Allopregnanolone During Short-Term Smoking Abstinence: Associations with Depressive Symptoms, Smoking-Related Symptomatology and Nicotine Response

Home , Cotinine

Allopregnanolone during Short-Term Smoking Abstinence: Associations with depressive symptoms, smoking-related symptomatology and nicotine response

A Dissertation SUBMITTED TO THE FACULTY OF UNIVERSITY OF MINNESOTA BY

Alicia Marie Allen

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

Adviser: Harry Lando, PhD

November, 2012

Acknowledgements

This project was funded by National Institute on Drug Abuse (NIDA)

Grants R01-DA08075 and R36-DA032539, and the J.B. Hawley Award (School of Public Health, University of Minnesota). Additional support comes from Grant

Number 1UL1RR033183 from the National Center for Research Resources

(NCRR) and by Grant Number 8UL1TR000114-02 from the National Center for

Advancing Translational Sciences (NCATS) of the National Institutes of Health

(NIH) to the University of Minnesota Clinical and Translational Science Institute

(CTSI). The University of Minnesota CTSI is part of a national Clinical and

Translational Science Award (CTSA) consortium created to accelerate laboratory discoveries into treatments for patients. Its contents are solely the responsibility of the authors, and do not necessarily represent the official views of CTSI, NIDA,

NCRR, NCATS or NIH. Additional thanks go to the research staff – Lindsay

Farnsworth, Kathryn Resner, Sara Paradise, Nicole Tosun, Jennifer Widenmier, and Danielle Young – for their dedication to participant recruitment and follow-up, as well as data collection, entry and management. Finally, I would also like to acknowledge Dr. Richard Hauger and Alan Turken at the University of California,

San Diego for analyzing the allopregnanolone samples.

Abstract

Background: Allopregnanolone (ALLO) is a neuroactive steroid metabolized from progesterone and, therefore, varies by menstrual phase in premenopausal women. Previously published literature has shown that risk for relapse to smoking varies by menstrual phase. Further, recent preclinical research indicates that ALLO may protect against drug abuse behaviors. Therefore, this dissertation project aims to characterize ALLO by menstrual phase in women with and without depressive symptoms (Paper #1) and explore the effect of ALLO on smoking-related symptomatology (SRS; Paper #2) and nicotine response (NR; Paper #3) during short-term smoking abstinence.

Methods: At screening, participants (n=87) were stratified by depressive symptoms status and, using a controlled cross-over study design, were randomized to testing order (i.e., follicular (F) menstrual phase followed by the luteal (L) phase or vice versa (L-F)). The six-day testing week consisted of two days of ad libitum smoking followed by four days of biochemically verified smoking abstinence. ALLO was measured twice during each testing week: during ad libitum smoking and on the fourth day of smoking abstinence. Participants completed daily forms to assess SRS during the testing week. On the fourth day of smoking abstinence, participants participated in a NR lab session. Growth curve and covariance pattern models, adjusted for menstrual phase and testing order, were used to assess the effect of ALLO on SRS and NR, respectively.

Results: In the first paper (n=84), a significant menstrual phase difference was observed in the change in ALLO level during smoking cessation. Specifically, ALLO decreased by 10% in the F phase and increased by 31% in the L phase (p<0.01). There were no significant differences in ALLO levels between the depressive symptoms groups. In the second paper (n=64), the absolute level of

ALLO on the day before quit was significantly associated with the following: (1) perceived stress on the day before quit ( β=-2.25, p<0.01), (2) the change in perceived stress during smoking abstinence (β=0.79, p<0.01), and (3) premenstrual symptoms of pain and water retention on the day before quit (β=1.09, p<0.01; β=1.08, p<0.01; respectively). The change in ALLO during smoking abstinence was significantly associated with the following: (1) positive and negative affect on the day before quit ( β=1.15, p<0.01; β=1.04, p=0.04; respectively), (2) perceived stress on the day before quit ( β=-1.77, p=0.01), (3) the change in perceived stress during smoking cessation (β=0.17, p<0.01), and (4) the change in depressive symptoms on the day of quit ( β=-1.52, p=0.02). Finally, in the third paper (n=77), ALLO had a significant, positive association with the following variables prior to initiation of nicotine nasal spray: systolic blood pressure ( β=0.85, p=0.04), diastolic blood pressure ( β=1.19, p<0.01), and subjective levels of physical symptoms ( β=0.58, p<0.01), dizziness ( β =0.88, p<0.01), jitteriness ( β=0.90, p=0.04) and pleasantness ( β=2.05, p=0.041). ALLO also had significant associations with changes in cognition from baseline to post nicotine nasal spray use: specifically, discriminability (a measure of attention; β=1.15, p=0.05), and bias (a measure of impulsivity; β=0.12, p=0.02).

Conclusion: ALLO varied significantly by menstrual phase and smoking status, and had a significant effect on several measures of SRS and NR. While several of these associations were favorable (i.e., perceived stress on the day before quit and pleasantness on the fourth day of smoking abstinence), some were not (i.e., premenstrual symptoms on the day before quit and increased subjective report of physical symptoms on the fourth day of smoking abstinence). Therefore, it remains unknown whether or not ALLO is a protective factor against drug abuse behaviors. Additional research is needed to explore the role of ALLO directly on smoking cessation outcomes.

iii

Table of Contents

Page Number List of Tables vi List of Figures vii A. Introduction 1 B. Significance, Background & Innovation 4 C. Parent Study Overview 12 D. Pape r #1: The Role of Menstrual Phase and Depressive Symptoms Status on 17 Allopregnanolone Levels during Short-Term Smoking Abstinence D.1. Overview 17 D.2. Background & Significance 18 D.3. Methods 20 D.4. Results 23 D.5. Summary of Results 24 E. Paper #2: Allopregnanolone and Smoking - Related Symptomatology during Short-Term 25 Smoking Abstinence E.1. Overview 25 E.2. Background & Significance 27 E.3. Methods 29 E.4. Results 31 E.3. Summary of Results 33 F. Paper #3: Allopregnanolon e and Nicotine Response during Short-Term Smoking 37 Abstinence F.1. Overview 37 F.2. Background & Significance 39 F.1. Methods 41 F.2. Results 44 F.3. Summary of Results 46

G. Discussion 47 G.1. Overview 47 G.2. Menstrual Phase 47 G.3. Depressive Symptoms 49 G.4. Perceived Stress 52 G.5. Premenstrual Symptoms 53 G.6. Smoking-Specific Symptomatology 54 G.7. Other Nicotine Response Measures 55 G.8. Strengths & Limitations 57 G.9. Future Directions 58 G.10. Conclusions 59 H. References 60

List of Tables

Page Number Table 1. Study Measures 13

Table 2. Frequency of Depressive Symptoms Status by Classification Tool 21

Table 3. Demographics & Smoking Behavior by 23 Depressive Symptoms Status (n=84)

Table 4. Allopregnanolone (ng/mL) by Menstrual Phase, Smoking Status and Depressive Symptoms 24 Status (n=84)

Table 5. Demographics, Smoking Behavior & Allopregnanolone Levels (n=64) 31

Table 6. Associations between Smoking-Related Symptomatology and Allopregnanolone (n=64) 34

Table 7. Demographics, Smoking Behavior & Allopregnanolone Levels (n=77) 44

Table 8. Associations between Nicotine Response 45 Variables and Allopregnanolone (n=77)

List of Figures

Page Number Figure 1. Estimated Estrogen and Progesterone Values during the Human Menstrual Cycle 5

Figure 2. Estimated Allopregnanolone Levels during the Human Menstrual Cycle 7

Figure 3. Central Hypothesis 11

Figure 4. Parent Study Flow Diagram 12

Figure 5. Nicotine Exposure Session Protocol 14

Figure 6. Smoking-Related Symptomatology during Short-Term Smoking Cessation by Allopregnanolone 35 Level on Day 2 (n=64)

Figure 7. Smoking-Related Symptomatology during Short-Term Smoking Cessation by Change in 36 Allopregnanolone Level from Day 2 to Day 6 (n=64)

vii

A. INTRODUCTION

Cigarette smoking persists as the leading cause of preventable morbidity and mortality in the United States (CDC, 2010). Despite the many advancements in smoking cessation interventions, the majority of smokers, especially women, relapse soon after a quit attempt (Benowitz, 2009; CDC, 2010; Perkins et al

2001; Perkins & Scott, 2008; USDHHS, 2001). Evidence continues to accumulate suggesting that sex hormones may influence risk for smoking relapse (e.g. Lynch & Sofuoglu, 2011). Recently, new efforts have begun to focus on the role of allopregnanolone (ALLO), which is a stress-reducing neuroactive steroid. ALLO is primarily, though not exclusively, metabolized from the sex hormone progesterone (Backstrom et al 1986; Ottander et al, 2005). Therefore, the level in ALLO in women varies by menstrual phase such that it is lower in the follicular phase (i.e., low progesterone) and higher in the luteal phase (i.e., high progesterone; Genazzani et al 1998; Nyberg et al 2007).

Within the animal literature, ALLO has been found to have protective effects against the drug abuse behaviors of escalation, reinstatement, and withdrawal; some of these associations varied by sex (Anker et al, 2008; Carroll

& Anker, 2010). Although ALLO has been shown to be associated with a variety of smoking relapse risk factors (including depressive symptoms, smoking-related symptomatology and variables associated with nicotine response), there is currently a lack of clinical research on the role of ALLO in addictive behaviors.

Therefore, the goal of this doctoral dissertation is to conduct a series of

1 secondary-data analyses to investigate the role of menstrual-timed endogenous levels of ALLO on these smoking relapse risk factors.

The “Menstrual Phase and Depressive Symptoms in Acute Smoking

Abstinence” (MPADS; 2006-2012) incorporated a controlled cross-over study design to collect data from women smokers with and without depressive symptoms during both the follicular (F; low ALLO) and luteal (L; high ALLO) menstrual phases. The two menstrual-timed six-day testing weeks included two days of ad libitum smoking followed by four days of biochemically-verified smoking abstinence during which participants completed daily assessment of smoking-related symptomatology and provided daily blood samples. On the fourth day of smoking abstinence, a nicotine response lab session was conducted. The following three papers, and related aims, are included in this dissertation project:

1. The Role of Menstrual Phase and Depressive Symptoms Status on

Allopregnanolone Levels during Short-Term Smoking Abstinence

a. Aim: Determine the ALLO levels by menstrual phase during ad

libitum smoking and smoking abstinence; Hypothesis: ALLO levels

will be higher in the L phase than the F phase during both ad libitum

smoking and smoking abstinence

b. Aim: Investigate the role of depressive symptoms on ALLO levels by

menstrual phase during short-term smoking cessation; Hypothesis:

Regardless of menstrual phase, women who have depressive

symptoms will have both lower absolute levels of ALLO, as well as a

blunted menstrual phase variation in ALLO during short-term

smoking cessation, versus women with no depressive symptoms.

2. Allopregnanolone and Smoking-Related Symptomatology during Short-

Term Smoking Abstinence

a. Aim: Characterize the association between ALLO levels with

smoking-related symptomatology during short-term smoking

abstinence; Hypothesis: There will be associations between the

absolute level of ALLO on the day before quit day with both the

absolute level of smoking-related symptomatology during ad libitum

smoking and with the change in smoking-related symptomatology

during short-term smoking abstinence. Further, the change of ALLO

from the day before quit date to the fourth day of smoking abstinence

will be significantly associated with both the absolutely level and

change in smoking-related symptomatology.

3. Allopregnanolone and Nicotine Response during Short-Term Smoking

Abstinence

a. Aim: Assess the association between ALLO and nicotine response

during short-term smoking abstinence; Hypothesis: We will observe

an inverse association with ALLO and cognitive measures, and a

positive association between ALLO and cardiovascular and

subjective measures.

Given that less than three percent of those who attempt to quit smoking remain abstinent six months later (Benowitz, 2009; CDC, 2010), there is an urgent need to gain a better understanding of the dynamic process involved in smoking relapse. This project will provide new knowledge regarding of the possible protective role of ALLO against smoking relapse. This paper will begin with an overall significance, background and invocation followed by an overall description of the parent study. Then, within each paper, evidence and methodology specific to each paper will be presented, along with the detailed results and summary of findings. Finally, a discussion of the application of these results to the field and future directions will conclude this paper.

