Citation
Effects of Exercise and Contingency Management on Craving and Cigarette Smoking in the Human Laboratory

Material Information

Title:
Effects of Exercise and Contingency Management on Craving and Cigarette Smoking in the Human Laboratory
Creator:
Kurti, Allison N
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (1 p.)

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Psychology
Committee Chair:
DALLERY,JESSE
Committee Co-Chair:
ROWLAND,NEIL E
Committee Members:
VOLLMER,TIMOTHY RAYMOND
POMERANZ,JAMIE L
Graduation Date:
8/9/2014

Subjects

Subjects / Keywords:
Anticipation ( jstor )
Athletic motivation ( jstor )
Cigarette smoking ( jstor )
Cigarettes ( jstor )
Craving ( jstor )
Discounting ( jstor )
Exercise ( jstor )
Heart rate ( jstor )
Motivation ( jstor )
Psychopharmacology ( jstor )
Psychology -- Dissertations, Academic -- UF
cessation -- cigarette -- contingency -- exercise -- management -- smoking
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Psychology thesis, Ph.D.

Notes

Abstract:
Innovative smoking cessation interventions may be inspired by viewing cigarette smoking within a temporal discounting framework. Contextualized in terms of temporal discounting, smokers are viewed as devaluing the delayed consequences that are likely to result from abstaining (e.g., better health, saving money) relative to the immediate reinforcement derived from smoking. Treatment strategies implied by this framework include increasing the reinforcing value of abstinence, decreasing the reinforcing value of smoking, decreasing the extent to which smokers devalue delayed reinforcers, or some combination of the three. Experiment 1 investigated effects of exercise under controlled, laboratory conditions. Using a within-subjects design, twenty-one smokers engaged in one session of inactivity, low, and moderate-intensity exercise. Moderate-intensity exercise most effectively reduced self-reports of anticipated positive consequences of smoking, thus all participants exercised at a moderate intensity in Experiment 2. Using an ABAB within-subjects design, Experiment 2 (N = 20) evaluated whether the relationship between exercise and participants' latencies (in minutes) to ad libitum smoking were mediated through anticipated positive consequences of smoking and/or relief from withdrawal. Latencies to smoke were significantly longer after exercise (M = 21 min) versus inactivity (M = 4 min), and effects of exercise on latency were mediated through anticipated positive consequences of smoking. In Experiment 3, we assessed whether the combination of exercise (a manipulation to decrease the value of smoking) plus a laboratory model of contingency management (a manipulation to increase the value of abstaining) more effectively reduced smoking than either component on its own. CM increased latencies to smoke and decreased total puffs (Ms = 26.6 min and 2.2 puffs, respectively) relative to CM-control (non-contingent incentives (Ms = 13.8 min and 14.0 puffs). Exercise decreased anticipated positive consequences of smoking and relief from withdrawal, but did not influence objective measures of smoking or the extent to which participants devalued delayed reinforcers. Thus, the combination of exercise plus CM may not be more effective than CM alone. However, given that substantial individual differences in effects of exercise were observed across all three experiments, future research should identify those variables that predict responsiveness to exercise interventions to promote smoking cessation. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (Ph.D.)--University of Florida, 2014.
Local:
Adviser: DALLERY,JESSE.
Local:
Co-adviser: ROWLAND,NEIL E.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2015-08-31
Statement of Responsibility:
by Allison N Kurti.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Embargo Date:
8/31/2015
Classification:
LD1780 2014 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

I used APA style



PAGE 1

EFFECTS OF EXERCISE AND CONTINGENCY MANAGEMENT ON CRAVING AND CIGARETTE SMOKING IN THE HUMAN LABORATORY By ALLISON N. KURTI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLME NT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2014

PAGE 2

© 2014 Allison N. Kurti

PAGE 3

To my family

PAGE 4

4 ACKNOWLEDGMENTS Numerous research assistants have helped with participant recruitment, data collection and data entry for the three studies that comprise my dissertation. These individuals include Bethany Mondor, Connie Farmer, Nicole Snyder, Jessica Brown, Tanvi Pendharkar, Jessica Cowan, and Lesleigh Craddock. Further, I than k my lab mates who have provided valuable input about this research over the past few years. These individuals include Steven Meredith, Philip Erb, Alana Rojewski, Brantley Jarvis, Rachel Cassidy, and Vanessa Minervini. Financial support was provided by th e University of Florida (Dallery overhead account), as well as through the College of Liberal Arts and Sciences Dissertation Research Award. I also thank parents, David and Wendy Kurti, along with my sister, Stephanie Kurti, for their support and encourage ment of my academic goals. Similarly, I thank those friends who have provided emotional support and levity all throughout graduate school. These individuals include Katherine Pankow, Brian Johnson, Rachel Parker, Natalie Hadad, Kati Connelly, and Tia Boliv ar. I thank my previous advisors, Robert Beck and Matthew Matell, for challenging me to be a better researcher and preparing me for graduate school. I also thank my graduate advisor, Jesse Dallery, for his constant guidance, support, and inspiration to bec ome a better scientist. I would also like to thank all of the faculty and students in the Behavior Analysis program at the University of Florida for contributing in numerous ways to my training. Finally, I thank my participants, without whom my research wo uld have been impossible.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 2 EXPERIMENT 1 ................................ ................................ ................................ ..... 20 Rationale ................................ ................................ ................................ ................. 20 Method ................................ ................................ ................................ .................... 24 Participants ................................ ................................ ................................ ....... 24 Screening ................................ ................................ ................................ ......... 24 Apparatus and Materials ................................ ................................ ................... 25 Experimental Design ................................ ................................ ........................ 25 Procedure ................................ ................................ ................................ ......... 26 Data Analysis ................................ ................................ ................................ ... 29 Results ................................ ................................ ................................ .................... 30 Manipulation Checks ................................ ................................ ........................ 30 Effects of Exercise Intensity on Smoking Motivation ................................ ........ 31 Reliability of Exercise Effects on Smoking Motivation ................................ ...... 32 Discussion ................................ ................................ ................................ .............. 33 3 EXPERIMENT 2 ................................ ................................ ................................ ..... 42 Rationale ................................ ................................ ................................ ................. 42 Method ................................ ................................ ................................ .................... 43 Participants ................................ ................................ ................................ ....... 43 Apparatus and Materials ................................ ................................ ................... 44 Experimental Design ................................ ................................ ........................ 44 Procedure ................................ ................................ ................................ ......... 44 Data Analysis ................................ ................................ ................................ ... 46 Results ................................ ................................ ................................ .................... 48 Manipulation Checks ................................ ................................ ........................ 48 Smoking Motivation ................................ ................................ .......................... 48 Latency to Smoke ................................ ................................ ............................. 49 Mediation Analysis ................................ ................................ ........................... 49 Discussion ................................ ................................ ................................ .............. 50

