The role of sex hormones in targeting stress-induced tobacco craving, stress-reactivity, and smoking with guanfacine among women who smoke

Abstract

Women who smoke are particularly vulnerable to tobacco craving, smoking behaviors, and relapse in the context of stress when compared to men who smoke. One factor in this sex difference may be sex hormones, including estradiol and progesterone; however, smoking cessation medication trials often do not explore the impact of sex hormones on drug effects. This secondary analysis of a double-blind, placebo-controlled study explored the impact of levels of actual estradiol and progesterone on guanfacine, a noradrenergic α2a agonist, which attenuates stress-induced smoking behaviors in women. Women who smoke (n = 43) completed a stress induction laboratory paradigm followed by an ad-libitum smoking period. Assessment of tobacco craving, and stress-reactivity (via cortisol response) occurred pre- and post-stress induction. Results indicated that guanfacine attenuated stress-induced tobacco craving (F = 10.94, p = 0.02) and cortisol response (F = 14.23, p < 0.001); however, high levels of estradiol overrode guanfacine’s effect on craving (F = 4.00, p = 0.05), cortisol response (F = 14.23, p < 0.001), and smoking during the ad-libitum period (F = 12.23, p = 0.001). Additionally, progesterone proved to be protective against tobacco craving and enhanced guanfacine’s medication effect on craving (F = 5.57, p = 0.02). The present study found that sex hormones had a significant impact on medication effects in a smoking cessation trial and thus underscore the importance of examining the role of sex hormones in future medication trials.

1. Introduction

Cigarette smoking has steadily declined in recent decades; however, estimates indicate that 19% of adults, approximately 47 million individuals, in the United States (US) continue to currently use tobacco products [1] and approximately 14% of adult men and 11% of adult women in the US smoke cigarettes regularly [1]. While cigarette smoking is more prevalent among men than women, these figures remain problematic for women, as they have lower rates of quitting smoking [2-4], are more likely to relapse to smoking after quitting [5,6], and face increased health consequences [7] in comparison to their male counterparts.

One factor for such differences in smoking cessation and relapse is sex-related vulnerabilities in negative affect. It is well established that negative affect impedes smoking cessation in both men and women. Negative affect and stress increase smoking behavior, stress-reactivity, tobacco craving, and negative emotions in both sexes [8]; however, women are more vulnerable to relapse in the context of negative affect and stress than men who smoke [9-11]. Among women who smoke, depression is related to increased likelihood to relapse following a cessation attempt [12] and women are more likely than men to relapse to smoking when facing a stressful event [13]. Furthermore, in one study following a negative mood induction in a laboratory paradigm, women more rapidly began smoking as compared to men [5]. These results illustrate the unique relationship between stress and smoking for women.

The relationship between stress and smoking among women is likely impacted by fluctuations in endogenous sex hormones, including estradiol and progesterone. Progesterone is generally protective against positive subjective ratings of tobacco and cravings, [14-16] and also attenuates smoking [16]. Whereas, estradiol is related to increased smoking behaviors, including relapse, continued smoking, and cravings (for review see Wetherill et al. [17] and Schiller et al. [18]). Estradiol is also related to elevated sensitivity to stress in women who smoke. [17] Conversely, there is emerging evidence that progesterone and its metabolites may be protective against stress-induced affect (for review see Peltier et al. [19]). This is particularly important as it is postulated that during times of high progesterone and moderate estradiol (e.g., midluteal phase [20]) women may respond to treatments targeting craving. Whereas, when estradiol is high and progesterone is low, but increasing (e.g., mid/late-follicular phase [20]), women may have heightened emotional responsiveness and thus benefit from treatments focused on stress-reactivity [17].

However, the literature exploring the relationship between sex hormones and smoking is marred by inconsistent methodology defining menstrual cycle phases and lack of exploration of reproductive status (e.g., peri- or postmenopausal) [20]. Accordingly, recent recommendations suggest presenting sex hormone levels as absolute values, as estradiol and progesterone fluctuate daily and rise/fall within each phase, including subphases of the follicular and luteal phases [20]. It is increasingly important to explore the relationship between smoking and actual hormone levels, irrespective of the subphases of the menstrual cycle and across reproductive status (e.g., peri- and postmenopause), in order to better understand the relationship between sex hormones and smoking behavior.