B. SIGNIFICANCE, BACKGROUND & INNOVATION

B.1. Epidemiology of Smoking.

Currently an estimated 46 million Americans smoke, resulting in more than

$190 billion annually for direct medical expenses and losses of productivity

(CDC, 2010). Among smokers, women are at increased risk of smoking-related morbidity and mortality including, but not limited to, myocardial infarction, cardiovascular disease, and lung cancer compared to their male counterparts

(Perkins et al 2001; Prescott et al, 1998; Baron et al, 1990; Zang & Wynder,

1996). While three out of four current smokers report wanting to quit smoking, a staggering 97 percent relapse within six months of a quit attempt indicating a lack of effective smoking cessation interventions (CDC, 2010; Benowitz, 2009).

B.2.Sex Differences & Smoking Behavior.

The animal literature clearly indicates that females are more susceptible to drug abuse behaviors on almost all stages of dependence (Carroll & Anker,

2010). For instance, females increase consumption at a faster rate than males during the acquisition stage (Carroll & Anker, 2010) and females are more sensitive to the rewarding effects of drugs compared to males (Finn et al, 2010).

Similar sex differences have been observed in clinical populations within the context of tobacco addiction. Compared to men, women have lower rates of smoking cessation (USDHHS, 2001; Perkins, 2001), and may receive less benefit from smoking cessation medications (e.g. Perkins & Scott, 2008). While substantial research has focused on sex differences in smoking relapse, the specific causal mechanisms remain unknown. One hypothesis proposed to explain these observations involves the role of sex hormones.

B.3. Sex Hormones & Smoking.

The relationship between sex hormones (often measured by the proxy menstrual phase as shown in Figure 1 from Carpenter et al, 2006) and smoking cessation appears to be quite complex. In general, the animal literature indicates estrogen is associated with an increase in self-administration of addictive drugs whereas progesterone is associated with a decrease (Carroll & Anker, 2010).

However, this relationship is less clear in clinical research. For instance, some

(but not all) of the literature indicates that menstrual phase has been shown to be associated with self-reported levels of craving, withdrawal symptoms, perceived stress and response to smoking cues (Allen et al, 2009a; Allen et al 2009b; Allen et al 2010; Carpenter et al, 2006; Franklin et al 2004; Gray et al 2010; Schiller et al, 2012). Additionally, while the circadian rate of smoking has been shown to vary by menstrual phase, other data indicate that total number of cigarettes smoked per day does not vary by phase (Allen et al, 2009c; Carpenter et al,

2006). This may be related to the possible association between sex hormones and nicotine absorption and metabolism (Allen et al, In Press; Benowitz et al,

2006; Hukkanen et al, 2005). These observations, although mixed, suggest sex hormones influences smoking behavior. Several studies have attempted to assess the impact of menstrual phase on smoking cessation outcomes, producing results that seem to conflict. Two of the four studies conducted used nicotine replacement therapy and observed better cessation rates among women who quit smoking in the follicular phase (Carpenter et al 2008; Franklin et al

2008).

However, the other two studies, one using behavioral counseling alone and one using bupropion (a non-nicotine smoking cessation aid), observed better cessation rates among women who quit in the luteal phase (Allen et al, 2008;

Allen et al, 2009d; Mazure et al, 2011). Based on these observations, a recent publication hypothesized that the timing of a quit attempt during the menstrual cycle may be as important as the cessation treatment provided (Franklin & Allen,

2009). Therefore, one possibility may be that successful smoking cessation efforts in women may be dependent on progesterone, and its metabolites, mediating the neurobiological response to nicotine .

B.4. Allopregnanolone.

The impact of progesterone on neurotransmission appears to have its most prominent effect via its A-ring metabolite allopregnanolone

(ALLO; Paul & Purdy, 1992).

ALLO is a neuroactive steroid, which means it has the capability of regulating neuronal activities, such as premenstrual syndrome, depression and anxiety (Zheng, 2009). As shown in

Figure 2 , in premenopausal women ALLO levels change daily and vary by

7 menstrual phase such that women in the luteal phase of the menstrual cycle have significantly higher ALLO levels compared to both women in the follicular phase and men (Genazzani et al 1998; Nyberg et al 2007). ALLO has been shown to be highly correlated to the levels of progesterone during the luteal phase, but not during the follicular phase (Genazzani et al, 1998). This is likely due to the predominate production of ALLO by the corpus luteum, with minimal production de novo in the brain in response to stress (Backstrom et al 1986;

Ottander et al, 2005).

ALLO readily crosses the blood-brain barrier, where it rapidly alters the central nervous system (Paul & Purdy, 1992; Purdy et al, 1991). It then produces the following neurocognitive effects within minutes - changes in physiological stress response (Girdler et al 2001; Girdler et al, 2006), variation of negative mood symptoms (Andreen et al 2005), confusion (Freeman et al, 1997), level sedation (Girdler et al, 2001; Girdler et al, 2006; Timby et al, 2006), and poor concentration (Freeman et al, 1997). ALLO also appears to have an inverse relationship with the severity of premenstrual symptoms (Wang et al, 1996), these symptoms may be the result of ALLO withdrawal (Sundstrom-Poromaa,

2008). Research also indicates that ALLO serves as an endogenous homeostatic mechanism restoring both GABA and HPA function to normal after acute stress (Girdler et al 2007; Girdler et al 2006; Concas et al 2006). Changes in ALLO are also associated with changes in dopamine release, which may be associated with the mood disturbances observed in premenstrual symptoms as

8 well as the development and maintenance in symptoms of depression and anxiety (Zheng, 2009). Overall, a recent review article noted “ALLO is one of the most important neurosteroids in the brain ” impacting neuroendocrine axes, as well as stress response and cognitive functioning (Zheng, 2009). In terms of drug abuse behavior, it is hypothesized that ALLO levels may mediate drug response via changes in GABAa and dopamine release (Zheng, 2009).

B.5. ALLO & Drug Abuse Behavior.

ALLO has been shown to be associated with the reduction of negative drug abuse behaviors in a number of animal studies (Carroll & Anker, 2010).

Specifically, ALLO produces an inhibitory influence on escalation of cocaine self- administration in rats (Anker et al, 2008) and a reduction of reinstatement behavior in cocaine-seeking female rats, but not in their male counterparts

(Anker & Carroll, 2010). Further, ALLO has been shown to blunt withdrawal effects in benzodiazepine- and morphine-addicted mice (Reddy & Kulkarni,

1997a; Reddy & Kulkarni, 1997b). Therefore, animal literature has concluded that

ALLO has reinforcing properties (Sinnott et al, 2002) and may be an effective substitute for ethanol abuse (Grant et al, 1997; Bowen et al, 1999).

The information on ALLO in drug abuse clinical research is scarce, especially within samples of women. To our knowledge only two studies have been conducted to investigate the association between ALLO and tobacco use. While neither of these studies observed a significant association between subjective effects, such as mood and level of craving, and ALLO, they both failed to include

9 a sample of women (Childs & de Wit, 2009; Marx et al, 2006). However, two studies that did include women, observed a significant protective association between ALLO and the behavioral effects of alcohol in a non-dependent sample of adolescent females (Torres & Ortega, 2003) and the severity of withdrawal in adult women addicted to alcohol (Hill et al, 2005). Overall, these studies indicate that the association between ALLO and drug abuse behaviors may have a sex- specific effect perhaps due, at least in part, to the menstrual cycle.

B.6. Formulation of Central Hypothesis.

A recent theory posits that nicotine withdrawal symptoms may be the result of a deficiency in dopamine release during attempted smoking cessation (Benowitz,

2009). We propose an expansion of this hypothesis given that the literature indicates: (1) luteal phase is associated with improved smoking cessation outcomes in the absence of nicotine replacement therapy (Allen et al 2008; Allen et al 2009d; Mazure et al, 2011), (2) ALLO levels peak in the luteal phase

(Genazzan et al, 1998), (3) increases in ALLO are associated with changes dopamine release (Zheng, 2009), and (4) ALLO has been shown to have protective effects against drugs with abuse potential (Anker et al 2008; Anker &

Carroll, 2010; Reddy & Kulkarni, 1997a; Reddy & Kulkarni, 1997b; Sinnott et al

2002; Grant et al 1997; Bowen et al 1999; Torres & Ortega, 2003; Hill et al

2005). Therefore, as displayed in Figure 3 , the central hypothesis for this project is that ALLO will have a protective effect against the stressful experience of smoking cessation by decreasing withdrawal symptoms and blunting response to

10 nicotine.

B.7. Innovation.

In light of the observed associations between menstrual phase and smoking cessation rates, recommendations for more attention to the dynamic processes involved in smoking cessation and relapse prevention have been made

(Shiffman, 2008). This project directly responds to that request by exploring the role of ALLO in smoking relapse risk factors among premenopausal women. The project is innovative as it will be among the first to investigate the association between ALLO and smoking behavior in a sample of females. Further, we will systematically investigate the role of ALLO levels by menstrual phase in short- term smoking abstinence. If our expectations are confirmed, the results will be of considerable interest given that the exogenous treatment of ALLO in humans has recently been proven to be safe with minimal side effects (Timby et al 2006).

C. PARENT STUDY

C.1. Overview.

The primary objective of the “Menstrual Phase and Depressive Symptoms in

Acute Smoking Abstinence” (MPADS) Study (funded by NIDA (DA08075-15)) is to investigate two factors that may affect smoking in women: (1) the presence or absence of depressive symptoms and, (2) the interaction of depressive symptoms with menstrual phase. This study utilized a controlled cross-over study design and completed data collection in April, 2012.

C.2. Study Procedures.

In brief, participants were recruited through mass media and internet advertising. Eligible participants were women between the ages of 18 and 40 with regular menstrual cycles who reported smoking at least five cigarettes per day for the past year. Exclusionary items included the use of exogenous hormones or psychotropic medications, and recent (< 3 months) breastfeeding or pregnancy. They also had to be, in the opinion of the study physician, in stable physical and mental health. Participants completed a phone interview and in- 12 person clinic visit to determine eligibility. After eligibility was confirmed, participants were stratified by depressive symptoms status (Figure 4 ; additional details on assessment of depressive symptoms can be found in Section D.2.2.) and randomized to testing order (i.e., F phase first then L phase, or vice versa).

All procedures were approved by the human subjects committee at the University of Minnesota.

After study enrollment, participants completed two six-day testing weeks.

During the testing weeks, participants attended visits daily to complete questionnaires, provide biological samples and confirm smoking status ( Table 1 ).

The six-day testing week started on the day after the onset of menses for F phase and two days after luteinizing hormone surge (determined by urine luteinizing hormone described in Allen et al, 2008) for the L phase. If schedule

13 conflicts occurred, the entire testing week was shifted by one day in either direction. Each testing week included daily clinic visits for six consecutive days.

On testing Days 1 and 2, participants were smoking ad libitum. At midnight on testing Day 2, participants quit smoking and remained abstinent for the rest of the testing week. Smoking abstinence was confirmed via carbon monoxide breathalyzer (<5 ppm), salivary cotinine (<15 ng/mL), and serum nicotine (< 2 ng/ml; Jarvis et al,

1987; Benowitz et al, 2008).

On Day 6, participants completed a nicotine exposure session which included a nicotine challenge via nicotine nasal spray and a timed series of cardiovascular, cognitive and subjective responses ( Figure 5 ). Upon completion of the first testing week, participants resumed ad libitum smoking for an approximate six weeks wash-out period (depending on menstrual cycle length), then repeated the identical testing procedures in the alternate menstrual phase.

C.3. ALLO Assays .

Up to a total of four serum samples per participant were analyzed for ALLO.

Specifically, samples from Day 2 (ad libitum smoking) and Day 6 (fourth day of

14 smoking abstinence) were included. A total of 296 serum samples were sent for

ALLO analysis. One of the samples was not analyzed due to an insufficient volume. Serum samples analysis was completed by Dr. Richard Hauger’s Brain

Receptor Signaling Lab at The University of California, San Diego. Samples were shipped overnight on dry ice via FedEx. Analyses were completed using radioimmunoassays methods in batches as previously described (Purdy et al,

1990) and as was recently completed in a similar project (Childs & de Wit, 2009).

Using these methods, intra- and inter-assay coefficient of variations were approximately 5% and 8%, respectively. Sensitivity is about 200 pg/ml, with a standard range of 0.2 to 50.0 ng/ml (Alan Turken, personal communication,

5/31/12).