PAGE 6

6 4 EXPERIMENT 3 ................................ ................................ ................................ ..... 59 Rationa le ................................ ................................ ................................ ................. 59 Method ................................ ................................ ................................ .................... 62 Participants ................................ ................................ ................................ ....... 62 Apparatus and Materials ................................ ................................ ................... 62 Experimental Design ................................ ................................ ........................ 63 Procedure ................................ ................................ ................................ ......... 63 Data Analysis ................................ ................................ ................................ ... 66 Objective measures of smoking ................................ ................................ . 67 Subjective measures of smoking motivation ................................ .............. 67 Monetary di scounting task ................................ ................................ ......... 67 Interobserver agreement (IOA) ................................ ................................ ......... 69 Results ................................ ................................ ................................ .................... 71 Manip ulation Checks ................................ ................................ ........................ 71 Objective Measures of Smoking ................................ ................................ ....... 71 Latency to smoke ................................ ................................ ....................... 71 Puff number ................................ ................................ ............................... 72 Puff volume ................................ ................................ ................................ 73 Subjective Measures of Smoking Motivation ................................ .................... 73 Monetary Discounting Task ................................ ................................ .............. 74 Interobserver Agreement (IOA) ................................ ................................ ........ 75 Correlates of Treatment Effects ................................ ................................ ........ 76 Discussion ................................ ................................ ................................ .............. 77 5 GENERAL DISCUSSION ................................ ................................ ....................... 94 APPENDIX: HYPOTHETICAL TEMPORAL DISCOUNTING DATA ........................... 105 LIST OF REFERENCES ................................ ................................ ............................. 109 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 122

PAGE 7

7 LIST OF TABLES Table page 2 1 Smoking Motivation Scores across Time (Experiment 1) ................................ ... 37 2 2 Smoking Motivation Scores during Replication Sessions (Experiment 1) ........... 38 4 1 Median scores on Indices of Temporal Discounting ................................ ........... 84 A 1 Hypothetical Temporal Discounting Data ................................ ......................... 106

PAGE 8

8 LIST OF FIGURES Figure page 1 1 Hypothetical discounting curves for individuals with different k values ............... 18 1 2 Hypo thetical discounting curves for smoking and abstinence ............................ 19 2 1 Average change in QSU scores at successive measurement times across different exercise intensities ................................ ................................ ............... 39 2 2 Indivi dual participant changes in QSU scores after treatment ............................ 40 2 3 Correspondence between QSU scores relative to baseline between initial sessions versus replication sessions ................................ ................................ .. 41 3 1 Hypothesized mediation model for effects of exercise on latency to ad libitum smoking. ................................ ................................ ................................ ............. 54 3 2 QSU scor es before and after exercise/inactivity ................................ ................. 55 3 3 Mean latencies to ad libitum smoking ................................ ................................ . 56 3 4 Individual participant latencies to ad libitum smoking ................................ ......... 5 7 3 5 Regression coefficients and significance levels for mediation analyses ............. 58 4 1 Mean latencies to smoke as a function of sesion type ................................ ........ 85 4 2 Individual participant latencies to smoke ................................ ............................ 86 4 3 Mean number of puffs per block across conditions ................................ ............. 87 4 4 Individual particip ................................ .................. 88 4 5 across conditions ................................ ...... 89 4 6 QSU scores before and after exercise/inactivity ................................ ................. 90 4 7 ty ... 91 4 8 Box plots for AUC values as a function of condition ................................ ........... 92 4 9 values as a function of conditio n ............................ 93 A 1 Calculation of k values from hypothetical discounting data .............................. 107 A 2 Calculation of AUC values from hypothetical discountin g data ......................... 108

PAGE 9

9 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EFFECTS OF EXERCISE AND CON TINGENCY MANAGEMENT ON CRAVING AND CIGARETTE SMOKING IN THE HUMAN LABORATORY By Allison N. Kurti August 2014 Chair: Jesse Dallery Major: Psychology Innovative smoking cessation interventions may be inspired by viewing cigarette smoking within a tempora l discounting framework. Contextualized in terms of temporal discounting, smokers are viewed as devaluing the delayed consequences that are likely to result from abstaining (e.g., better health, saving money) relative to the immediate reinforcement derived from smoking. Treatment strategies implied by this framework include increasing the reinforcing value of abstinence, decreasing the reinforcing value of smoking, decreasing the extent to which smokers devalue delayed reinforcers, or some combination of th e three. Experiment 1 investigated effects of exercise under controlled, laboratory conditions. Using a within subjects design, twenty one smokers engaged in one session of inactivity, low, and moderate intensity exercise. Moderate intensity exercise most effectively reduced self reports of anticipated positive consequences of smoking, thus all participants exercised at a moderate intensity in Experiment 2. Using an ABAB within subjects design, Experiment 2 (N = 20) evaluated whether the relationship betwee libitum smoking were mediated through anticipated positive consequences of smoking and/or relief from withdrawal. Latencies to smoke were significantly longer after exercise

PAGE 10

10 ( M = 21 min) versus inac tivity ( M = 4 min), and effects of exercise on latency were mediated through anticipated positive consequences of smoking. In Experiment 3, we assessed whether the combination of exercise (a manipulation to decrease the value of smoking) plus a laboratory model of contingency management (a manipulation to increase the value of abstaining) more effectively reduced smoking than either component on its own. CM increased latencies to smoke and decreased total puffs ( Ms = 26.6 min and 2.2 puffs, respectively) re lative to CM control (non contingent incentives ( Ms = 13.8 min and 14.0 puffs). Exercise decreased anticipated positive consequences of smoking and relief from withdrawal, but did not influence objective measures of smoking or the extent to which participa nts devalued delayed reinforcers. Thus, the combination of exercise plus CM may not be more effective than CM alone. However, given that substantial individual differences in effects of exercise were observed across all three experiments, future research s hould identify those variables that predict responsiveness to exercise interventions to promote smoking cessation.