Given these observed sex-specific vulnerabilities, it is unsurprising to find that there are sex differences in medication effectiveness on smoking cessation. For instance, women are less likely to have a successful cessation attempt using transdermal nicotine patches [2,21] and varenicline has shown efficacy to improve short-term cessation for women compared to bupropion or transdermal nicotine patches [22,23]; however, these medications do not directly target stress-related relapse, which is particularly salient for women who smoke. Emerging evidence suggests guanfacine, a noradrenergic α2a agonist, reduces smoking and craving by targeting stress-precipitated use via the noradrenergic system [24]. Notably, in a randomized clinical trial powered to explore sex differences, guanfacine attenuated the effect of stress-induced smoking in women who smoke and normalized stress-reactivity among women who smoke [25]. However, to the best of the authors’ knowledge, no study has explored the effect of actual levels of endogenous sex hormones on medications targeting stress-related smoking behavior.

1.1. The present study

Accordingly, the present study sought to explore the effect of estradiol and progesterone on guanfacine versus placebo medication among women who smoke. The present study sought to explore actual hormone levels as opposed to using proxies of reproductive states, including menstrual cycle subphases and/or peri- and postmenopause. It was hypothesized that estradiol levels would be associated with higher craving and smoking behavior, whereas progesterone levels would be associated with lower craving and smoking. Additionally, it was hypothesized that medication effects would be greater among individuals with higher levels of estradiol.

2. Material and methods

The present data were collected in a between-subjects, double-blind, placebo-controlled trial comparing guanfacine (3 mg/day immediate release) to a placebo control (0 mg/day). Following a 21-day titration to steady-state medication levels, participants (n = 43 women) completed a laboratory session designed to model stress-induced, smoking lapse. Information regarding procedures of the parent study is available else-where [24].

2.1. Participants

Participants were recruited from the community for a smoking laboratory study; all participants provided informed consent. Participants were between 18 and 60 years old. All participants had smoked at least 10 cigarettes per day for the past year, had urine cotinine levels ≥ 150 ng/ml, and had normal blood pressures (sitting blood pressure >90/60 and <160/100 mmHg). Participants were excluded from the study if they met criteria for DSM-5 psychiatric disorders, except for nicotine use disorder [26]. Participants were also excluded if they were using other substances (assessed by urine toxicology screen, including amphetamines, cocaine, opioids, phencyclidine, and methadone). Of note, participants were not excluded if they tested positive for cannabis but were excluded if they met DSM-5 criteria for a cannabis use disorder. Participants were excluded if they had engaged in smoking cessation treatment within the past six months, or had medical conditions/concurrent medication use that was contraindicated for guanfacine use/smoking behavior. Medical conditions were assessed via physical exam, which included an electrocardiogram and basic blood chemistries. The Yale Human Investigations Committee approved all study procedures.

2.2. Guanfacine treatment and placebo

Approximately 1–3 weeks following the intake assessment, participants were randomized (stratified by sex) to receive guanfacine or placebo. The trial was double-blinded, and the Yale Investigational Drug Service maintained the randomization schedule/blinding. Guanfacine was administered twice per day and titrated to steady-state levels over 21 days as follows: 0.5 g mg on days 1–3, 1.5 mg on days 4–7, 2 mg on days 8–12, 2.5 mg on days 13–15 and 3 mg on days 16–21). This dose was chosen as we had previously demonstrated that it reduces nicotine craving and increases the ability to resist smoking following stress [24]. Following titration, participants were maintained on 3 mg/day guanfacine or placebo for the duration of the study and completed the laboratory sessions, followed by 4-week treatment period which included weekly, brief behavioral support. Adverse events were assessed twice weekly during the titration period and weekly during the treatment phase of the study; all adverse events were minimal or mild and no participant discontinued or required a medication dose adjustment. Additional information on the treatment period and adverse events reported can be found in the parent study [24].

2.3. Laboratory assessment of stress-precipitated smoking lapse

Each participant completed a 6.5 h laboratory session, which was completed following the 21-day titration of guanfacine/placebo. The laboratory sessions started at 9:00am. Participants were instructed to smoke a final cigarette at 10:00pm the night prior to the laboratory session. Abstinence was confirmed via carbon monoxide (CO) reading. Abstinence was defined as having a CO reading that was less than half of the participant’s highest CO reading at previous study appointments; all participants who completed the laboratory assessment met abstinence criteria. At onset of session, an intravenous (IV) cannula was inserted to collect blood samples. Baseline assessments of breath CO, breath alcohol, urine drug screen, urine pregnancy screen, and vital signs were collected. In a previous study session (see McKee et al. [24] for description of full methods), participants completed assessments including cigarettes per day (CPD), the Fagerström Test of Nicotine Dependence (FTND [27]) and the Center for Epidemiological Studies Depression Scale (CES-D [28]), among other measures. Medication administration (1.5 mg or placebo) occurred at 10:00am the morning of the study. Participants were provided with a standardized lunch at 11:15am to control for time since last food consumption. From 10:00am to 12:30pm, participants were able to watch television or read.