C.4. Subsample Selection .

To fully capitalize on the cross-over study design, participants had to have at least two corresponding blood samples available to be analyzed for ALLO (i.e., two Day 2 samples and/or two Day 6 samples, or a Day 2 and Day 6 sample from the same testing week) to be included in this dissertation. Therefore, of the total 121 study participants completed the parent study, 27 (79%) were excluded from the present study due to a lack of blood samples available. An additional seven participants were excluded due to an inability to remain abstinent during the testing week (n=4; 12%), and having progesterone and estradiol hormone levels inconsistent with menstrual phase of testing (n=3; 9%). As shown in

Figure 4, this resulted in a total sample size of 87, including 36 (41%) in the no

15 depressive symptoms group and 51 (59%) in the depressive symptoms group.

Overall, 39 (45%) were randomized to complete the testing weeks in order of F then L and 48 (55%) were randomized to complete testing in the L then F. Of the 87 participants, 70 (80%) had ALLO values for Day 2 (ad libitum smoking) for both testing weeks and 77 (89%) had ALLO values for Day 6 (fourth day of smoking abstinence) for both testing weeks.

Power was calculated based on Paper #2, which was the limiting paper.

Effect sizes were based on data from a previous research study with a similar study sample. Specifically, effect sizes of self-reported withdrawal of 10.35 (S.E.

0.53) and 11.61 (S.E. 0.54) during the F and L menstrual phases, respectively

(Allen et al, 2010) were used. Given the cross-over study design, a conservative estimate of a within subject correlation of 0.40 for outcome variables was used.

With a fairly modest sample size (n=62), the cross-over study design is very powerful as each participant is compared to herself. Therefore, using a two-sided test with an alpha level of 0.05, a sample size of 62 participants would result in a power of greater than 99% to detect differences in symptoms (i.e., withdrawal) by menstrual phase (SAS Version 9.2). Given this high level of power, interpretation of study results should be done with caution as the public health and/or clinical significance is not known (Oakes, 2012).

D. PAPER #1: THE ROLE OF MENSTRUAL PHASE AND DEPRESSIVE

SYMPTOMS STATUS ON ALLOPREGNANOLONE LEVELS DURING SHORT-

TERM SMOKING CESSATION

D.1. Overview

Introduction: Women with depressive symptoms (DS) are at high risk for smoking relapse. Allopregnanolone (ALLO), a neuroactive steroid that varies by menstrual phase and DS, can be protective against stress. Therefore, the purpose of this paper is to explore changes in ALLO by menstrual phase during short-term smoking cessation among women with and without DS.

Methods: At screening, study participants (n=84) were classified as either having DS (n=48) or not (NDS; n=36) via the Composite International Diagnostic

Interview and Patient Health Questionnaire-9. Then, participants were randomly assigned an order to complete two testing weeks in the follicular (F, i.e., low

ALLO) and luteal (L, i.e., high ALLO) menstrual phases. In each menstrual phase, blood samples were collected during ad libitum smoking and on the fourth day of biochemically verified smoking abstinence.

Results: Participants were, on average, 30.1 ± 6.7 years old. Most (57%) were White and had achieved more than a high school education (68%). ALLO levels were significantly higher in the L phase than the F phase, regardless of smoking status (p<0.001). ALLO significantly decreased by 10% in the F phase

(p=0.0047) and significantly increased by 31% in the L phase (p<0.0001). There were no significant differences in ALLO levels by DS group.

Conclusion: ALLO levels varied by menstrual phase and smoking status in premenopausal women, but not by DS. It may be possible that the increase in

ALLO during short-term cessation in the L phase may be one of the factors protecting against relapse whereas a decrease in ALLO, as observed in the F phase, may increase risk for relapse. Additional research is needed explore this theory.

D.2. Background & Significance.

The relationship between depression and smoking is complex and bidirectional. Those with depressive symptoms are predisposed to smoking initiation (Glassman et al, 2001). Once they become smokers, they are more likely to experience depressive symptoms during attempted cessation and also more likely to relapse to smoking after a quit attempt (Glassman et al, 2001).

Unfortunately, women are at increased risk to experience both depressive symptoms and smoking relapse (Perkins, 2001). Further, the effect of depressive symptoms on smoking behavior is stronger in women than in men

(Borrelli et al, 1996; Husky et al, 2008; Pratt & Brody, 2010).

ALLO is a neurosteroid (steroid that is synthesized in the central and peripheral nervous system) that has profound effects on mood (Birzniece et al,

2006). The dysregulation of ALLO has been implicated in a number of psychiatric conditions, particularly mood disorders (Girdler & Klatzkin, 2007). For example, compared to normal controls, women with either a history of or current major depressive disorder (MDD) had lower levels of both naturally occurring

18 endogenous ALLO, as well as lower ALLO produced following administration of exogenous progesterone (Girdler & Klatskin, 2007; Strohle et al 2000; Klatskin et al, 2006). Increases in ALLO levels after successful treatment of MDD via SSRIs have been observed (Birzniece et al, 2006). The same positive relationship may exist between ALLO and depressive symptoms as higher ALLO levels were associated with an improved sense of well-being in premenopausal women and less depressive illness during the postpartum period (Wang et al, 1996; Nappi et al, 2001). Additional evidence comes from the preclinical literature, where models of depression (such as social isolation) have been shown to reduce

ALLO levels (Porcu, 2003; Roselli et al 2011). However, Andreen and colleagues (2006) concluded that the association between ALLO and negative mood symptoms is not a linear relationship; rather, they suggest it is a U-shape curve between ALLO and negative mood. Specifically, in postmenopausal women treated with exogenous progesterone, ALLO levels in the ‘mid-range’

(0.57-0.76 ng/mL or 1.5-2.0 nmol/L) had the highest negative mood scores, whereas women with lower (<0.57 ng/mL or <1.5 nmol/L) and higher (>0.76 ng/mL or >2.0 nmol/L) ALLO levels had lower levels of negative mood. Overall, these data suggest depressive symptoms and ALLO may have either a linear relationship (i.e., higher ALLO, lower depressive symptoms) or there may be an ideal therapeutic range - either way, it is clear that ALLO and depressive symptoms are related.

ALLO varies dramatically by menstrual phase (Genazzani et al, 1998) and may be protective against drug abuse behaviors (Carroll & Anker, 2010).

Further, women who have depressive symptoms may have an inability to recover from the stressful experience of smoking cessation due to an inappropriate ALLO response (Klatskin et al, 2006). Therefore, there is a need to better understand the fluctuation of ALLO by menstrual phase during smoking cessation in premenopausal women within the context of depressive symptoms.

Consequently, utilizing a controlled cross-over study design with a sample of women with and without depressive symptoms, this paper aims to: (1) determine the ALLO levels by menstrual phase during ad libitum smoking and smoking abstinence, and (2) investigate the role of depressive symptoms on ALLO levels by menstrual phase during short-term smoking abstinence. We hypothesized that ALLO levels will be higher in the luteal phase during both ad libitum smoking and smoking abstinence. Further, regardless of menstrual phase, women who have depressive symptoms will have both lower absolute levels of ALLO, as well as a blunted menstrual phase variation in ALLO during short-term smoking abstinence, versus women with no depressive symptoms.

D.3. Methods

D.3.1. Study Participants. Of the 87 participants selected for inclusion in this project, 84 were included in this paper. Three were excluded from the analyses given they did not complete the full depressive symptoms assessment

(specifically the Patient Health Questionnaire - 9 was not completed by three participants as it was included in the study after the study was initiated).

D.2.2. Measures – Depressive Symptoms Status. As noted above in

Section C.2., upon confirmation of study eligibility, participants were stratified into one of two groups – depressive symptoms (DS) and no depressive symptoms

(NDS). Using the Composite International Diagnostic Interview (CIDI) Computer

Assisted Interview (Wittchen et al 1991) to assess the DSM-IV criteria (APA,

1994) and the Patient Health Questionnaire-9 (PHQ9; Kroenke et al, 2001), the

DS group was defined as meeting at least one of the following criteria ( Table 2 ):

(1) lifetime presence of depressed mood or loss of interest/pleasure for at least

14 consecutive days, (2) lifetime presence of four or more DSM-IV behavioral symptoms, and/or (3) a score of five or greater on the

PHQ9. Those who met criteria for

Premenstrual

Dysphoric

Disorder

(PMDD) or MDD (i.e., criteria one and two above) within the last six months were

21 excluded and referred for treatment. The NDS group was defined as not meeting any of the criteria for the DS group. After stratification, participants were randomized to begin testing in the follicular phase followed by testing in the luteal

(F-L) phase or vice versa (L-F).

D.3.3. Measures – ALLO. All four ALLO blood samples were used in this paper. These samples were collected on Day 2 (ad libitum smoking) and Day 6

(fourth day of smoking abstinence) during each testing week. The analysis of these samples is described above in Section C.3.

D.2.4. Statistical Analysis. Descriptive statistics were calculated for demographics and smoking behavior variables including mean and standard deviation for continuous variables, and percent for categorical variables. Log transformation was considered for some variables when distribution was non- normal. Differences by depressive symptoms groups in demographic and smoking behavior variables were assessed by t-tests and Chi-square tests.

Paired t-tests were used to compare differences in ALLO levels by phase and smoking status. Multiple regression models were used to assess differences in absolute levels of ALLO during ad libitum smoking and smoking abstinence by menstrual phase. A random coefficient model with fixed linear time (i.e., menstrual cycle day) and an order effect was used to investigate the difference in

ALLO levels by depressive symptoms status during the different menstrual phase and smoking status conditions. In subsequent analyses, a depressive symptoms group and time (i.e., menstrual cycle day) interaction variable was

22 included in the model, and cigarettes smoked per day was assessed as a possible confounder using hierarchical regression. P-values less than 0.05 were deemed statistically significant. No adjustments for multiple comparisons were made. SAS V9.1.3 (SAS Institute, Cary, NC) was used for the analyses.

D.4. Results

D.4.1. Study Sample. Of the 84 participants included in this analysis, 36

(43%) were stratified into the NDS group and 48 (57%) into the DS group ( Table

2). Participants were, on average, 30.1 (± 6.7) years old, 57% were White and

68% had more than a high school education. More participants in the DS group had more than a high school education (77%) than those in the NDS group

(56%). However, this difference did not reach statistical significance ( Χ2=8.22, p- value=0.084). No other significant differences in demographics or smoking behavior by depressive symptoms group were noted ( Table 3 ).

D.4.2. ALLO Levels by Menstrual Phase & Smoking Status. During the F phase, the average ALLO level decreased by 10% from ad libitum smoking to smoking abstinence ( Table 4 ; t-value=2.03, p-value=0.047). Conversely, during

23 the L phase, ALLO levels increased by 31% from ad libitum smoking to smoking abstinence (t-value=4.71, p-value<0.001). During both ad libitum smoking and smoking abstinence, L phase ALLO levels were significantly higher than their corresponding F phase levels (t-value=10.67, p-value<0.001; t-value=16.00, p- value<0.0001; respectively).

D.4.3. ALLO by Depressive Symptoms Status. After adjusting for cigarettes/day, there were no significant differences in absolute ALLO by depressive symptoms status in either menstrual phase during smoking ad libitum or smoking abstinence conditions (Table 3). Although time (i.e., menstrual cycle day) was a significant independent predictor (f-value=192.37, p-value<0.001) of ALLO level, depressive symptoms status (f-value=0.05, p-value=0.831) did not predict ALLO level or the change in ALLO over time (i.e., menstrual cycle day; f-value=0.64, p- value=0.424). These results remained unchanged when controlling for cigarettes/day.

D.5. Summary of Results

In this controlled cross-over study in which women with (n=48) and without

(n=36) depressive symptoms completed two testing weeks, including ad libitum smoking followed by four days of smoking abstinence during the follicular and luteal menstrual phases, we had two main findings. First, there were significant differences in ALLO levels by menstrual phase and smoking status such that

ALLO levels decreased significantly in the follicular phase and increased significantly in the luteal phase during short-term smoking cessation. Second, there were no significant differences between the women with and without depressive symptoms in the absolute or the change in ALLO levels by menstrual phase and smoking status. This study is the first to assess menstrual phase differences in ALLO levels in women with and without depressive symptoms during short-term smoking cessation. Additional research is needed to verify this observation in a sample of premenopausal women who are attempting to quit smoking permanently and explore the effect of ALLO on risk for relapse during attempted smoking cessation.