PAGE 11

11 CHAPTER 1 INTRODUCTION Despite the abundance of smoking cessation treatments, approximately 19% of the current US population smokes (Cent ers for Disease Control and Prevention (CDC), 2012), and cigarette smoking remains the largest preventable risk factor for morbidity and mortality. Among current smokers, 70% report a desire to quit, and 52% attempt to quit each year (CDC, 2011). Nonethele ss, the national prevalence of cigarette smoking has stabilized over the past seven years (CDC, 2012). For these remaining smokers, new approaches to promoting cessation are needed (e.g., devising novel cessation interventions, enhancing the efficacy of ex isting cessation interventions). These interventions may be inspired by contextualizing cigarette smoking in terms of temporal discounting. Temporal discounting refers to the loss in subjective value of a commodity as the delay to receipt of that commodity increases (Ainslie, 1974; Green, 1982; Mazur, 1987). With respect to smoking, the delayed consequences that are likely to result from abstaining (e.g., better health, saving money) are devalued relative to the immediate reinforcers derived from smoking. T he extent to which smokers devalue the likely, delayed consequences associated with abstaining can be quantified using a parameter k , which indexes the rate at which reinforcers decline in value as a function of the delay until their receipt. k is estimate d using the equation (Mazur, 1987): (1 1) In this equation, V refers to the discounted value of the delayed reinforcer, A refers to the amount or magnitude of the reinforcer, and D refers to the delay. K is a free parameter that is estimated using least squares regression. Higher s reflect steeper

PAGE 12

12 discounting or greater impulsivity (i.e., reinforcers quickly lose value as the delay until their receipt increases), whereas lower s reflect shallower discounting. Figure 1 1 shows hypothetical discountin g curves for individuals with different k values. For individuals with higher k values (i.e., more impulsive individuals), reinforcers decline more sharply in value as a function of the delay until their receipt. Thus, if the individuals in Figure 1 1 were smokers, the less impulsive smoker (left) would resist smoking until he or she was face to face with smoking stimuli, at which point behavior would be allocated to smoking rather than alternative, non smoking activities. In contrast, the more impulsive sm oker (right) would allocate behavior to smoking much earlier. The point at which the smoker begins allocating behavior to smoking rather than alternative, non smoking activities is represented by the delay at which the curves cross one another. Users of ci garettes, alcohol, marijuana, cocaine, and heroin have been shown to discount the delayed value of both money and their drug of choice more steeply than non users (Baker, Bickel, & Johnson, 2003; Critchfield & Kollins, 2001; Kirby, Petry, & Bickel, 1999; M adden & Bickel, 2009; Rosenthal, Edwards, Ackerman, Knott, & Rosenthal, 1990). These data have been interpreted as supporting the validity of temporal discounting as a framework for quantifying impulsivity. One advantage of conceptualizing drug use in ter ms of temporal discounting is that strategies to minimize impulsive choice (e.g., cigarette smoking) are directly implied by this framework. In a choice framework, one chooses between an immediate versus a delayed alternative, thus Equation 1 1 can be appl ied to the value of each alternative (Figure 1 2). Choice between the two alternatives can be described by the equation:

PAGE 13

13 (1 2) where the B subscripts s and a represent choice f or smoking and abstinence (i.e., non smoking activities considered as an aggregate), respectively. The value, V , of each alternative is determined by Equation 1 1. One strategy to reduce impulsive choice for cigarettes, therefore, is to diminish the value of the immediate, cigarette reinforcer by decreasing A in Equation 1 1. Hypothetically, this could be achieved by taking varenicline (Chantix ® ) or smoking denicotinized cigarettes, both of which may diminish the reinforcement that is typically derived fro discounting ( k ). Figure 1 1 suggests that the less impulsive smoker (i.e., the smoker with the lower k ) must be face to face with smoking stimuli before he or she chooses to smoke. I f his or her k was further decreased, this individual would choose to abstain even when directly confronted with smoking stimuli (i.e., the two curves would not cross one another). One final strategy to limit impulsive choice is to decrease the delay until the receipt of the likely delayed reinforcers for abstinence ( D in Equation 1 1). This could be achieved by delivering positive consequences contingent on drug abstinence. The notion that decreasing the delay to delayed reinforcement can promote abstinenc e underlies the effectiveness of contingency management (CM) for treating drug addiction. CM allows individuals to earn motivational incentives (e.g., vouchers exchangeable for goods and services), contingent on meeting objectively verifiable goals (e.g., smoking abstinence; Higgins, Alessi, & Dantona, 2002; JABA , 2008; Lussier, Heil, Mongeon, Badger, & Higgins, 2006). CM has been shown to promote