In a prior study session, participants completed personalized, stress-induction scripts using guided-imagery methods [8]. These stress-imagery scripts were developed by asking participants to identify and describe highly stressful experiences occurring within the past six months. Only events rated an 8 or greater on a Likert scale (1 = “not at all stressful”; 10 = “the most stress they recently felt in their life”) were used for script development. Scripts were developed by a doctoral-level clinician and audiotaped for presentation during the laboratory session. At 12:55pm on the laboratory session day, participants listened to the stressful script via headphones.

Following completion of the audiotaped stress-induction script and a delayed smoking period (for description of procedures see McKee et al. [24]), participants began an ad-libitum smoking session for 60 min. During ad-libitum smoking, participants were provided with eight cigarettes from their preferred brand and were instructed to “smoke as little or as much as you wish. ” Participants were monitored via camera for the duration of the laboratory sessions. Cigarettes and the ashtray were inspected following the paradigm to ensure an accurate count of cigarettes smoked. Participants were discharged at 3:15pm.

2.4. Assessments

The primary measures included the number of cigarettes smoked during the ad-libitum period and tobacco craving (measured pre- and post-stress induction via Questionnaire of Smoking Urges-Brief [QSU] [29]). Women had a negative pregnancy test during screening and the morning of the laboratory session. Menstrual cycle phase and menopausal status were not controlled, and hormonal contraception was not exclusionary. Other measures were collected but are not included in this report (for description see McKee et al. [24]).

2.4.1. Biochemical assessments

Blood samples were taken the morning of the laboratory session to determine baseline hormone levels for the laboratory session, including estradiol and progesterone. Five milliliters (mL) of blood were collected at the onset of the laboratory session to assess serum estradiol and progesterone. Estradiol and progesterone samples were analyzed using the ELISA kits (ALPCO; estradiol sensitivity 10 pg/mL; progesterone sensitivity 0.1 ng/mL). Cortisol samples were collected −30, −15, +5, +20, +40, and +60 min post-stress induction. Five mLs of blood were collected at each timepoint to assess plasma cortisol and collected in heparinized tubes for each assay. Each tube was placed on ice immediately following drawing and centrifuged. Cortisol was analyzed with the ELISA kit (ALPCO; sensitivity 0.4 μg/dl). Biochemical samples are centrifuged, aliquoted, and stored at −80C. All assays were conducted by the Core Laboratory of the Yale Center for Clinical Investigation.

2.5. Statistical analysis

Baseline participant demographic characteristics were compared between guanfacine and placebo conditions via analysis of variance (ANOVA) and chi-square analyses. Repeated measures ANOVAs (utilizing generalized linear models; IBM SPSS Data Editor Version 28) were used to examine the within-subject effects of time (pre- versus post-stress induction) by effects of medication condition (guanfacine versus placebo) and estradiol and progesterone levels, on craving (QSU- Factor 1, QSU- Factor 2). Estradiol levels were separated via median split to explore the role of high and low levels and the same procedure was used for progesterone levels. The actual hormone levels were analyzed in separate models. Random effects regression models of cortisol levels collected prior to smoking were conducted for actual estradiol and progesterone levels. Additionally, separate ANOVAs were used to examine the effect of hormones by medication condition on number of cigarettes smoked during the adlib period. Age, depression (assessed via CES-D), marital status, education, race, and nicotine dependence (FTND) were evaluated as covariates across all outcomes and were retained if they were significantly associated with each outcome. Supplemental analysis was repeated to include the covariate of regular versus none/irregular menses, to account for various stages of reproductive status; however, results were not substantively altered. Thus, the full sample is reported.

