.0022	12.4 (6.5)	13.6 (7.9)	0.5002
Minutes to first cigarette 1
Within 5 min	47% (89)	50% (75)	0.8003	52% (27)	57% (17)	0.5003,4	45% (62)	48% (58)	0.8003
6–30 min	34% (64)	33% (49)		33% (17)	23% (7)		34% (47)	35% (42)
31–60 min	10% (19)	11% (16)		6% (3)	13% (4)		12% (16)	10% (12)
After 60 min	10% (18)	7% (10)		10% (5)	7% (2)		9% (13)	7% (8)
Smokes menthols (Yes) 1	88% (167)	86% (129)	0.6003	92% (48)	93% (28)	>0.9003,4	86% (119)	84% (101)	0.6003
Marijuana use (Yes) 1	21% (40)	25% (37)	0.4003	21% (11)	23% (7)	0.8003	21% (29)	25% (30)	0.5003
Depressive Symptoms Score	5.4 (5.3)	4.4 (4.6)	0.1402	6.3 (5.6)	4.9 (5.3)	0.2002	5.1 (5.1)	4.3 (4.4)	0.4002
Anxiety Symptoms Score	4.8 (5.5)	4.1 (4.9)	0.3002	6.0 (5.7)	5.3 (5.4)	0.6002	4.4 (5.4)	3.8 (4.7)	0.5002
Perceived Stress Score	4.3 (3.5)	4.3 (3.2)	0.7002	5.2 (3.7)	4.3 (3.4)	0.3002	4.0 (3.4)	4.3 (3.2)	0.2002
Smoking Urges Score	32.5 (19.1)	32.7 (17.0)	0.6002	35.0 (17.5)	29.5 (14.9)	0.2002	31.6 (19.7)	33.5 (17.4)	0.2002

¹%(n).

²Wilcoxon rank-sum test.

³Pearson’s chi-squared test.

⁴Fisher’s exact test.

Cessation-Related Sex Differences in the Full Sample

Most participants reported quitting for at least 24 h (females: 63%, males: 65%) and a smaller portion with CO-verified abstinence (females: 36%, males: 29%; Table 2). A trend indicated males had a greater reduction in cigarettes/day (10.0 ± 6.8) compared to females (8.6 ± 6.1; p-value = .085). However, after adjusting for baseline cigarettes/day this difference was null (p-value = .726). There were also no significant differences in cessation-related outcomes by age group in sex-stratified analyses (Supplementary Table S1).

Table 2.

Smoking-Related Outcomes at Two Weeks Post Attempted Cessation by Sex and Age (n = 340)

	Overall				Age < 45 years				Age ≥ 45 years
Characteristic mean (SD)	Female n = 1901	Male n = 1501	Unadjusted p-value	Adjusted p-value2	Female n = 521	Male n = 301	Unadjusted p-value	Adjusted p-value2	Female n = 1381	Male n = 1201	Unadjusted p-value	Adjusted p-value2
Abstinence status (% Yes)	36% (69/190)	29% (43/150)	0.1403	0.337	29% (15/52)	27% (8/30)	0.8003	0.471	39% (54/138)	29% (35/120)	0.0933	0.138
Quit for 24 hours or more (% yes)	63% (119/190)	65% (97/150)	0.7003	0.352	58% (30/52)	63% (19/30)	0.6003	0.196	65% (89/138)	65% (78/120)	>0.9003	0.733
Reduction in cigarettes/day from baseline	8.6 (6.1)	10.0 (6.8)	0.0854	0.726	7.9 (5.3)	11.3 (6.1)	0.0084	0.991	8.9 (6.4)	9.6 (6.9)	0.5004	0.810

¹ %(n).

²Adjusted for baseline cigarettes/day.

³Pearson’s chi-squared test.

⁴Wilcoxon rank-sum test.

Cessation-Related Sex Differences in < 45 Years Old

Quit attempts were similar by sex within this age group (females < 45: 58% versus males < 45: 63%), as was CO-verified abstinence (females < 45: 29% versus males < 45: 27%). The male < 45 group had a significantly greater reduction in cigarettes/day compared to the female < 45 group (11.3 ± 6.1 versus 7.9 ± 5.3, respectively; p-value = .008), though adjustment for baseline cigarettes/day accounted for this group difference (p-value = .991). There were no significant in the sensitivity analyses excluding those who were using hormonal contraceptives (Supplementary Table S2).

Cessation-Related Sex Differences in ≥ 45 Years Old

There were no significant sex differences in quit attempts (females ≥ 45: 65% versus males ≥ 45: 65%), CO-verified abstinence (females ≥ 45: 39% versus males ≥ 45: 29%), or reduction in cigarettes/day (females ≥ 45: 8.9 ± 6.4 versus males ≥ 45: 9.6 ± 6.9) within this age group in both unadjusted and adjusted analyses (p-values > .100).