PAGE 14

14 abstinence in both laboratory (Dallery & Raiff, 2007; Higgins, Heil, & Lussier, 2004; Packer, Howell, McPherson , & Roll, 2012) and naturalistic settings ( Dallery, Glenn, & Raiff, 2007; Roll, Higgins, & Badger, 1996; Sigmon, Lamb, & Dallery, 2008; Stitzer & Bigelow, 1984 ). In lab models of CM, monetary reinforcement is contingent upon choosing small sums of money r ather than puffs on a cigarette, and in naturalistic settings, monetary reinforcement is contingent upon demonstrating predetermined reductions in breath carbon monoxide (CO) levels. Although CM attempts to increase the value of abstinence (i.e., CM decrea ses D in Equation 1 1 by providing alternative commodities with relatively short D value of smoking (i.e., CM might not decrease A ). Thus, for some smokers, the delayed reinforcers likely to resul t from abstaining continue to be devalued relative to the immediate, reinforcing value of smoking. Consequently, these individuals do not achieve drug free periods during CM (Dallery & Raiff, 2007; Iguchi et al., 1996; Roll et al., 1996; Silverman et al., 1996). discounting rates would be an ideal addition to CM, which attempts to increase the value of abstinence. One promising approach to achieving this is physical exercise. A single, brief bout of exercise has been shown to diminish self reported craving (Daniel, Cropley, & Fife Schaw, 2007; Everson, Daley, & Ussher, 2008; Scerbo, Faulkner, Taylor, & Thomas, 2010; Taylor, Katomeri, & Ussher, 2005, 2006; Ussher, Cropley, Playle, Mohidin, & West, 2009) and withdrawal (Daniel et al., 2007; Everson et al., 2008; Ussher et al. 2009), and to increase mood (Arbour Nicitopoulos, Faulker, Hsin, &

PAGE 15

15 Selby, 2011; Elibero, Janse Van Rensburg, & Drobes, 2011; Everson et al., 2008; Thayer, Taka hashki, & Birkhead Flight, 1993; Williams et al., 2011). Exercise has also been shown to increase the delay to ad libitum smoking relative to non physical activities (Katomeri, 2009; Reeser, 1983; see Roberts, Maddison, Simpson, Bullen, & Prapavessis , 2012 for a review; Taylor & Katomeri, 2007; Thayer et al., 1993). Further, exercise decreases brain activity in areas associated with drug motivation and reward in the presence of smoking cues (Janse Van Rensburg, Taylor, Hodgson, & Benattayallah, 2009, 2011), which further suggests that exercise might decrease the reinforcing value exercise), and increasing the value of abst inence (via CM), the combination of exercise plus CM may reduce smoking more effectively than either approach on its own. Thus, the overarching purpose of this dissertation was to evaluate whether a combined manipulation to simultaneously decrease the valu k values, and increase the value of abstaining (i.e., exercise plus CM, respectively), more effectively reduced smoking than either exercise or CM alone. En route to evaluating this combined approach, three Experiments were con ducted: Experiment 1 was conducted to determine the exercise intensity that most effectively decreased self reports of the anticipated positive and negative reinforcement associated with smoking (i.e., A ). Moderate intensity exercise more effectively decre ased self reports of the anticipated positive consequences associated with smoking relative to low intensity exercise or inactivity. However, whether effects of exercise on self reported smoking motivation translated into actual reductions in

PAGE 16

16 smoking was n ot assessed. Thus, Experiment 2 was conducted to determine effects of moderate intensity exercise on both self reported smoking motivation and objective indices of smoking (e.g., latency to ad libitum smoking). Experiment 2 demonstrated that exercise incr smoke relative to inactivity, and mediation analyses indicated that this effect was mediated through anticipated positive consequences of smoking. Aside from smoking motivation, however, Experiment 2 did not evaluate other behavioral mechanisms through which exercise may decrease smoking (e.g., temporal discounting rates). In addition, results of the experiment revealed individual differences in effects of exercise on both positive and negative reinforcement associated with smoking, as well as latency to smoke. Interpreted within a temporal discounting framework, the results of Experiment 2 suggested that among participants for whom exercise had minimal influence on smoking related variables, the value of the immediate conse quences of smoking continued to outweigh the value of the delayed consequences likely to result from abstaining. Thus, it seemed that a treatment to increase the value of abstinence would be a useful adjunct to exercise for individuals whose smoking was no t influenced by exercise alone. Consequently, Experiment 3 was conducted to evaluate whether a combined manipulation to simultaneously decrease the value of smoking and/or (CM) would more effectively reduce smoking than either component on its own. The three experiments that comprise this dissertation were conducted in a laboratory as opposed to a naturalistic setting for several reasons. For example,

PAGE 17

17 laboratory settings pe rmit an assessment of smoking related variables (e.g., effects of deprivation on smoking motivation) on a shortened time course. Laboratory studies of smoking are also efficient, cost and time effective, likely to reduce attrition, and they allow researche rs to conduct parametric manipulations and evaluate hypothesized mechanisms that underlie a phenomenon under controlled conditions (Chow, 1995; McKee, 2009; Kamboj et al., 2012; Shadel et al., 2011; Sweitzer, Delinger, & Donny, 2012). Importantly, laborato ry based smoking models have replicated the results of full scale clinical trials in terms of identifying predictors of smoking abstinence (e.g., smoking reinforcement, Perkins, Stitzer, & Lerman, 2006; MacKillop et al., 2008; Murphy, MacKillop, Tidey, Bra zil, & Colby, 2011; latency to smoke after a deprivation period, Shadel et al., 2011; see Sweitzer et al., 2012, for a review), and approaches to outcomes in naturalistic settings (Haney & Spealman, 2008). By systematically investigating the effectiveness of a combined approach to reducing smoking in the laboratory, this dissertation represents a first step towards a continuum of research intended to inspire further develop ment and refinement of innovative, behavioral

PAGE 18

18 Figure 1 1. Hypothetical discounting curves for individuals with lower (left) versus higher (right) k values. The slopes indicate the rate at which reinforcers are devalued (discounted) as a function of delay. The shorter vertical bar in each panel represents the immediate alternative (e.g., smoking) and the taller bar represents the delayed alternative (e.g., abstinence). The point at which the curves cross represents the delay at which the individual begins allocating behavior to the immediate alternative as opposed to the delayed alternative. This crossover point occurs much earlier for the individual with the higher k value. Delay (days)

PAGE 19

19 Figure 1 2. Hypothetical discounting curves for smoking and abstinence (i.e., non smoking activities considered as an aggregate). The discounting equation applies to both alternatives, and the value ( V ) of each alternative determines the amount of respond ing ( B ) allocated to that alternative (where V a and V s refer to the value of abstinence versus smoking, respectively).