3. Results

Participants did not differ on any baseline characteristics by medication condition (see Table 1). The average age was 37.35 years (SD = 11.44), and participants were primarily White (51.2%), not married (83.7%), and had at least a college degree (51.2%). Participants smoked on average 15.50 (SD = 8.31) cigarettes per day and averaged 4.86 (SD = 2.07) on the FTND, indicating low to moderate levels of dependence. They also scored a 9.00 (SD = 7.48) on the CES-D, thus were below the range of clinical depression measured by the scale. 9.4% (n = 3) of participants endorsed taking a hormonal form of birth control. Additionally, 48.8% of participants endorsed regular menses (defined as occurring about once per month), while the remaining participants endorsed either having irregular menses or not having menses. The estradiol median split was 24.67 pg/mL (SD = 51.21); the progesterone median split was 0.57 ng/mL (SD = 2.23). The high estradiol group included 21 women (n = 10, placebo; n = 11, guanfacine) and the high progesterone group also included 21 women (n = 11, placebo; n = 10, guanfacine).

Table 1.

Baseline demographics by medication (n = 43).

	Guanfacine (n = 21)	Placebo (n = 22)
Age, mean (SD)	34.00 (9.08)	40.55 (12.70)
Laboratory Session Assays	50.18 (59.78)	33.20 (40.63)
Estradiol pg/mL level, mean (SD)
Progesterone ng/mL level, mean (SD)	1.08 (1.44)	1.52 (2.83)
CPD, mean (SD)	14.21 (5.74)	16.73 (10.16)
FTND, mean (SD)	4.71 (1.93)	5.00 (2.23)
CESD, mean (SD)	8.38 (7.49)	9.59 (7.59)
Race, n (%)
White	12 (57.1%)	10 (45.5%)
Other	9 (42.9%)	12 (54.5%)
Marital Status n (%)
Married	3 (14.3%)	4 (18.2%)
Not Married	18 (85.7%)	18 (81.8%)
Education n (%)
High School or less	11 (52.4%)	10 (45.5%)
College or more	10 (47.6%)	12 (54.5%)

Note. All baseline comparison across medication groups were not significant (p > 0.05); CPD = cigarettes per day; FTND = Fagerström Test of Nicotine Dependence, range 1–10; CES-D = Center for Epidemiological Studies Depression Scale, range 0–60.

3.1. Tobacco craving

3.1.1. Estradiol & craving

When examining the intention and desire to smoke (QSU-Factor 1), there was a main effect of medication (F = 10.94, p = 0.02), in that participants receiving guanfacine reported lower rates of craving (M = 50.52, SE = 5.27) versus those receiving placebo medication (M = 74.85, SE = 5.11). Additionally, a three-way interaction between time, medication condition, and estradiol level (F = 4.00, p = 0.05), demonstrated that high levels of estradiol override the effect of guanfacine on desire to smoke based craving (See Fig. 1). Among participants receiving guanfacine, pre-stress induction levels of craving were lower among those with low levels of estradiol (M = 30.31, SE = 8.48), as compared to those with high levels of estradiol (M = 58.55, SE = 7.56). This was also true among post-stress induction levels of cravings, in that cravings among individuals with low levels of estradiol (M = 47.40, SE = 8.33) were lower than those with high levels of estradiol (M = 65.83, SE = 7.42). Levels of craving remained consistent between pre- and post-stress induction among the placebo condition in both the low estradiol group (pre-induction, M = 77.60, SE = 7.53; post-induction, M = 76.95, SE = 7.39) and the high estradiol group (pre-induction, M = 67.48, SE = 7.88; post-induction, M = 77.36, SE = 7.74).

Fig. 1. Time, medication, and estradiol levels on cravings related to desire to smoke.

Note. In the low estradiol group, there is a significant guanfacine versus placebo difference. Guanfacine reduced craving. In the high estradiol group, there is no guanfacine versus placebo difference on craving.

Regarding the anticipated relief from smoking (QSU-Factor 2), a main effect of time was observed (F = 5.07, p = 0.03); craving increased following the stress-induction procedure (pre-induction, M = 31.16, SE = 4.31; post-induction, M = 41.73, SE = 5.17).

3.1.2. Progesterone & craving

When examining the intention to smoke (QSU-Factor 1), among those receiving guanfacine, high levels of progesterone were further protective against craving related to anticipated relief from smoking; there was a three-way interaction between time, medication condition, and progesterone level (F = 5.42, p = 0.03; See Fig. 2a). Guanfacine protected against cravings related to anticipated relief of smoking among individuals with high levels of progesterone (pre-induction, M = 53.72, SE = 8.28; post-induction, M = 59.83, SE = 7.46) compared to placebo (pre-induction, M = 78.50, SE = 7.90; post-induction, M = 88.52, SE = 7.12). Similar differences in levels of cravings related to anticipated relief were observed between medication groups among individuals with low levels of progesterone (guanfacine: pre-induction, M = 33.40, SE = 8.29; post-induction, M = 55.63, SE = 7.47; placebo: pre-induction, M = 66.06, SE = 8.34; post-induction, M = 64.25, SE = 7.51).