E. PAPER #2: ALLOPREGNANOLONE AND SMOKING-RELATED

SYMPTOMATOLOGY DURING SHORT-TERM SMOKING ABSTINENCE

E.1. Overview

Introduction: Most women relapse from a quit attempt within the first few days. Smoking-related symptomatology (SRS) has been shown to be associated with relapse risk. Allopregnanolone (ALLO) can protect against negative mood

25 symptoms (one aspect of SRS); however, it is unknown how ALLO may affect

SRS during smoking cessation.

Methods: Women, ages 18-40, were recruited to participate in a controlled cross-over study in which they attended daily clinic visits during a six- day testing week in both the follicular and luteal phases. During the testing week, ALLO was measured on Day 2 (ad libitum smoking) and Day 6 (fourth day of confirmed abstinence). Participants completed several measures of SRS daily during the testing week. Growth curve models were used to investigate the association between ALLO and SRS.

Results: Participants (n=64) were, on average, 30.3 (± 6.7) years old and most (55%) were White. The absolute level of ALLO on the day before quit was significantly associated with the following: (1) perceived stress on the day before quit ( β=-2.25, p<0.01), (2) the change in perceived stress during smoking abstinence ( β=0.79, p<0.01), and (3) premenstrual symptoms of pain and water retention on the day before quit ( β=1.09, p<0.01; β=1.08, p<0.01; respectively).

The change in ALLO during smoking abstinence was significantly associated with the following: (1) positive and negative affect on the day before quit ( β=1.15, p<0.01; β=1.04, p=0.04; respectively), (2) perceived stress on the day before quit

(β=-1.77, p=0.01), (3) the change in perceived stress during smoking cessation

(β=0.17, p<0.01), and (4) the change in depressive symptoms on the day of quit

(β=-1.52, p=0.02).

Conclusions: ALLO had a significant association with affect and premenstrual symptoms during short-term smoking abstinence, but not with items specific to smoking such as craving. Additional research is needed to confirm this relationship in premenopausal women who are quitting smoking permanently and evaluate how these associations may impact risk for smoking relapse.

E.2. Significance & Background

Withdrawal from drugs, including nicotine, is associated with various well- characterized subjective symptoms including irritability, anxiety, depressed mood, restlessness and craving (Concas et al, 2006). The onset of these symptoms typically occurs within 24 hours of the initiation of smoking abstinence, followed by a peak within the first week (Hughes, 2007). Both menstrual phase variability in withdrawal symptoms during ad libitum smoking and more severe withdrawal symptoms during attempted cessation are associated with an increased risk for smoking relapse within younger female smokers (e.g. Allen et al 2010; Allen et al 2009b). Because there is a large overlap between symptoms of withdrawal and premenstrual symptoms, it is recommended to consider the severity of premenstrual symptoms when women attempt to quit smoking

(Carpenter et al, 2006). In fact, more severe premenstrual symptoms have also been shown to be predictive of risk for smoking relapse (Allen et al, 2009b; Allen et al, 2010).

As described above in Paper #1 (Section D.2.), ALLO is associated with

27 negative mood. Similar associations have been observed in premenstrual symptomatology. For example, higher ALLO levels are associated with less severe premenstrual symptoms during the luteal phase and lower levels of anxiety in women with Premenstrual Dysphoric Disorder (PMDD; Wang et al,

1996; Girdler et al, 2001). However, decreases in ALLO are associated with improvement of PMDD symptoms (Freeman et al, 2002). These seemingly discrepant findings may indicate a threshold effect of ALLO on premenstrual symptomatology, similar to that hypothesized to be present in the relationship between ALLO and depressive symptoms. Despite these observations, no literature exists on the possible relationship between ALLO and smoking-related symptomatology (SRS) during smoking cessation in premenopausal women.

As recently hypothesized, increased SRS during attempted smoking cessation may be the result of a deficiency in dopamine release (Benowitz,

2009). ALLO modulates the dopamine release over the course of the menstrual cycle (Zheng, 2009). Therefore, it may be possible that ALLO is affecting SRS severity via changes in dopamine release. Thus, the goal of this paper is to characterize the association between endogenous ALLO levels with SRS, specifically withdrawal symptoms (total withdrawal, craving, and smoking urges), affect (depressive symptoms, positive and negative affect, perceived stress), and premenstrual symptoms (affect, pain and water retention) during short-term smoking abstinence in a sample of premenopausal women who participated in a cross-over study in which they completed two testing weeks during the follicular

(low ALLO) and luteal (high ALLO) phases. We hypothesize that: (1) there will be associations between the absolute level of ALLO on the day before quit day with both the absolute level of SRS during ad libitum smoking and with the change in SRS during short-term smoking cessation, and (2) the change of ALLO from the day before quit date to the fourth day of smoking abstinence will be significantly associated with both the absolute level and change in SRS.

E.3. Methods

E.3.1. Study Participants. To be eligible for inclusion in this paper, participants had to have all four samples (i.e., Day 2 and Day 6 samples in Week

1 and Day 2 and Day 6 samples Week 2). This resulted in 64 participants of the

87 in the subgroup included in this paper.

E.3.2. Measures - Smoking-Related Symptomatology. Study participants completed the following forms daily for five days (Day 2 to Day 6): (1) Minnesota

Nicotine Withdrawal Scale (MNWS): With this item, participants rated their nicotine withdrawal symptoms on eight-items on a scale of ‘0’ (not present) to ‘4’

(severe) resulting in summary scores for craving and overall withdrawal (sum of all items excluding craving; Hughes & Hatsukami, 1998); (2) Questionnaire on

Smoking Urges-Brief (QSU-Brief): This ten-item version of the original form

(QSU) was developed by Tiffany and Drobes (1991). Each item is rated on a seven-point Likert-type scale and results in two factors: ‘Factor 1’ is a measure of primary intention and desire to smoke (QSU-F1), and ‘Factor 2’ is a desire to smoke in anticipation of relief from negative affect (QSU-F2); (3) Positive and

Negative Affect Scale (PANAS): Participants rated 20 words associated with positive or negative affect using a five-point Likert-type scale to yield a ‘Positive

Affect’ (PA) and ‘Negative Affect’ (NA) score (Watson et al, 1998); (4) Cohen

Perceived Stress Scale (PSS): This item uses a five-point Likert-type scale to assess ten questions resulting in a score indicating the perception of stress

(Roberti et al, 2006); (5) Center for Epidemiologic Studies – Depression (CES-D):

This form contains 20 items to assess symptoms typically associated with depression. The items are ranked using a four-point Likert-type scale and yield a single score representing level of depressive symptoms (Knight et al, 1997); and

(6) Premenstrual Assessment Form (PAF): This form include ten items that are assessed on a six-point Likert-type scale, resulting in three subscales for Affect

(PAF-Affect), Pain (PAF-Pain), and Water Retention (PAF-Water; Allen et al,

1991).

E.3.2. Measures – ALLO. All four ALLO blood samples were used in this paper. These samples were collected on Days 2 (ad libitum smoking) and 6

(fourth day of smoking abstinence) during each testing week. The analysis of these samples is described above in Section C.3.

E.3.3. Statistical Analysis. Descriptive statistics were calculated for demographics and smoking behavior variables including mean and standard deviation for continuous variables, and percent for categorical variables.

Participants who had data for all four ALLO values were included in the analyses, including those who had incomplete SRS data. Growth curve models were used

30 to examine the effect of ALLO on the pattern of change in SRS during five days of data collection. The effect of ALLO on SRS was assessed two ways - the absolute value of ALLO on Day 2, and the change in ALLO during the testing week (via a change score that was computed by subtracting the Day 2 ALLO value from the Day 6 ALLO value). For SRS items that appeared to have a linear pattern (Craving, Withdrawal, QSU-F1, QSU-F2, PSS, PAF-Affect, PAF-Water,

PAF-Pain), linear growth curve models were used. Those with a non-linear pattern (CES-D, PA, NA), a non-linear growth curve model were used. All models were adjusted for menstrual phase (F vs. L), and testing order (1 st week vs. 2 nd week). Log transformation was used for variables with a non-normal distribution (NA, PAF-Affect, and PAF-Water). For illustrative purposes, raw

ALLO values were split into quartiles and plotted against the average SRS values

(Figures 6 and 7). P-values less than 0.05 were deemed statistically significant.

No adjustments for multiple comparisons were made. SAS V9.1.3 (SAS Institute,

Cary, NC) was used for the analyses.

E.4. Results

E.4.1.

Study Sample.

The 64 participants included in this analysis were,

31 on average, 30.3 (Standard Deviation (S.D.) ± 6.7) years old and smoked an average of 12.5 (S.D. ± 4.9) cigarettes per day. Most were White (55%) and had at least a high school education (77%). The average ALLO value on Day 2 was

1.92 (S.D. ± 1.54) ng/mL and ALLO increased, on average, by 21% during the testing week ( Table 5 ).

E.4.2. Effect of the Absolute Level of ALLO during Ad Libitum Smoking on

Smoking Related Symptomatology during Short-Term Abstinence. Higher ALLO on the day before quit was significantly associated to the following ( Table 6,

Figure 6 ): (1) lower perceived stress on the day before quit ( β=-2.25, t- value=2.76, p-value=0.006), (2) a greater increase in perceived stress during short-term smoking abstinence ( β=0.792, t-value=4.00, p-value<0.001), (3) a higher level of PAF-Pain on the day before quit ( β=1.096, t-value=2.69, p=0.007), and (4) a higher level of PAF-Water on the day before quit ( β=1.109, t-

Value=2.65, p=0.008). There were no other significant associations between the absolute value of ALLO on the day before quit and SRS.

E.4.3. Effect of the Change in ALLO during Short-Term Smoking

Abstinence on Smoking Related Symptomatology. A greater change in ALLO during short-term smoking abstinence was significantly associated with the following ( Table 6, Figure 7 ): (1) a less drastic change in depressive symptoms on day of quit ( β=-1.524, t-value=2.37, p-value=0.018), (2) a higher positive affect score on the day before quit ( β=1.146, t-value=2.63, p-value=0.009), (3) a higher negative affect score on the day before quit ( β=1.041, t-value=2.09, p-

32 value=0.037), (4) a lower level of perceived stress on the day before quit ( β=-

1.77, t-value=2.52, p-value=0.013), and (5) a greater change in perceived stress during short-term smoking abstinence ( β=0.168, t-value=3.28, p-value=0.001).

Two trends that were also noted between the change in ALLO were depressive symptoms on the day before quit ( β=0.494, t-value=1.81, p-value=0.071) and the change in PAF-Pain ( β=1.015, t-value=1.88, p-value=0.061). The change in

ALLO during short-term smoking abstinence was not significantly associated with any other SRS.

E.5. Summary of Results

In this controlled cross-over trial, in which premenopausal female study participants quit smoking during both the follicular (low ALLO) and luteal (high

ALLO) menstrual phases, we observed a significant effect of ALLO on several measures of affect and premenstrual symptoms, but not on items specific to smoking withdrawal. Specifically, on the day before quit date, a higher absolute level of ALLO was associated with lower perceived stress and greater premenstrual symptoms of pain and water. The higher absolute levels of ALLO on the day before quit day also predicted a greater change in perceived stress during short-term smoking cessation. The more dramatic change in ALLO during short-term smoking cessation was a significant predictor of higher negative affect, lower perceived stress and, possibly, lower depressive symptoms on the day before quit day. Finally, a greater change in ALLO during short-term smoking cessation was also a predictor of greater change in perceived stress

33 and a blunted change in depressive symptoms during smoking cessation. This study is the first to assess the effect of ALLO levels on smoking related symptomatology in premenopausal women during short-term smoking cessation.

Because affect and premenstrual symptoms have been shown to be predictive of risk for relapse in premenstrual women (Allen et al, 2009d; Allen et al, 2010), additional research is warranted to verify these results in premenopausal women who are intending to quit smoking permanently and explore the possible therapeutic effects of ALLO during smoking cessation.