PAGE 20

20 CHAPTER 2 EXPERIMENT 1 Rationale As explained previously, smokers choose a smaller, immediate reinforcer (e.g., a euphoric nicotine buzz) at the expense of larger, delayed consequences that are likely to result from abstaining (e.g., better health, saving money). In the context of the temporal discounting equation, the reinforcing value of smoking is reflected by A (i.e., the amount o r magnitude of the reinforcer). Therefore, decreasing A is one strategy to minimize smoking. For an individual smoker, the value of smoking ( A ) at any given moment is dynamic (MacKillop, Menges, McGeary, & Lisman, 2007), and is presumably determined by bot h positive reinforcement (e.g., pleasurable consequences of smoking) and negative reinforcement (e.g., relief from withdrawal symptoms) that have historically accompanied smoking. By decreasing both the anticipated, positive consequences of smoking and rel ief from withdrawal, exercise might represent a manipulation that decreases A. Previous research has shown that a brief bout of exercise decreases cigarette craving relative to non physical activities (Bock, Marcus, King, Borrelli, & Roberts, 1999; Eliber o et al., 2011; Taylor et al., 2005, 2006). However, most prior research has approached craving as a one dimensional construct and employed single item measures of craving (see Roberts et al., 2012, for a review), as opposed to a conceptualization of cravi ng based on positive and negative reinforcement aspects of smoking motivation. In fact, there is growing support for this latter, multidimensional conceptualization of craving (Davies, Wilner, & Morgan, 2000; Eissenberg, Adams, Riggins, & Likness, 1999; To ll, Katulak, & McKee, 2006). In addition, emerging

PAGE 21

21 laboratory studies characterize craving not as a variable that independently produces contributes to impulsive choice alon g with other processes such as delay discounting (Loewenstein, 1996; MacKillop et al., 2007; MacKillop et al., 2012; Perkins, Grobe, & Fonte, 1997; Sayette, Martin, Wertz, Shiffman, & Perrott, 2001). conceptualized as a physiological state resulting from smoking deprivation, in which one tacts (i.e., privately identifies) the positive and negative reinforcement that have historically resulted from operation (EO) that increases the value of the anticipated positive and negative dent manipulations (e.g., smoking deprivation) that (1) increase the value of a particular consequence as a reinforcer (e.g., increase the positive and negative reinforcement associated with smoking), and (2) increase the behavior emitted to earn that rein forcer (e.g., procuring a pack of cigarettes, lighting a cigarette, Laraway, Snycerski, Michael, & Poling, 2003; Michael, 1993). Given the presumed effect of smoking deprivation on the value of smoking (i.e., smoking deprivation presumably increases A in E quation 1 1), it is important to assess the capacity for exercise to decrease A using a valid conceptualization of craving. To this end, a multidimensional craving measure reflecting both positive and negative reinforcement aspects of smoking motivation (i .e., the Questionnaire of Smoking Urges Brief; Cox, Tiffany, & Christen, 2001) was administered in Experiment 1.

PAGE 22

22 In addition to the QSU being broadly consistent with a behavior analytic account of craving (i.e., it addresses the private tacts one may emit in a nicotine deprived state), the validity of the QSU is supported by research conducted in both laboratory (e.g., Willner, Hardman, & Eaton, 1995) and naturalistic settings (e.g., Allen, Bade, Hatsukami, & Center, 2008; Davies, Willner, & Morgan, 2000). In the laboratory, Willner et al. (1995) evaluated the extent to which deprivation periods of differing lengths modulated the relationship between QSU scores and the number of progressive ratio responses that participants made to earn access to one cigaret te puff. After a 4 hour smoking deprivation period, craving based on anticipated relief from withdrawal symptoms (i.e., negative reinforcement items) better predicted progressive ratio responding than craving based on anticipated positive consequences of s moking (i.e., positive reinforcement items). In contrast, positive reinforcement items better predicted progressive ratio responding when smokers were not nicotine deprived. These data were taken to support the construct validity of the two QSU subscales. Differential response patterns on the positive and negative reinforcement subscales of the QSU have also been observed among smokers who exhibit different (i.e., in dividuals who smoke less than five cigarettes per day) reported higher craving based on anticipated positive consequences of smoking relative to craving based on anticipated relief of withdrawal symptoms. In addition, participants who had not smoked for se veral hours provided higher ratings on both dimensions of craving relative to participants who had not been nicotine deprived. In another study, Allen et al. (2008) showed that, among smokers undergoing a quit attempt, those who relapsed within the

PAGE 23

23 first 3 0 days provided higher ratings on the negative reinforcement items at five administration times preceding their relapse relative to those who abstained despite the absence of differences in baseline craving between smokers versus abstainers. Thus, despite the availability of many other methods to assess craving (e.g., single item measures, free response procedures, psychophysiological procedures, see Rosenberg, 2009, for a review), a multidimensional self report measure was selected for use in Experiment 1 because (1) it is broadly consistent with a behavior analytic interpretation of craving, and (2) research supports the use of the QSU in both laboratory and naturalistic settings. Although the majority of research on exercise intensity and smoking supports the use of moderate intensity exercise for reducing craving and tobacco related withdrawal symptoms (Daniel, Cropley, Ussher, & West, 2004; Elibero et al., 2011), other research suggests that some smokers are unwilling or unable to substantially elevate t heir heart rates (Pomerleau et al., 1987). Exercising at higher intensities has also been shown to reported well being and increase psychological distress (Everson, Daley, & Ussher, 2008). Few researchers have evaluated whether there are individual differences in the exercise intensity that most effectively reduces craving, but it is feasible that the efficacy of low versus moderate intensity exercise at reducing craving differs as a function of various antecedent variables (e.g., nic otine dependence level, years smoking). Thus, the purpose of Experiment 1 was to assess effects of low intensity exercise, moderate intensity exercise, and inactivity on positive and negative reinforcement aspects of smoking motivation.