Fig. 2. Progesterone effects on tobacco cravings.

Note. Guanfacine reduced craving both related to desire to smoke and anticipated relief and this effect was augmented among women in the high progesterone group.

Regarding QSU-Factor 2, anticipated relief from negative affect, there was a main effect of time (F = 5.01, p = 0.03), in that craving increased from pre-stress induction (M = 31.04, SE = 4.24) to post-stress induction (M = 41.33, SE = 4.81). Furthermore, there was a two-way interaction between time and medication (F = 4.75, p = 0.04); individuals receiving guanfacine had lower ratings of craving at both pre- and post-stress induction (pre-induction, M = 22.09, SE = 6.08; post-induction, M = 37.51, SE = 6.89) compared to individuals receiving the placebo medication (pre-induction, M = 39.91, SE = 5.94; post-induction, M = 45.14, SE = 6.73). Finally, there was also a significant three-way interaction (F = 8.95, p = 0.01; see Fig. 2b), in that among individuals with high progesterone, guanfacine attenuated cravings (pre-induction, M = 24.29, SE = 8.59; post-induction, M = 35.24, SE = 9.73) compared to placebo medication (pre-induction = 46.18, SE = 8.19; post-induction, M = 60.98, SE = 9.28). Among individuals with low progesterone, guanfacine attenuated cravings prior to the stress-induction (M = 19.90, SE = 8.60) compared to the placebo condition (M = 33.79, SE = 8.65); however, following the stress induction, individuals taking guanfacine with low progesterone levels had higher ratings (M = 39.78, SE = 9.75) than that of the placebo group (M = 29.31, SE = 9.80).

3.2. Cortisol

A two-way interaction between estradiol levels and medication (F = 14.23, p < 0.001; See Fig. 3) indicated that high levels of estradiol override the medication effect of guanfacine on cortisol. Among participants receiving guanfacine, those with lower levels of estradiol, guanfacine normalized cortisol levels (M = 10.33, SE = 0.72), while those with higher levels of estradiol demonstrated blunted cortisol levels (M = 8.13, SE = 0.92). The cortisol level among those with higher estradiol was similar to the levels observed in the placebo condition for both hormone groups (high estradiol: M = 7.30, SE = 0.72; low estradiol: M = 7.68, SE = 0.82). The main and interaction effects of progesterone were not significant.

Fig. 3. Estradiol effects on cortisol levels.

Note. *p < 0.05; high levels of estradiol override the medication effect of guanfacine on normalized cortisol response in comparison to the low estradiol group.

3.3. Cigarettes smoked

There was a main effect of estradiol levels on ad-libitum cigarettes smoking (F = 12.23, p = 0.001; See Fig. 4). Participants with high levels of estradiol smoked more cigarettes (M = 2.10, SE = 0.22) during the ad-libitum smoking period than those with low levels of estradiol (M = 1.03, SE = 0.22). No effects of progesterone were observed.

Fig. 4. Estradiol effects on cigarettes smoked during ad-libitum period.

Note. *p < 0.05; there was a difference between high and low estradiol in ad-libitum cigarettes smoked; participants with high levels of estradiol smoked more cigarettes during the ad-libitum period.

4. Discussion

The present study provides initial evidence that estradiol and progesterone levels impact guanfacine’s effect on stress-induced tobacco craving, stress-reactivity, and cigarettes smoked among women who smoke. Interestingly, high levels of estradiol overrode guanfacine’s effect on attenuating women’s desire to smoke. Both pre- and post-stress induction ratings of craving were lower in women receiving guanfacine with low levels of estradiol, while ratings of craving for women receiving guanfacine with high levels of estradiol were similar to those in the placebo medication group. Similarly, women with lower levels of estradiol who were taking guanfacine had higher ratings of cortisol than those with high levels of estradiol taking guanfacine and those in the placebo group.