F. PAPER #3: ALLOPREGNANOLONE AND NICOTINE RESPONSE DURING

SHORT-TERM SMOKING ABSTINENCE

F.1. Overview

Introduction: Nicotine response may predict susceptibility to smoking relapse. Allopregnanolone (ALLO), a neuroactive steroid metabolized from progesterone, has been shown to be associated with several symptoms of nicotine response. Therefore, it may protect against nicotine response and subsequent relapse to smoking.

Methods: Women smokers were enrolled into the study and completed two nicotine response lab sessions in their follicular (low ALLO) and luteal (high

ALLO) menstrual phases on the fourth day of biochemically confirmed smoking abstinence. During the lab session, participants used nicotine nasal spray and completed a timed series of cardiovascular, cognitive and subjective responses.

A blood sample, for measurement of ALLO, was collected immediately before the start of the nicotine response lab session. The association between ALLO and nicotine response were assessed using covariance pattern models.

Results: Study participants (n=77) were, on average, 29.9 (±6.8) years old and smoked 12.2 (±4.9) cigarettes per day. ALLO had a significant, positive association with the following variables prior to initiation of nicotine nasal spray: systolic blood pressure ( β =0.85, p=0.04), diastolic blood pressure ( β =1.19, p<0.01), and self-report of physical symptoms ( β =0.58, p<0.01), dizziness ( β

=0.88, p<0.01), jitteriness ( β =0.90, p=0.04) and pleasantness ( β =2.05, p=0.041). ALLO also had significant positive associations with the following changes from baseline to post nicotine nasal spray use: discriminability (a measure of attention; β =1.15, p=0.05), and bias (a measure of impulsivity; β

=0.12, p=0.02).

Conclusions: ALLO was significantly associated with a number of measures of nicotine response both prior to and after use of nicotine nasal spray.

The results suggest that ALLO may impact the cardiovascular and the subjective physical state during acute smoking abstinence, as well as the cognitive

38 response to nicotine. Additional research is needed to explore the impact of these associations on smoking cessation outcomes.

F.2. Background & Significance

Nicotine response, defined as the cardiovascular, cognitive, and subjective reaction to nicotine, reinforces cigarette-smoking behavior (Stolerman & Jarvis,

1995). The responses to nicotine may serve as indicators of both sensitivity to nicotine and susceptibility to nicotine dependence via an association with likelihood of smoking relapse (Pillitteri et al, 1997; Pomerleau et al, 1998;

Pomerleau, 1995). Sensitivity to nicotine is theorized to increase after smoking cessation due to a decrease in tolerance (Perkins et al, 2002). Understanding the acute effects of nicotine during short-term abstinence is important in understanding sensitivity to nicotine and subsequent smoking relapse (Perkins et al, 2001; Perkins et al, 2002). Some preliminary evidence has suggested that nicotine response varies by menstrual phase – specifically the follicular phase is associated with increases in impulsivity and decreases in nicotine metabolism compared to the luteal phase in women without depressive symptoms (Allen et al, 2009e), suggesting that sex hormones, and perhaps their metabolites, may play a role.

Although the association between nicotine response and ALLO has yet to be investigated, several individual symptoms of nicotine response have been found to be significantly associated with ALLO. For example, complaints of memory impairment in pregnant women often begin in the second trimester when ALLO

39 peaks and resolve soon after childbirth when ALLO rapidly declines (Birzniece et al, 2006). Women with premenstrual dysphoric disorder (PMDD) also show difficulty concentrating and impairment in working memory during the luteal phase of the menstrual cycle when ALLO is at its highest. (Birzniece et al, 2006).

Changes subjective scores of sedation are correlated with increasing serum

ALLO levels (Sundstrom-Poromaa, 2008). In addition to ALLO’s association with cognition, ALLO has been shown to have a positive correlation with diastolic blood pressure (Childs et al, 2010). Additional evidence comes from the animal literature, such that animals injected with ALLO learned significantly less compared to those who were not injected with ALLO (Sundstrom-Poromaa,

2008). Overall these observations suggest there may be an association between

ALLO and cognition, sedation and blood pressure, all measures of nicotine response.

ALLO may have an effect on nicotine response through the effect on GABAa receptors (see Section B.4 for additional details); this could, subsequently, impact nicotine response and, therefore, relapse to smoking. Thus, the overall goal of this paper is to assess the association between ALLO and nicotine response during short-term smoking cessation. We hypothesize an inverse association with ALLO and cognitive measures (i.e., higher ALLO will be associated with lower cognitive function), and a positive association between

ALLO and cardiovascular and subjective measures (i.e., higher ALLO will be associated with greater cardiovascular response and favorable subjective

40 response).

F.3. Methods

F.3.1. Study Participants. To be eligible for inclusion in this paper, participants had to have ALLO values available for both Day 6 visits. Of the 87 participants included in this project, 77 met this criteria and were included in this paper.

F.3.2. Measures - Nicotine Response . The nicotine lab session included a four-hour timed protocol in which the study participants used nicotine nasal spray

(2 mg dose; two sprays per nostril separated by two minutes) at Time 0 minutes.

Dependent measures, assessing cardiovascular, cognitive, and subjective response to nicotine, were collected starting at Time -30 minutes and continuing until Time 90 minutes ( Figure 5 ). This series was repeated twice. Given the purpose of this paper is to assess initial response to nicotine, the data from two time points were analyzed: (1) prior to using nicotine nasal spray (i.e baseline values), and (2) pre-determined peak time points (i.e., expected greatest change from baseline values).

First, cardiovascular response was measured by the following measures: (1)

Blood pressure (Diastolic (DBP) and Systolic (SBP)) and heart rate (HR) was measured using an automatic blood pressure machine (Time -1 and 5 minutes).

Second, cognitive response was assessed with the following three measures:

(1) Mathematical Skills Task (Math): This test is a revised version of the serial addition/subtraction task from the Walter Reed Performance Assessment

Battery. The Math task assesses cognition, specifically general information processing. Each trial involves a presentation of 50 simple addition or subtraction problems and the participant indicates if the problem is true or false

(Thorne et al, 1985) resulting in a total number correct, as well as response time

(Time -30 and 20 minutes); (2) Fingertapping Task (FT): Using the index finger on the dominant hand, participants tapped a key on a computer as quickly as possible for 30 seconds for two successive trials (Hindmarch, 1980). The average of the two trials is used as an indicate of motor speed (Time -30 and 20 minutes); and (3) Immediate Memory Task (IMT): In this task, a measurement of attention and impulsivity, a series of randomly generated 5-digit numbers (e.g.,

54983) is displayed on a computer monitor for ½ a second at a rate of one per second. The participant was instructed to press a button when the number displayed is identical to the preceding number. Outcome variables include

Discriminability (A’), which ranges from 0.5 (chance) to 1.0 (perfect discrimination) and Bias (beta) which ranges from 0.0 (liberal bias) to 2.0

(conservative bias; Dougherty et al, 2002; Time -30 and 10 minutes).

Third, the following two measures were used to assess the subjective response to nicotine: (1) Subjective State Scale (SSS; al’Absi et al, 2004; al’Absi et al, 2005): This measure contains 24-items to which participants respond to using a six-point Likert-type scale. These items yield five subscales: Negative

Affect, Positive Affect, Physical Symptoms, Withdrawal and Craving (Time -30 and 5 minutes), and (2) Visual Analog Scale (VAS): This subjective item was

42 used to measure potentially rapid changes in negative or positive drug effects including the following items: Alert, Dizzy, Head Rush, Jittery, Pleasant, Relaxed,

Stimulated and Urge to Smoke. The participant used a 100-mm line to mark her experience ranging from ‘not at all’ to ‘very much’ (Jones et al, 1999; Time -30 and 5 minutes).

F.3.3. Measures – ALLO. ALLO values from samples collected on the day of the nicotine response lab session (i.e., Day 6, fourth day of smoking abstinence) during each testing week were used in this paper. The analysis of these samples is described above in Section C.3.

F.3.4. Statistical Analysis. Descriptive statistics were calculated for demographics and smoking behavior variables including mean and standard deviation for continuous variables, and percent for categorical variables. All participants who had data for ALLO values on Day 6 were included in the analyses, including those who had incomplete nicotine response data. To measure the response to nicotine, change scores for each item were calculated by subtracting the baseline value (i.e., Time -30 or -1 minutes) from the expected peak value (i.e., Time 5, 10 or 20 minutes). The effect of ALLO on the baseline value and the change score was assessed via a covariance pattern models with an unstructured pattern. All models were adjusted for menstrual phase (F vs. L), and testing order (first week vs. second week). The model with change score as an outcome also adjusted for the corresponding baseline value for each variable.

Log transformation was used for variables with a non-normal distribution (IMT

Discriminability, VAS Dizzy, and VAS Jittery). P-values less than 0.05 were deemed statistically significant. No adjustments for multiple comparisons were made. SAS

V9.1.3 (SAS Institute,

Cary, NC) was used for the analyses.

F.4. Results

F.4.1. Study Sample. A total of 77 women were included in this analysis.

Participants were, on average, 29.9 (Standard Deviation (S.D.) ± 6.8) years old

(Table 7 ). Most were non-Hispanic White (54%) with more than a high school education (67%). They smoked, on average, 12.2 (S.D. ± 4.9) cigarettes per day with the first one occurring 40.9 (S.D. ± 41.0) minutes of waking. The average

ALLO level was 2.3 (S.D. ± 1.9) ng/mL.

3.2 Association between ALLO and Baseline Levels of Nicotine Response.

A one-unit increase in ALLO was significantly associated with the following

(Table 8 ): (1) 0.8% increase in SBP (p=0.037), (2) 1.7% increase in DBP

(p<0.001), (3) 13.3% increase in SSS Physical Symptoms (p<0.001), (4) 8.1% increase in VAS Dizzy (p=0.008), (5) 5.0% increase in VAS Jittery (p=0.041), and

(6) 3.1% increase in VAS Pleasant (p=0.041). Higher ALLO levels had

44 associations that approached statistical significance with the following items: (1) greater Response Time on the Math ( β=0.078, p=0.064), (2) more VAS Head

Rush ( β =1.391, p=0.094), and (3) less VAS Urge to Smoke ( β=-2.228, p=0.087).

3.3 Association between ALLO and Nicotine Response. Higher ALLO levels were significantly associated with the following greater changes in nicotine response from baseline to 10 minutes post nicotine nasal spray ( Table 8 ): (1)

IMT-Discriminability (i.e., attention; β=1.154, p=0.046), and (2) IMT-Bias (i.e., impulsivity; β=0.124, p=0.022). A trend was also noted suggesting that a higher

ALLO level may be associated with a decrease in SSS Negative Affect at five minutes after using the nicotine nasal spray ( β=-0.354, p=0.086).

F.4. Summary of Results

In this controlled cross-over study, in which premenopausal smokers completed two nicotine response lab sessions on the fourth day of biochemically verified smoking abstinence during the follicular (low ALLO) and luteal (high

ALLO) menstrual phases, we observed several significant associations between

ALLO and nicotine response (i.e., cardiovascular, cognitive, and subjective response). Specifically, prior to using the nicotine nasal spray, higher ALLO levels were associated with higher blood pressure, favorable subjective measures, and, possibly, decreased information processing. Higher ALLO also significantly impacted the change in cognition after using the nicotine nasal spray via greater changes in attention and increases in impulsivity. While most of these significant associations occurred prior to the administration of a nicotine challenge, these relationships may still play a role in risk for smoking relapse.

Therefore, future research should explore the mechanisms of these relationships and how these associations may play a role in smoking cessation efforts.

G. DISCUSSION

G.1. Overview of Discussion.

This dissertation project includes a series of papers which focused on the association between ALLO and several risk factors associated with smoking relapse (depressive symptoms, smoking-related symptomatology, and nicotine response). These papers yielded a number of significant associations between

ALLO and study outcomes, resulting in the identification of several themes. A discussion of each of these themes is presented below and included ALLO’s association with the following: menstrual phase, depressive symptoms, perceived stress, premenstrual symptoms, smoking-specific symptomatology and other aspects of nicotine response. These data have some strengths and limitations, as well as applicability to future research. This information is also presented below.

G.2. ALLO and Menstrual Phase.