PAGE 24

24 Method Participants Twenty one smokers (14M, 7F) were recruited via classified advertisements and flyers posted throughout the community. The eligibility criteria were (a) 18 60 years of age, (b) self reported smoking of > 10 cigarettes/day, (c) self reported desire to qui t smoking (to enhance the clinical validity of our short term, laboratory study, see Perkins, Stitzer, & Lerman 2006), (d) drug free urine (with the exception of marijuana), (e) blood alcohol concentration (BAC) of 0.0%, and (f) breath carbon monoxide (CO) sample of > 10 parts per million (ppm). Exclusionary criteria were (a) current drug abuse/drug dependence (excluding nicotine and caffeine), (b) medication use that would interfere ric illness (e.g., taking psychotropic medication) within the past 6 months, (d) evidence of any condition that might contraindicate physical activity (e.g., exercise induced asthma), (e) Questionnaire (PAR Q, see Measures), (f) classifying as high risk for cardiovascular disease according to the Health Status Questionnaire (HSQ, see Measures), and (g) being pregnant or lactating. Participants were compensated $40.00 after completing all f our sessions, in the form of a check that was mailed to their homes. Screening Applicants who responded to advertisements and flyers underwent a brief telephone screening. They were asked (a) their age, (b) how many cigarettes per day they smoked, (c) whet her they had any health problems that made exercise dangerous for them, and (d) whether they wanted to quit smoking. Applicants who met the inclusion criteria (see Participants) were invited into the lab for a 30 minute in person screening

PAGE 25

25 During the scree ning, participants first signed an informed consent. They then provided a urine sample for drug testing (CupOne Kits; Varian; Lake Forest, CA) and pregnancy testing (Calhoun Industries; Fort Smith, AZ). Participants were also tested for evidence of recent drinking (measured via hand held breathalyzer; Alco Sensor IV, Intoximeters, Inc.; Saint Louis, MO), recent smoking (measured via hand held monitor; Bedfont Scientific Ltd.; Kent, England), and their resting heart rates were measured. Participants also com pleted several questionnaires, including a psychosocial history, the Physical Activity Readiness Questionnaire (PAR Q; Hafen & Hoeger, 1994), the Health Status Questionnaire (HSQ; Radosevich, Wetzler, & Wilson, 1994), and the Fagerstrom Test for Nicotine D ependence (FTND; Heatherton et al., 1991). The psychosocial history contains questions related to demographics, general health, and medication use. The PAR Q identifies individuals for whom low to moderate intensity exercise is not recommended. Participant s who reported experiencing chest pain, dizziness, etc., during exercise, were excluded. The HSQ identifies participants as being at a low, medium, or high risk for cardiovascular disease. All high risk participants were excluded. Apparatus and Materials Experimental sessions took place in small, windowless, well ventilated smoking rooms. The rooms were equipped with a chair, PC with monitor, television with a DVD player, health laptop computer with internet access. Experimental Design A within subjects design was employed, in which participants returned to the

PAGE 26

26 low intensity exercise, and moderate intensity exercise (with session order randomized across participants). On their fourth visit, one of these sessions was randomly selected to be replicated. Although replicating all three sessions would have been preferable, the present method was selected to address issues concerning reliability while simultaneously minimizing participant attrition. Procedure in a minimum of one session per week and one day between sessions. The time of sessions was kept as consistent as possible within participants. At the beginning of all sessions, participants smoked half of one cigarette of their preferred brand (supplied by research staff) in order to equate smoke exposure across sessions, and to equate smoke exposure between participants as much as possible. To ensure that participants smoked the instructed amount, cigarettes were measured, and a line was drawn at the ha lfway point (measured from filter to tip). No participants smoked beyond this line. Participants then underwent a 1 hour, no smoking period, during which they had access to a television with a DVD player, laptops with Internet access, and health related ma indicates that self reported craving and withdrawal symptoms can occur within 30 minutes of not smoking (Hendricks, Ditre, Drobes, & Brandon, 2006), and that 1 hour corresponds to the onset of craving in sm 2011). After the 1 hour deprivation period, participants completed a brief version of the Questionnaire of Smoking Urges (QSU; Cox et al., 2001). This 10 item measure is

PAGE 27

27 composed of two factors, with 5 it ems reflecting positive reinforcement aspects of smoking motivation (e.g., a cigarette would taste good now), and the other 5 items reflecting negative reinforcement aspects of smoking motivation (e.g., smoking would make me less depressed). This measure h as good reliability and validity (Cappelleri et al., 2007; Toll, Katulak, & McKee, 2006), and has been used in several previous, laboratory based exercise studies (e.g., Janse Van Rensburg & Taylor, 2008; Janse Van Rensburg, Elibero, Kilpatrick, & Drobes, 2013). We adapted the QSU from its original Likert style format into a 100 point VAS to permit a more fine grained analysis (0 = not at all , 100 = extremely , McKee, Weinberger, Shi, Tetrault, & Coppola, 2012; Perkins et al., 2008). Research suggests that i n most cases, converting Likert scales to visual analog scales (or vice versa) does not diminish the psychometric properties of the original measure (e.g., Borjeson et al., 1997; Jaeschke, Singer, & Guyatt, 1990; van Laerhoven & van der Zaag Loonen, 2004). After completing the QSU, participants underwent 20 minutes of low intensity exercise, moderate intensity exercise, or inactivity (control). Exercise consisted of riding a stationary bike (Schwinn 140 Upright Exercise Bike), and moderate intensity was de fined as 40 Sports Medicine, 2010). However, based on research suggesting that smokers may be unwilling and/or unable to elevate their heart rates substantially (e.g., Pomerleau et al., 19 87), participants were instructed to exercise at 60% HRR to increase the likelihood intensity range. Low intensity exercise was defined as 20 40% of HRR. The upper range of intensity exercise,