Previous research has shown that individuals who smoke have diminished responses to stress. One study demonstrated that attenuated physiological responses, including cortisol reactivity in response to an acute laboratory-based stressor, predicted relapse among individuals who smoke [30]. Guanfacine has been associated with increased levels of cortisol following stress-induction [24], illustrating its potential to normalize the blunted stress-response typically observed in individuals who smoke. This is congruent with its effectiveness in women who smoke, who are more vulnerable to stress-induced relapse and would benefit from a normalization of the hypothalamic-pituitary-adrenal (HPA) axis activation in response to stress [13,24].

However, guanfacine’s function to normalize stress-reactivity and craving was attenuated by high levels of estradiol in the present study. In a previous study of stress-induction among individuals who smoke, periods of high estradiol (e.g., follicular phase) were associated with increased stress-reactivity and arousal compared to responses in men [31]. This provides further evidence that estradiol increases stress response among women and thus overrode the medication effects of guanfacine in women with high estradiol levels. Moreover, in the current study high levels of estradiol were associated with smoking more cigarettes during the ad-libitum smoking period regardless of medication group. Previous data has demonstrated that estradiol is associated with increased smoking [32] and past studies have demonstrated that guanfacine reduces the amount of cigarettes smoked during ad-libitum smoking periods [24]. However, high estradiol levels appear to have attenuated such medication effects.

This is suggestive that personalizing cessation attempts in women may be useful if there is consideration of actual hormone levels. For example, one potential strategy may be to time smoking cessation attempts in women during times when estradiol levels are lower; however, it is worth noting that such measures should comprise of actual hormone levels to account for between-cycle hormonal fluctuations and variations versus a dichotomization of menstrual cycle phase [33]. Alternatively, women with higher actual levels of estradiol may benefit from adjunctive treatments to specifically target stress-reactivity during a cessation attempt [17], such as mindfulness interventions.

Progesterone proved to be protective against increased craving from smoking and notably guanfacine enhanced this effect. This is consistent with previous literature suggesting progesterone, including exogenous progesterone, reduces tobacco cravings [32]; while little evidence to date suggests that progesterone impacts negative affect [32], increasing progesterone levels have been related to decreasing HPA-axis hormones [34] and thus may be protective against stress-related craving, which is particularly salient in women and would complement guanfacine’s effectiveness. This relationship should be explored in future studies.

Overall, it is important to consider the present study’s findings in the context of the larger sex-specific literature to date. Few clinical trials adequately power their samples to explore sex-specific medication effects and it is assumed that medication response is equitable across sexes [35]. The lack of exploration regarding pharmacological responses and actual hormone levels is especially problematic; there is a bidirectional relationship, in which sex hormones impact pharmacokinetics and pharmacodynamics, including drug distribution, binding, absorption, and elimination, and in turn medications impact hormone levels [35]. One example of this is aspirin’s antithrombotic effect, which is observed primarily in men and may be mediated by testosterone [36]. However, despite this information, very few studies explore the impact of actual sex hormones on medication effects and even fewer are adequately powered to do so [35], primarily using post-hoc analyses [37] and examining medication effects by menstrual cycle phase [38], which has historically been problematic [33].

The scant literature regarding medication effects and actual hormone levels remain true across smoking cessation medication trials, in which the majority of studies have not been designed to explore sex differences in the Federal Drug Administration (FDA)-approved cessation medications [39] and are thus not powered to explore sex-specific variables including how hormones impact medication effects. This is increasingly important as the present study illustrates that high actual levels of estradiol attenuated craving, stress-reactivity, and ad-libitum cigarettes smoked.

Despite the current study’s strengths, it is important to acknowledge that the present study did not explicitly control for hormone use or menopausal status. However, supplemental analysis controlled for reproductive status (which included regular menses, n = 21) to account for the various factors that may impact individual, actual hormone levels; this supplemental analysis did not substantively change the pattern of results. Also, we followed best practice procedures for the collection and analysis of hormone levels [20]. Nonetheless, we were able to detect important differences based upon actual hormone levels that warrant further exploration. Future studies should be powered to explore a priori hypotheses related to medication effects by actual hormone levels and account for hormonal factors, including reproductive status.

5. Conclusion

The present findings are consistent with previous research that guanfacine reduced stress-related craving and stress-reactivity among women who smoke; however high actual levels of estradiol attenuated guanfacine’s function to normalize stress-reactivity and reduce craving among women who smoke. Additionally, findings provide further support that progesterone plays a role in decreasing craving in women smoking and enhances guanfacine’s effects. Overall, these findings underscore the importance of examining the role of sex hormones in medication trials.

Data Availability

Data will be made available on request.