As hypothesized, and consistent with prior research (Genazzani et al, 1998), we observed significantly higher ALLO levels in the luteal phase compared to the follicular phase. One of our novel observations was a significant 10% decrease in ALLO during the follicular phase whereas in the luteal phase ALLO significantly increased by 31% during short-term smoking abstinence. Given we observed the change in ALLO during the luteal phase (from approximately menstrual cycle day 17 to 21), the significant increase may be attributed to the

47 natural fluctuation of ALLO levels during the course of the menstrual cycle as it typically increases dramatically from ovulation (i.e., approximately menstrual cycle day 14) until later in the luteal phase (i.e., approximately menstrual cycle day 22; Genazzani et al, 1998; Nyberg et al 2007). However, the observation in the follicular phase is not as easily explained as ALLO levels characteristically remain flat during this phase (Genazzani et al, 1998; Nyberg et al 2007). While additional research is needed to replicate these findings, these observations offer further support for the current theory that the luteal menstrual phase is associated with improved smoking cessation outcomes in the absence of nicotine replacement therapy (Franklin & Allen, 2009). ALLO has been shown to be associated with the reduction of negative drug abuse behaviors in a number of preclinical studies, especially in females. For example, in females, but not males, ALLO is associated with a reduction of reinstatement behavior in cocaine- seeking rats (Anker & Carroll, 2010). Thus, given that ALLO is thought to be protective against drug abuse behaviors (Sinnott et al, 2002; Grant et al, 1997;

Bowen et al, 1999), an increase in ALLO during smoking cessation during the luteal phase may protect against relapse whereas a drop in the follicular phase, as observed in this study, may increase risk for relapse. However, whether or not ALLO is protective against drug abuse behavior in humans remains to be seen.

G.3. ALLO and Depressive Symptoms.

Contrary to our hypotheses, we did not identify a significant difference in either absolute levels or menstrual phase variability of ALLO based on depressive symptom status. This observation conflicts with a recent study reporting that women with a history of depressive symptoms had a blunted stress-induced ALLO response compared to those without a history of depressive symptoms (Klatskin et al 2006). While both of these studies contained samples of premenopausal women, there are a number of important methodological differences to consider. First, our sample consisted entirely of established smokers whereas Klatskin and colleagues included few, if any, smokers (<14%). Given that cotinine levels have been shown to be positively correlated with ALLO levels in a sample of men (Marx et al 2006), it is possible that cigarette smoking may impact the production of ALLO. However, our results did not change when adjusting for cigarettes smoked per day and the association between ALLO and cotinine in women is unknown. Second, Klatskin and colleagues used an acute mental stressor (<20 minutes) to induce an ALLO response whereas we investigated change in ALLO during the first four days of smoking abstinence, which could be considered a more chronic stressor. It is unknown how ALLO response to stress varies by different types (i.e., acute versus chronic) stressors. Therefore, it is difficult to know what kind of effect, if any, this may have had on our results. Third, Klatskin and colleagues compared those with a history of major depressive disorder (MDD) to those without a history of MDD, whereas we compared any depressive symptoms (i.e., history of

MDD, or history or current subclinical depressive symptoms) to those without any depressive symptoms. While our results remain unchanged in ad hoc analyses that compared history of MDD versus those without a history of MDD within the depressive symptoms group (data not shown), it remains unknown how different sub-types of depression and/or severity of depressive symptoms may impact

ALLO production. Interestingly, Klatskin and colleagues initially aimed to assess menstrual phase differences in stress-induced ALLO response; however, they were unable to assess follicular phase levels given numerous undetectable

ALLO levels. In the present study, we were able to consistently detect ALLO levels during both the follicular and luteal menstrual phases. Despite this, we were unable to identify a significant difference between women with and without depressive symptoms in either menstrual phase. In sum, additional research is needed to further explore ALLO response to stress in women with and without depressive symptoms. Given the effect of depressive symptoms on risk for smoking relapse along with the potential for ALLO to protect against relapse, elucidating this relationship among women who smoke is of particular interest.

In the second paper, we observed a substantial effect of ALLO during short- term cessation on affect, including depressive symptoms. While previous literature has consistently observed significant associations between allopregnolone levels and negative mood symptoms in women, the pattern (i.e., linear versus biphasic) of this relationship remains unknown (Andreen et al 2005;

Nappi et al 2001). Interestingly, in a sample of male smokers, ALLO was not

50 significantly associated with negative affect (Marx et al, 2006); suggesting effect of ALLO on affect may differ by sex. This would concur with the preclinical literature (Anker & Carroll, 2010). This paper expands the literature by examining the effect of ALLO on affect during smoking cessation, while adjusting for menstrual phase. Smoking cessation is associated with a dramatic increase in negative affect, including depressive symptoms (Hughes, 2007). Women are particularly susceptible to the effect of depressive symptoms during smoking cessation (Borelli et al, 1996; Husky et al, 2008). Further research should continue to explore the role ALLO plays in affect, including depressive symptoms. If further studies confirm such an association, then the clinical implications of exogenous manipulation with ALLO to limit negative affect during attempted smoking cessation should be explored as a possible avenue for improved smoking cessation rates in women.

In the third paper where we observed several significant associations between ALLO and affect prior to the administration of nicotine nasal spray.

Overall higher ALLO levels were associated with more favorable subjective measures (i.e., more pleasantness). With the possible exception of negative affect, there were no significant associations between ALLO and the change in affect from 30 minutes prior to administration of nicotine nasal spray to five minutes post administration. We observed a trend between ALLO and the change in negative affect, suggesting that less change in negative affect may occur when higher levels of ALLO are present.

Although speculative, the data from these three papers seem to suggest that higher levels of ALLO may be associated with an improved affect during short-term smoking abstinence. If this is confirmed, ALLO may protect against smoking relapse via preventing an initial slip which tends to occur in response to negative mood in women (Nakajima & al’Absi, 2012).

G.4. ALLO and Perceived Stress.

In the second paper, both the absolute level and change in ALLO during short-term smoking abstinence had a significant effect on levels of perceived stress. Acute stressors had no effect on ALLO production in men (Childs & de

Wit, 2009); however, in women ALLO response to stress was dependent on menstrual phase such that the luteal phase, ALLO production was more responsive to acute stressors (Childs et al 2010). In the first paper in this series, we demonstrated that ALLO levels change during short-term smoking cessation and the direction of the change is dependent on menstrual phase (decrease during follicular phase versus increase in luteal phase). The results of this paper expand these observations by examining the effect of ALLO on perceived stress.

Based on the first paper of this series, ALLO increases with short-term smoking cessation that occurs in the luteal phase. Further, increases in ALLO levels during short-term smoking cessation are associated with initial lower levels of perceived stress. Therefore, these observations offer further evidence for a possible mechanism to explain the improved smoking cessation outcomes in the luteal phase (Allen et al 2008). The effect of ALLO on perceived stress remained

52 significant even after adjusting for menstrual phase. However, it is possible that the seemingly different patterns of association between ALLO and perceived stress may be reflective of hormonal changes during cessation. Given that stress predicts quitting distress (al’Absi et al 2004) and subsequent relapse

(al’Absi et al 2005), combined with ALLO’s ability to modulate GABAa receptors

(Zheng, 2009), ALLO may have a place in smoking cessation treatment, where there is currently a lack of therapeutic options to directly address this relapse trigger (Rose, 2010).

G.5. ALLO and Premenstrual Symptoms.

To our knowledge, this paper is the first to document a relationship between

ALLO and premenstrual symptoms in a sample of premenopausal women without PMDD. Three studies to-date have investigated the association between ALLO and premenstrual symptoms in women with PMDD during the luteal phase. One of these studies (Nyberg et al, 2007), observed a decrease in

ALLO was associated with a decrease in negative symptomatology, whereas the other two studies observed an increase in ALLO was associated with a decrease in negative symptomatology (Gridler et al, 2001; Wang et al, 1996). One possible explanation for these seemingly discrepant findings may be a threshold effect of ALLO on symptomatology given the ALLO levels observed in Nyberg and colleagues’ study was substantially lower than those observed in the other studies (0.3-0.4 ng/ml versus 1.1 ng/ml and 1.0-2.6 ng/ml, respectively). Like

Nyberg and colleagues (2007), we observed positive relationship between ALLO

53 and negative symptomatology, which remained significant even after adjusting for menstrual phase. Unlike the previously published literature, we observed a wide range of ALLO levels (0.3-8.6 ng/mL). Given premenstrual symptoms have been shown to be related to risk for relapse in women (Allen et al, 2009; Allen et al, 2010 ), the effect of ALLO on these symptoms offers another potential advantage to manipulating levels of ALLO in women who are attempting to quit smoking. Further, the application of these data may extend into other fields (e.g. treatment of premenstrual symptoms and PMDD).

G.6. ALLO and Smoking-Specific Symptomatology.

One of the most interesting observations in this dissertation project was the lack of an association between ALLO and items specific to smoking withdrawal, such as craving. There were virtually no associations between ALLO and items specific to smoking withdrawal during short-term smoking abstinence. While we did observe an inverse association between urge to smoke and ALLO, this observation was a trend. Further, no other associations were observed (most notably was the lack of the association between ALLO and ‘craving’ during the same time points in the nicotine response lab session). Therefore, it is quite possible that this trend was observed by chance. Prior literature demonstrated a lack of an effect of ALLO on craving and nicotine dependence in a sample of men (Marx et al 2006). The present study supports and expands this observation in a sample of premenopausal women. We have previously observed higher levels of withdrawal and smoking urges during the luteal phase

54 versus the follicular phase (Allen et al, 2009), and that craving has a differential menstrual phase effect on risk for relapse (Allen et al, 2010). Taken together these data suggest that there is a menstrual phase effect on smoking withdrawal, but this effect is not caused by ALLO. An alternative explanation may be the ratio of progesterone to estradiol as suggested by Schiller and colleagues (2012).

G.7. ALLO and Other Nicotine Response Measures.

ALLO also had effects on cardiovascular and cognitive measures during the nicotine response lab session. First, as hypothesized, both systolic and diastolic blood pressure, but not heart rate, had a significant and positive association with

ALLO prior to the administration of nicotine nasal spray. The change in cardiovascular measures of response from one minute prior to the administration of nicotine nasal spray to five minutes after were not significantly related to

ALLO. Overall, these data may support prior clinical literature. Specifically, in a study by Childs and colleagues (2010) a positive correlation (p=0.06) between change in diastolic blood pressure and change in ALLO was observed in a sample of men and women who were participating in a lab session designed to induce a stress response. In paper three, we did not observe a significant association between the change in ALLO and change in cardiovascular measures as the Childs and colleagues (2010) study did. Given that both studies observed a significant association of ALLO with cardiovascular measures suggests interplay between the two. Additional research is needed to

55 understand the mechanisms of action and how this information may relate to smoking cessation.

Second, in terms of cognition, ALLO appeared to have significant relationships with several cognitive response measures. Specifically, ALLO levels had a significant association with the change in both IMT Discriminability and Bias after the dose of nicotine was administered. This suggests that when

ALLO levels are high and a dose of nicotine is introduced during a withdrawal state, a greater change in attention and impulsivity occurs. On the other hand, at baseline (i.e., Time -30 minutes), we observed a near significant and positive relationship between ALLO and response time on the Math Task. This suggests that higher levels of ALLO may be associated with slower information processing, which concurs with our hypothesis and previously published literature which have demonstrated that higher levels of ALLO are associated with lower cognitive function. For example, in the preclinical literature, animals injected with ALLO learned significantly less than those who were not injected. In the clinical literature pregnant women and women with PMDD had lower levels of memory and concentrations with higher levels of ALLO (Brizniece et al 2006). While we observed the same in our study prior to the administration of nicotine nasal spray

(i.e., higher ALLO was associated with slower information processing), the opposite was observed after administration of nicotine (i.e., higher ALLO was associated with greater change in attention and impulsivity). Nicotine itself is associated with improved cognitive functioning (e.g. Wignal & de Wit, 2011; Bell

56 et al, 1999; Wesnes & Warburton, 1983), ALLO may be involved in this relationship via its effect on GABAa receptors which also effect learning.

Additional research is needed to explore the mechanisms involved in this relationship.

G.8. Strengths and Limitations.