Discussion

This secondary data analysis of a Black-only sample explored differences in cessation-related outcomes by sex, age, and self-reported menopausal status. Overall, in those less than 45 years old, males smoked more cigarettes/day at baseline versus females. In addition, unadjusted analyses suggested that males had a larger reduction in cigarettes/day compared to females over the first 2 weeks of attempted smoking cessation, but when adjusted for baseline cigarettes/day this trend became null. In contrast to our hypotheses, we did not observe significant differences in cessation-related outcomes by sex, regardless of age and menopausal status.

Despite prior evidence suggesting lower NRT efficacy in females,^2,3 we did not observe any significant sex differences in cessation outcomes after adjusting for baseline cigarettes/day in this Black-only sample. While our observations are limited to the first 2 weeks of cessation, this is an important indicator for longer-term success.¹² Specifically, per the parent study, those who demonstrated CO-verified abstinence after 2 weeks had 4.6 greater odds of achieving abstinence at the end of treatment relative to those who did not achieve early treatment abstinence (28.7% versus 7.8%, p-value < .001). Similar to other findings,⁵ we observed that males smoked more cigarettes/day at baseline as compared to females. This sex difference is statistically significant in those who were of reproductive age but not in those who were older. Baseline cigarettes/day is a well-known predictor of smoking cessation outcomes²⁰ and should be considered in planning for smoking cessation. Moreover, although psychosocial and craving behaviors can also predict smoking behaviors differently by sex, these variables were comparable among males and females in our sample. Similar to our observations, other research has demonstrated that gender differences in mood during acute abstinence were more pronounced in non-Hispanic White populations versus non-Hispanic African American populations.²¹ More research needs to be done to determine if this lack of sex differences is commonly observed in a Black population and, if so, how this may inform smoking cessation approaches in this population.

Estradiol plays a complex role in smoking behaviors, with higher levels adversely affecting cravings and withdrawal symptoms.²² Premenopausal females typically experience hormonal fluctuations throughout the menstrual cycle which can be predictive of cessation outcomes.²³ While hormonal contraceptive use dampens estradiol levels and emerging evidence suggests they influence smoking-related behaviors,²⁴ we did not observe any sex differences in cessation outcomes in our main analyses or in the sensitivity analyses excluding those using hormonal contraception. Postmenopausal females, in contrast, have a low steady estradiol level,⁹ suggesting that postmenopausal females may have cessation outcomes similar to men. While we did observe similar cessation outcomes across sexes among those over 45 years of age, we were unable to verify menopausal status with objective measurements (e.g., ovarian hormone levels). Without these values, caution in interpreting these observations should be used as errors may be present. In line with recommendations for the evaluation of ovarian hormones in smoking behavior among women of reproductive age,²⁵ further research on the role of menopause in smoking-related outcomes should include direct measurement of hormones.

This is the first study to examine sex differences by age and menopausal status within a Black-only sample. However, several limitations are present. First, we have limited information to classify menstrual phase and menopausal status. For example, we had the first day of the last menstrual period collected at baseline only and, thus, were unable to identify phase menstrual phase at subsequent timepoints. We also do not have measured hormone values. Relatedly, it is possible that our group definitions (including the cut-off age of 45) may have led to a lack of identification of more nuanced sex differences that occur with age and misclassification may have occurred. In addition, our findings have limited generalizability due to the focus on the first two weeks, exclusive use of NRT, lack of a non-Black comparison group, and limited statistical power from the sample size. Finally, it must be considered that a few measures are self-reported usage. While expired CO was used to validate abstinence reports, other variables were not.

In summary, we did not observe any sex differences in cessation-related outcomes in this all-Black sample, regardless of age and menopausal status. However, cigarettes/day at baseline were greater among men than women in those of reproductive age. These observations should be replicated in future studies with more rigorous methodology and examining differing cessation approaches. Effective smoking cessation approaches need to be studied further in a Black population with consideration of sex and age.

Declaration of Interest

Dr. Nollen is a member of the scientific advisory board for Qnovia, a startup company that developed a smoking cessation medication for FDA approval.

Author Contributions

Ryan A. Park (Methodology, Visualization [supporting], Writing—original draft, Writing—review & editing [lead]), Alicia M. Allen (Conceptualization [lead], Methodology [supporting], Supervision [equal], Writing—original draft [supporting], Writing—review & editing [equal]), Devashri Prabhudesai (Formal analysis [lead], Visualization [equal], Writing—review & editing [equal]), Emmery E. Kelly (Methodology [supporting], Visualization, Writing—review & editing [equal]), Janet L. Rothers (Formal analysis, Supervision [supporting], Writing—review & editing [equal]), and Nikki Nollen (Conceptualization, Data curation [equal], Funding acquisition, Investigation, Methodology, Project administration [lead], Writing—review & editing [equal])

Data Availability Statement

Requests for data may be made to the parent study principal investigator (Dr. Nikki Nollen).