PAGE 28

28 lculated using the Karvonen person screening) were subtracted from their maximum heart rates (i.e., 220 age). This number was then used to cal HR. For example, the following calculation was used to determine the target HR for moderate intensity exercise sessions: (.60 X HRR) + RHR. Participants began the exercise session with a two minute warm up, during which they inc reased their heart rate to 40 60% (or 20 40%) of HRR. This target HR was maintained for 17 minutes, and participants received 1 minute to cool down. Researchers were present during exercise sessions to ensure that participants maintained a HR within their target HR range for the entire 17 minutes. During control sessions, participants sat quietly for 20 minutes with access to health related magazines, a TV, and a laptop with internet access. Participant heart rates were also monitored during control session s. Following exercise/inactivity, participants completed the QSU for a second time, and at 10 minutes, 20 minutes, 30 minutes, 40 minutes, 50 minutes, and 60 minutes thereafter (a computer generated tone indicated when to complete successive assessments). Participants had access to magazines, TV, etc., during this time. After completing the QSU at the 60 minute time point, participants left the lab. Regardless of exercise intensity, all sessions proceeded according to the schematic below: Smoke 1 hr Q SU Brief (2 3 min) Exercise/Control (20 min) QSU Brief at 10 min intervals (1 hr)

PAGE 29

29 Data Analysis Statistical analyses were conducted using SPSS (version 21). The outcome variables were positive and negative reinforcement components of smoking motivatio n (QSU positive and QSU negative, respectively). Because the homogeneity of variance assumption did not hold and neither variable was normally distributed, Friedman tests were conducted to detect differences in pre post QSU positive/QSU negative change sco res (i.e., difference between baseline smoking motivation and smoking motivation at successive time points) (significance criterion = p < .05), and post hoc, Wilcoxon signed rank follow up tests were conducted to determine which sets of change scores diffe red from one another. Because preliminary data analyses using Mann Whitney U tests (e.g., for gender) and Kruskal Wallis tests (e.g., session number) indicated that outcomes did not differ based on these variables, they were not used as covariates in subse quent analyses (Janse Van Rensburg et al., 2013). In addition to the group based analyses described above, visual analysis of individual participant data was employed to (1) evaluate the extent to which exercise effects on QSU scores observed at the group level were replicated within individual participants, and (2) detect individual differences in the exercise intensity that most effectively reduced QSU scores. Recall that each participant had a replication of one of the three exercise intensity sessions (e.g., inactivity, low, or moderate intensity exercise). Data from the replication session were excluded from the analyses described above. However, to evaluate whether exercise had replicable effects on smoking motivation, Wilcoxon signed rank tests were conducted to assess differences between pre post QSU -

PAGE 30

30 positive/QSU replication session. For example, for smokers who experienced a replication of the moderate intensity exercise session , these analyses were conducted to evaluate whether the magnitude of exercise effects on QSU scores differed between the initial session and the replication session. Significant effects would indicate that the magnitude of exercise effects differed between the two sessions. In contrast, non significant effects would indicate that the magnitude of exercise effects was comparable on the initial session and the replication session (i.e., that exercise had reliable effects on self reported smoking motivation). In addition to the statistical analysis described above, visual analysis was employed to evaluate whether there were any visually detectable differences in the session and his or her replication session. Results Participants were 21 smokers (14M, 7F). They had a mean age of 35.5 years ( SD = 13.2), had been smoking for 16.1 years ( SD = 11.8), and smoked 18.6 cigarettes per day ( SD = 7.4). Twelve participants were Caucasian, s even were Black, one was Hispanic, and one reported more than one race. Participants reported a mean of 3.6 previous attempts to quit smoking ( SD = 3.0), and had a mean FTND score of 5.5 ( SD = 2.5). Manipulation Checks A one way analysis of variance (ANO VA) was conducted to validate the significantly as a function of condition, F (2,58) = 106.0, p < .001. Post hoc, Bonferroni -

PAGE 31

31 corrected follow r during moderate intensity exercise ( M = 132.1 bpm, SD = 13.6) relative to low intensity ( M = 109.3 bpm, SD = 8.2) and inactivity ( M = 76.6 bpm, SD = 13.7), and during low intensity versus inactivity ( ps < .001). Seventeen participants (71.4%) achieved th e moderate intensity criterion. Of the end of the range (60% HRR) by 1 and 6 bpm. N ineteen participants (95.2%) achieved the low intensity criterion. The participant who did not achieve this criterion exhibited a HR that fell 1 bpm above the higher end (< 40%) of this range. Effects of Exercise Intensity on Smoking Motivation Effec ts of exercise intensity on QSU positive/QSU negative relative to baseline are shown in Figure 2 1. Table 2 1 displays raw values and indicates which conditions differed at successive time points. Although baseline smoking motivation in the control conditi on was higher than craving in the other conditions, these differences were not statistically significant. Nonetheless, analyses were conducted on change scores to account for differences in smoking motivation that were present at baseline. Pre post change in QSU positive immediately after low and moderate intensity exercise was significantly greater than the change in QSU immediately after inactivity (control). Suppression of smoking motivation in the control versus moderate intensity conditions differed up to 30 minutes, after which change relative to baseline did not differ between any conditions. No significant differences in change relative to baseline were observed for QSU negative at any exercise intensity. Visual analysis of the group aggregate data s uggested that the biggest differences in QSU positive/QSU negative relative to baseline were present immediately