While this dissertation project contained several strengths including the controlled cross-over study design, which limits bias and confounding, with a fairly diverse study sample, along with the detailed measurement of smoking status and ALLO levels and a well-designed nicotine response lab session including a variety of measures, it also has some limitations. First, our study sample included a convenience sample of women who did not want to quit smoking permanently. Given that our study participants knew they could return to smoking within a matter of days, it is possible that the ALLO response to smoking cessation may be different in women who are intending to abstain from smoking indefinitely. Therefore, the results of this study cannot be generalized to women who are attempting to quit smoking permanently. Second, we have a relatively small sample size, which may have limited our power to detect smaller differences in ALLO levels and the effect of ALLO on smoking-related symptomatology and nicotine response. However, due to the cross-over study design, we have greater than 99% power to detect menstrual phase differences in study outcomes with a smaller sample size than was included (n=52 versus n=64-84). Therefore, any significant differences that were not detected due to

57 limits in power may have questionable clinical and public health significance.

Third, ALLO is but one of many possibly explanations of the previously observed menstrual phase differences in smoking cessation outcomes. Additional research is needed to explore the effects of ALLO on smoking-related symptomatology and nicotine response within the context of other possible risk factors, such as progesterone, estradiol and the progesterone/estradiol ratio.

Despite these limitations this project fills gaps in the literature by being the first to systematically investigate the role of ALLO on several risk factors associated with smoking relapse.

G.10. Future Directions.

Due primarily to its reinforcing properties, ALLO has been identified as a possible substitute for drugs with abuse potential in studies of drug discrimination in animals (Bowen et al, 1999; Grant et al, 1997; Sinnott et al, 2002). The present project provides preliminary evidence for a similar effect of ALLO in humans. These observations will be immediately applied to two newly funded research projects. First, a P50 grant funded by the National Institute on Drug

Abuse and Office on Research on Women’s Health (PI Marilyn Carroll) aims to explore the effect of exogenous progesterone (versus a placebo) on preventing smoking and cocaine relapse in men and women. Weekly blood samples (that will be later analyzed for ALLO) will be collected for a 12-week period during attempted smoking cessation. In this study we will be able explore the effect of

ALLO risk for smoking and cocaine relapse, as well as secondary outcomes of

58 changes in mood and stress, in both men and women. Second, a newly funded

R01 grant by the National Institute on Drug Abuse (PI Sharon Allen) will investigate the role of progesterone (endogenous in pregnant women and exogenous in non-pregnant women) on smoking-related symptomatology and nicotine response. This study will directly build on the present dissertation by studying the same smoking outcomes within the context of extreme progesterone levels. These extreme progesterone levels will likely result in the production of extreme ALLO levels. Future projects that have not yet been funded could focus on explore the following: (1) the effect of exogenous ALLO on smoking outcomes using a lab-based setting, (2) the effect of endogenous or exogenous

ALLO paired with various pharmacotherapies on smoking cessation outcomes, or

(3) the effect of ALLO on other addictive substances and behaviors.

G.11. Conclusions.

In conclusion, ALLO levels were higher in the luteal phase, regardless of smoking status and depressive symptoms status. Higher ALLO levels did not have a consistent effect on affect but, overall, it was associated with favorable levels of affect and perceived stress, as well as measures of nicotine response during short-term smoking abstinence. After nicotine administration during short- term smoking abstinence, higher ALLO levels were associated greater changes in cognition. Future research is needed to explore the mechanisms of these relationships, confirm these findings in premenopausal women who are

59 attempting to quit smoking permanently, and investigate how ALLO may impact risk for smoking relapse.

H. REFERENCES al’Absi, M., Hatsukami, D., Davis, G.L., & Wittmers, L.E. (2004). Prospective

examination of effects of smoking abstinence on cortisol and withdrawal

symptoms as predictors of early smoking relapse. Drug and Alcohol

Dependence; 73: 267-278. al’Absi, M., Hatsukami, D., & Davis, G.L. (2005). Attenuated adrenocorticotropic

responses to psychological stress are associated with early smoking relapse.

Psychopharmacology; 181: 107-117.

Allen, A. M., Allen, S. S., al’Absi, M., & Hatsukami, D. K. (2009e). Menstrual

phase differences in nicotine response after acute smoking abstinence. Paper

presented at the College of Problems on Drug Dependence Annual Meeting,

Reno, Nevada. Retrieved from http://www.srnt.org/conferences/past/2009

/pdf/2009_Poster_Sessions.pdf

Allen, A.M., Allen, S.S., Lunos, S., & Pomerleau, C.S. (2010). Severity of

withdrawal symptomatology in follicular versus luteal quitters: The combined

effects of menstrual phase and withdrawal on smoking cessation outcome.

Addictive Behaviors; 35: 549-552.

Allen, A.M., Allen, S.S., Widenmier, J., & al’Absi, M. (2009a). Patterns of cortisol

and craving by menstrual phase in women attempting to quit smoking.

Addictive Behaviors; 34: 632-635.

Allen, S.S., Allen, A.M., Lunos, S., & Hatsukami, D. (2009d). Patterns of self-

selected smoking cessation attempts and relapse by menstrual phase.

Addictive Behaviors; 34: 928-931.

Allen, A.M., Mooney, M., Chakraborty, R., & Allen, S.S. (2009c). Circadian

patterns of ad libitum smoking by menstrual phase. Human

Psychopharmacology; 24: 503-506.

Allen, S.S., Allen, A.M., Kotlyar, M., Lunos, S., al’Absi, M., & Hatsukami, D. (In

Press). Menstrual phase and depressive symptoms difference in

physiological response to nicotine following acute smoking abstinence.

Nicotine & Tobacco Research.

Allen, S.S., Allen, A.M., & Pomerleau, C.S. (2009b). Influence of phase-related

variability in premenstrual symptomatology, mood, smoking withdrawal and

smoking behavior during ad libitum smoking, on smoking cessation outcome.

Addictive Behaviors; 34: 107-111.

Allen, S.S., Bade, T., Center, B., Finstad, D., & Hatsukami, D. (2008). Menstrual

phase effects on smoking relapse. Addiction; 103: 809-821.

Allen, S.S., McBride, C.M., & Pirie, P.L. (1991). The shortened premenstrual

assessment form. Journal of Reproductive Medicine; 36: 769-772.

American Psychological Association (APA). Diagnostic and Statistical Manual of

Mental Disorders. 4th ed. Washington, D.C.: American Psychological

Association, 1994. Print.

Andreen, L., Sunstrom-Poromaa, I., Bixo, M., Andersson, A., Nyberg, S., &

Backstrom, T. (2005). Relationship between allopregnanolone and negative

mood in postmenopausal women taking sequential hormone replacement

therapy with vaginal progesterone. Psychoneuroendocrinology; 30: 212-224.

Anker, J.J., & Carroll, M.E. (2010). Sex differences in the effects of

allopregnanolone on yohimbine-induced reinstatement of cocaine seeking in

rats. Drug and Alcohol Dependence; 107: 264-267.

Anker, J.J., Perry, J.L., Gliddon, L.A., & Carroll, M.E. (2008). Impulsivity predicts

the escalation of cocaine self-administration in rats. Pharmacology,

Biochemistry and behavior; 93: 343-348.

Backstrom, T., Andersson, A., Baird, D.T., & Selstam, G. (1986). The human

corpus luteum secretes 5 alpha-pregnane-3,20-dione. Acta Endocrinology;

111:116-121.

Baron, J.A., Becchia, C., & Levi, F. (1990). The anti-estrogenic effect of cigarette

smoking in women. American Journal of Obstetrics and Gynecology; 162:

502-514.

Benowitz, N.L., Bernert, J.T., Caraballo, R.S., Holiday, D.B. & Wang, J. (2008).

Optimal serum cotinine levels for distinguishing cigarette smokers and

nonsmokers within different racial/ethnic groups in the United States between

1999 and 2004. American Journal of Epidemiology; 169: 236-248.

Benowitz, N.L., Lessov-Schlagger, C.N., Swan, G.E., & Jacob, P. (2006).

Female sex and oral contraceptive use accelerate nicotine metabolism.

Clinical Pharmacology & Therapeutics; 79: 480-488.

Benowitz, N.L. (2009). Pharmacology of Nicotine: Addiction, Smoking-Induced

Disease, and Therapeutics. Annual Review of Pharmacology and

Toxicology; 49: 57-71.

Birzniece, V., Backstrom, T., Johansson, I.M., et al. (2006). Neuroactive steroid

effects on cognitive functions with a focus on the serotonin and GABA

systems. Brain Research Reviews: 51: 212-239.

Borrelli, B., Bock, B., King, T., Pinto, B., & Marcus, B. H. (1996). The impact of

depression on smoking cessation in women. American Journal of Preventive

Medicine; 12: 378-387.

Bowen, C.A., Purdy, R.H., & Grant, K.A. (1999). Ethanol-like discriminative

stimulus of endogenous neuroactive steroids: Effect of ethanol training dose

and dosing procedure. Journal of Pharmacology and Experimental

Therapeutics; 289; 405-411.

Carpenter, M.J., Saladin, M.E., Leinbach, A.S., LaRowe, S.D., & Upadhyaya,

H.P. (2008). Menstrual phase effects on smoking cessation: A pilot feasibility

study. Journal of Women’s Health; 17: 293-301.

Carpenter, M.J., Upadhyaya, H.P., LaRowe, S.D., Saladin, M.E., & Brady, K.T.

(2006). Menstrual cycle phase effects on nicotine withdrawal and cigarette

craving: A review. Nicotine & Tobacco Research; 8: 627-638.

Centers for Disease Control and Prevention (CDC). Tobacco Control State

Highlights, 2010. Atlanta: U.S. Department of Health and Human

Services, Centers for Disease Control and Prevention, Center for Chronic

Disease Prevention and Health Promotion, Office on Smoking and Health,

2010.

http://www.cdc.gov/tobacco/data_statistics/state_data/state_highlights/201

0/pdfs/highlights2010.pdf

Childs, E. & de Wit, H. (2009). Hormonal, cardiovascular, and subjective

responses to acute stress in smokers. Psychopharmacology; 203: 1-12.

Childs, E., Dlugos, A., & de Wit, H. (2010). Cardiovascular, hormonal, and

emotional responses to the TSST in relation to sex and menstrual cycle

phase. Psychophysiology; 47: 550-559.

Concas, A., Sogliano, C., Porcu, P. Marra, C., Brundo, A. & Biggio, G. (2006).

Neurosteroids in nicotine and morphine dependence. Psychoparmacology;

186: 281-292.

Dougherty, D.N., Marsh, D.M., & Mathias, C.W. (2002). Immediate and delayed

memory tasks: A computerized behavioral measure of memory, attention and

impulsivity. Behavioral Research Methods: Instruments and Computers; 34:

391-398.

Finn, D.A., Beckley, E.H., Kaufman, K.R., & Ford, M.W. (2010). Manipulation of

GABAergic steroids: Sex differences in the effects of alcohol drinking- and

withdrawal-related behavior. Hormones & Behavior; 57: 12-22.

Franklin, T.R., & Allen, S.S. (2009). Influence of menstrual cycle phase on

smoking cessation treatment outcome: A hypothesis regarding the discordant

findings in the literature. Addiction; 1941-1942.

Franklin, T.R., Ehrman, R., Lynch, K.G., Harper, D., Sciortion, N., O’Brien, C.P.,

& Childress, A.R. (2008). Menstrual cycle phase at quit date predicts

smoking status in an NRT treatment trial: A retrospective analysis. Journal of

Women’s Health; 17: 287-292.

Franklin, T.R., Napier, K., Ehrman, R., Gariti, P., O’Brien, C.P., & Childress, A.R.

(2004). Retrospective study: Influence of menstrual cycle on cue-induced

cigarette craving. Nicotine & Tobacco Research; 6: 171-175.

Freeman, E.W., Frye, C.A., Rickels, K., Martin, P.A., & Smith, S.S. (2002).

Allopregnanolone levels and symptom improvement in severe premenstrual

syndrome. Journal of Clinical Psychopharmacology; 22: 516-520.

Freeman, E.W., Purdy, R.H., Coutifaris, C., Rickels, K., & Paul, S.M. (1993).

Anxiolytic metabolites of progesterone: Correlation with mood and

performance measures following oral progesterone administration to healthy

female volunteers. Neuroendocrinology; 58: 478-484.

Genazzani, A.R., Petraglia, F., Bernardi, F. et al (1998). Circulating levels of

allopregnanolone in humans: Gender, age and endocrine influences. Journal

of Clinical Endocrinology and Metabolism; 83: 2099-2103.

Glassman, A.H., Covey, L.S., Stetner, F., & Rivelli, S. (2001). Smoking

cessation and the course of major depression: A follow-up study. Lancet;

357: 1929-1932.

Girdler, S.S., & Klatzkin, R. (2007). Neurosteroids in the context of stress:

Implications for depressive disorders. Pharmacology & Therapeutics; 116:

125-139.

Girdler, S.S., Mechlin, M.B., Light, K.C., & Morrow, L.A. (2006). Ethnic

differences in allopregnanolone concentrations in women during rest and

following mental stress. Psychophysiology; 43: 331-336.

Girdler, S.S., Straneva, P.A., Light, K.C., Pedersen, C.A., & Morrow, A.L. (2001).

Allopregnanolone levels and reactivity to mental stress in premenstrual

dysphoric disorder. Biological Psychiatry; 49: 788-797.

Grant, K.A., Azarov, A., Shively, C.A, & Purdy, R.H. (1997). Discriminative

stimulus effects of ethanol and 3 alpha-hydroxy-5 alcpha-pregnan-20-one in

relation to menstrual cycle phase in cynomolgus monkeys.

Psychopharmacology; 130: 59-68.

Gray, K.M., DeSantis, S.M., Carpenter, M.J., Saladin, M.E., LaRowe, S.D. &

Upadhyaya, H.P. (2010). Menstrual cycle and cue reactivity in women

smokers. Nicotine & Tobacco Research; 12: 174-178.

Hill, M., Popov, P., Havlikova, H., Kancheva, L., Vribikova, J., Kancheva, R., et

al. (2005). Altered profiles of serum neuroactive steroids in premenopausal

women treated for alcohol addiction. Steroids; 70: 515-524.

Hindmarch, I. (1980). Psychomotor function and psychoactive drugs. British

Journal of Clinical Pharmacology; 10: 189-209.

Hughes, J., & Hatsukami, D. K. (1998). Errors in using tobacco withdrawal scale.

Tobacco Control; 7: 92-93.

Hughes, J.R. (2007). Effects of abstinence from tobacco: Valid symptoms and

time course. Nicotine & Tobacco Research; 9: 315-327.

Hukkanen, J. Gourlay, S.G., Kerkare, S. & Benowitz, N.L. (2005). Influence of

menstrual cycle on cytochrome P450 2A6 activity and cardiovascular effects

of nicotine. Clinical Pharamcology & Therapeutics; 77: 159-169.

Husky, M.M., Mazure, C.M., Paliwal, P., McKee, S.A. (2008). Gender differences

in the comorbidity of smoking behavior and major depression. Drug & Alcohol

Dependence; 11: 176-179.

Jarvis, M. J., Tunstall-Pedoe, H., Feyerabend, C., Vesey, C., and Saloojee, Y.

(1987) Comparison of tests used to distinguish smokers from nonsmokers.

American Journal of Public Health 77, 1435-1445.

Jones, H.E., Garrett, B.E., & Griffiths, R.R. (1999). Subjective and psychological

effects of intravenous nicotine and cocaine in cigarette smoking cocaine

abusers. Journal of Pharmacology and Experimental Therapeutics; 288: 188-

197.

Klatzkin, R.R., Morrow, L.A., Light K.C., Pedersen, C.A., & Girdler, S.S. (2006).

Associations of histories of depression and PMDD with allopregnanolone

concentraions following the oral administration of micronized progesterone.

Psychoeuroendocrinology; 31: 1208-1219.

Knight, R. G., Williams, S., McGee, R., et al. (1997). Psychometric Properties of

the Centre for Epidemiologic Studies Depression Scale (CES-D) in a Sample

of Women in Middle Life. Behavior Research & Therapy 35, 373-380.

Kroenke, K., Spitzer, R.L., & Williams, J.B. (2001). The PHQ-9: Validity of a

brief depression severity measure. Journal of General Internal Medicine; 16:

606-613.

Lynch, W.J., & Sofuoglu, M. (2010). Role of Progesterone in Nicotine Addiction:

Evidence form initiation to relapse. Experimental and Clinical

Psychoparmacology; 18: 451-461

Marx, C.E., Trost, W.T., Shampine, L., Behm, F.M., et al. (2006). Neuroactive

steroids, negative affect and nicotine dependence severity in male smokers.

Psychopharmacology; 186: 462-472.

Mazure, C.M., Toll, B., McKee, S.A., Wu, R., & O’Malley, S.S. (2011). Menstrual

cycle phase at quit date and smoking abstinence at 6 weeks in an open-label

trial of bupropion. Drug & Alcohol Dependence; 114: 68-72.

Nappi, R.E., Petraglia, F., Luisi, S., Polatti, F., Farina, C., & Genazzani, A.R.

(2001). Serum allopregnanolone in women with postpartum “blues.”

Obstetrics & Gynecology; 97: 77-80.

Nyberg, S., Backstrom, T., Zingmark, E., Purdy, R., & Poromaa, I.S. (2007).

Allopregnanolone decreass with symptom improvement during placebo and

gonadotropin-releasing hormone agonist treatment in women with severe

premenstrual syndrome. Gynecological Endocrinology; 23: 257-266.

Oakes, J. Michael. (2012). "Statistical Power and Sample Size: Considerations

for Clinician-Researchers." in Essentials of Clinical Research, edited by

Stephen P Glasser. New York: Springer

Ottander, U., Sundstrom-Poromaa, I., Bjurulf, E., Skytt, A., Backstrom, T.,

Olofsson, J.I. (2005). Allopregnanolone and pregnanolone are produced by

the human corpus luteum. Molecular and Cellular Endocrinology; 239: 37-44.

Paul, S.M., & Purdy, R.H. (1992). Neuroactive steroids. The FASEB Journal; 6:

2311-2311.

Perkins, K.A., Broge, M., Gerlach, D., Sanders, M., Grobe, J.E., Cherry, C., &

Wilson, A.S. (2002). Acute nicotine reinforcement, but not chronic tolerance,

predicts withdrawal and relapse after quitting smoking. Health Psychology;

21: 332-339.

Perkins, K.A., & Scott, J. (2008). Sex differences in long-term smoking cessation

rates due to nicotine patch. Nicotine & Tobacco Research; 10: 1245-

1251.

Perkins, K.A. (2001). Smoking cessation in women: Special considerations.

CNS Drug; 15: 391-411.

Pillitteri, J.L., Kozlowski, L.T., Sweeney, C.T., & Heatherton, T.F. (1997).

Individual differences in the subjective effects of the first cigarette of the day:

A self-report method for studying tolerance. Experimental Clinical

Psychopharmacology; 5: 83-90.

Pomerleau, O.F., Pomerleau, C.S., & Namenek, R.J. (1998). Early experiences

with tobacco among women smokers, ex-smokers, and never-smokers.

Addiction; 93: 595-599.

Pomerleau, O.F. (1995). Individual differences in sensitivity to nicotine:

Implications for genetic research on nicotine dependence. Behavioral

Genetics: 25: 161-177.

Porcu, P. Sogliano, C., Cinus, M., Purdy, R.H., Biggio, G., & Concas, A. (2003).

Nicotine-induced changes in cerebrocortical neuroactive steroids and plasma

corticosterone concentrations in the rat. Pharmacology Biochemistry &

Behavior; 74: 683-690.

Pratt, L.A., & Brody, D.J. (2010). Depression and smoking in the U.S. household

population aged 20 and over, 2005-2008. NCHS Data Brief, 34. Hyattsville,

MD: National Center for Health Statistics.

http://www.cdc.gov/nchs/data/databriefs/db34.pdf

Prescott, E., Hippe, M., Schnohr, P., et al. (1998). Smoking and risk of

myocardial infarction in women and men: Longitudinal population study.

British Medical Journal; 316: 1043-1047.

Purdy, R.H., Moore, P.H., Rao, P.N., et al. (1990). Radioimmunoassay of 3-

alpha-hydroxy-5-alpha-pregnan-20-one in rat and human plasma. Steroids;

55: 290-296.

Reddy, D.S., & Kulkarni, S.K. (1997a). Chronic neurosteroid treatment prevents

the development of morphine tolerance and attenuates abstinence behavior

in mice. European Journal of Pharmacology; 337: 19-25.

Reddy, D.S., & Kulkarni, S.K. (1997b). Neurosteroid coadministration prevents

development of tolerance and augments recovery from benzodiazepine

withdrawal anxiety in hyperactivity mice. Experimental Clinical

Pharmacology; 19: 395-405.

Roberti, J., Harrington, L., Storch, E. (2006). Further Psychometric Support for

the 10-Item Version of the Perceived Stress Scale. Journal of University

Counseling; 9: 135-147.

Roselli, C.E., Fin, T.J., Ronnekleiv-Kelly, S.M., Tanchuck, M.A., Kaufman, K.R.,

Finn, Deboarh A. (2011). Localization of brain 5-alpha-reductase meseenger

RNA in mice selectively bred for high chronic alcohol withdrawal severity.

Alcohol; 45: 763-772.

Schiller, C.E., Saladin, M.E., Gray, K.M., Hartwell, K.J., & Carpenter, M.J. (2012).

Association between ovarian hormones and smoking behavior inw omen.

Experimental and Clinical Psychopharmacology; 20: 251-257.

Shiffman, S. (2008). How when matters for quitting and relapse. Addiction; 103:

822-823.

Sinnott, R.S., Mark, G.P., & Finn, D.A. (2002). Reinforcing effects of the

neurosteroid allopregnanolone in rats. Pharmacology, Biochemistry and

Behavior; 72: 923-929.

Stolerman, I.P., & Jarvis, M.J. (1995). The scientific case that nicotine is

addictive. Psychopharmacology; 117: 14-20.

Strohle, A., Pasini, A., Romeo, E., et al. (2000). Fluoxetine decreases

concentrations of 3 alpha, 5 alpha tetrahydrodeoxycortisoterone (THDOC) in

major depression. Journal of Psychiatry; 34: 181-186.

Sundstrom-Poromaa, I. (2008). Action of progesterone and progesterone

metabolites in menstrual-cycle-related disorders. Headache; 48: S90-S98.

Thorne, D.R., Genser, S.G., Sing, H.C., & Hegge, F.W. (1985). The Walter Reed

performance assessment batter. Neurobehavioral Toxicology & Teratology;

7: 415-418.

Tiffany, S. T., & Drobes, D. J. (1991). The development and initial validation of a

questionnaire on smoking urges. British Journal of Addiction; 86:1467-1476.

Timby, E., Balgard, M., Nyberg, S., Spigset, O., Andersson, A., et al. (2006).

Pharmacokinetic and behavioral effects of allopregnanolone in healthy

women. Psychopharmacology; 186: 414-424.

Torres, J.M., & Ortega, E. (2003). Alcohol intoxication increases

allopregnanolone levels in female adolescent humans.

Neuropsychopharmacology; 28: 1207-1209.

U.S. Department of Health and Human Services (USDHHS). (2001). Women and

Smoking: A report of the Surgeon General. Public Health Service, Office of

the Surgeon General, Rockville, Maryland.

http://www.cdc.gov/tobacco/data_statistics/

sgr/2001/complete_report/index.htm

Wang, M., Seippel, L., Purdy, R.H., & Backstrom, T. (1996). Relationship

between symptom severity and steroid variation in women with premenstrual

syndrome: Study on serum pregnenolone, prenenolone sulfate, 5 alpha-

pregnenlone-3,20-dione and 3 alpha-hydroxy-5 alpha-pregnan-20-one.

Journal of Clinical Endocrinology and Metabolism; 81: 1076-1082.

Watson, D., Clark, L.A., Carey, G. (1988). Positive and Negative Affectivity and

their Relation to Anxiety and Depressive Disorders. Journal of Abnormal

Psychology; 97: 346-353.

Wittchen, H. U., Robins, L. N., Cottler, L. B., Sartorius, N., Burke, J. D., and

Regier, D. (1991) Cross-cultural feasibility, reliability and sources of variance

of the Composite International Diagnostic Interview (CIDI). The multicenter

WHO/ADAMHA Field Trials. British Journal of Psychiatry; 159: 645-653.

Zang, E.A., & Wynder, E.L. (1996). Differences in lung cancer risks between

men and women: Examination of the evidence. Journal of National Cancer

Institute; 88: 183-192.

Zheng, P. (2009). Neuroactive steroid regulation of neurotransmitter release in

the CNS: Action, mechanism and possible significance. Progress in

Neurology; 89: 134-152.