PAGE 32

32 after exercise. In other words, participants displayed (on average) large decreases in smoking motivation immediately after exercise that dissi pated quickly, and these decreases appeared to be greater after moderate intensity exercise relative to low intensity exercise or inactivity. These impressions were confirmed by a visual analysis of the individual participant data. Figure 2 2 displays indi vidual differences in the acute effects of exercise on positive and negative reinforcement components of smoking motivation (i.e., change scores between baseline and immediately after exercise/inactivity). The post exercise drops in QSU positive after mode rate (and low) intensity exercise were replicated in 19 (and 13) of 21 participants, respectively, and the magnitude of decreases was greater after moderate intensity exercise relative to low intensity exercise or inactivity. Reliability of Exercise Effect s on Smoking Motivation Table 2 2 displays raw values for QSU positive/QSU negative scores at g motivation levels relative to baseline (i.e., differences in QSU positive/QSU negative) did not differ significantly between the initial sessions versus the replication sessions ( ps > .05). This non significant effect suggested that effects of exercise i ntensity on smoking motivation were replicable across sessions of the same type. positive/QSU negative scores relative to baseline corresponded between their initial sessions and their replication sessions are displayed in Figure 2 3. A zero would indicate that the magnitude of change was identical across the two sessions, irrespective of what the raw values were. In contrast, positive values (i.e., data points > 0) across successive measurement times

PAGE 33

33 ind icate that the magnitude of change (e.g., exercise induced decrease in smoking motivation relative to baseline) was greater during the initial session than during the replication session. Negative values indicate that the magnitude of change was greater d uring the replication session than during the initial session. Visual analysis of these data suggests that effects of inactivity on QSU positive/QSU negative scores were the most reliable (i.e., values fell closest to zero on the control graphs). In contr ast to inactivity, there were visually detectable differences in the influence of low and moderate intensity exercise on smoking motivation between farther from 0 duri ng exercise sessions than they did during inactivity, indicating that the magnitude of exercise effects differed between the initial session and the replication session for some participants. With the exception of effects of moderate intensity exercise on QSU positive, the largest discrepancies (i.e., those cases in which the values were farthest from 0) appear to be the result of greater decreases in QSU scores Discussion Results o f Experiment 1 suggested that moderate intensity exercise more effectively decreased anticipated positive consequences of smoking than low intensity exercise or inactivity. However, although change relative to baseline differed between the moderate intensi ty and control conditions up to 30 minutes, this was partially attributable to the fact that smoking motivation increased across time after inactivity. The rapid increases in both positive and negative reinforcement associated with smoking that occurred af ter inactivity may have important clinical implications. By suppressing rises in smoking motivation that typically occur in the absence of exercise,

PAGE 34

34 brief bouts of exercise conducted throughout the day may eliminate common precipitates of smoking lapses ( based cessation interventions. The finding that moderate intensity exercise more effectively diminished anticipated positive consequences of smoking than low intensity exercise is consistent with Daniel et al. (2004), in which 5 minutes of moderate intensity cycling more low intensity cycling. These results also extend Daniel et al. by replicating effects of exercise on smoking motivation as measured by the QSU, as opposed to a single item index of smoking motivation. Because smoking motivation presumably differs both between participants (e.g., based on having different smoking histories) and within particip ants (e.g., changes dynamically over time based on the length of smoking deprivation), measuring both positive and negative reinforcement aspects of smoking motivation reflects a more valid conceptualization of the construct than single item indices. The finding that baseline smoking motivation based on positive reinforcement was greater than smoking motivation based on negative reinforcement is consistent with previous research indicating response differences across the two dimensions (Davies, Wilner, & Morgan, 2000; Eissenberg, Adams, Riggins, & Likness, 1999; Taylor et al., 2005; Tiffany & Drobes, 1991; Toll, Katulak, & McKee, 2006; Toll, McKee, Krishnan negative scores is that participa cause substantial increases in smoking motivation on this component. Specifically,

PAGE 35

35 QSU negative reflects anticipated relief from withdrawal symptoms, which are likely somewhat mild aft er 1 hour of not smoking. Low baseline QSU negative scores may have contributed, at least partly, to a floor effect in the statistical analyses, in that there was minimal opportunity for exercise to cause significant decreases on this component. Although previous research has employed parametric manipulations of exercise intensity, no research to date has evaluated individual differences in effects of exercise intensity on smoking motivation. Although Figure 2 2 indicated that (a) there were more decrease s in QSU positive immediately after moderate intensity exercise relative to low intensity exercise, and (b) the decreases in QSU positive induced by moderate intensity exercise tended to be larger than those induced by low intensity exercise, Figure 2 2 di d to moderate intensity exercise. displayed greater decreases in QSU positive after low relative to moderate intensity exercise, but for four of these participants, the difference ranged between a mere .4 to 4 points on a 100 point VAS. The fact that only four participants displayed substantially greater decreases in QSU positive after l ow relative to moderate intensity exercise (i.e., range = 13.2 48.4) combined with the fact that decreases in QSU positive immediately after moderate intensity exercise were replicated in 19 of 21 participants suggests that, for most smokers, moderate inte nsity exercise may be most likely to decrease A. Because the most convincing way to demonstrate the reliability of an effect is to replicate it, participants in Experiment 1 underwent one replication of one of the three exercise intensities. Statistical analyses indicated that, at all three exercise intensities,

PAGE 36

36 changes in smoking motivation relative to baseline did not differ significantly between individual part icipant data suggested that this may be due to the fact that, at least for low and moderate intensity exercise, participants varied in terms of whether the magnitude of exercise effects were greater during the initial session or the replication session. In other words, exercise exerted reliable effects on smoking motivation (i.e., values fell close to 0) for only a portion of participants. Had we assessed the reliability of exercise effects on smoking motivation using statistical analyses alone, variability in the extent to which exercise exerted reliable effects on smoking motivation would have gone undetected. In this sense, the results of Experiment 1 illustrate the utility of visual analysis as a more conservative method to evaluate the effects of a mani pulation on some outcome. It will be important for future research to incorporate both statistical and visual analyses into assessments of the reliability of exercise effects on smoking motivation in clinical, exercise based cessation interventions, which prescribe multiple exercise bouts over extended time frames. In sum, effects of moderate intensity exercise on the positive reinforcement component of smoking motivation suggested that moderate intensity exercise may be a treatment that decreases A. Howeve r, Experiment 1 was limited in that no objective measures of smoking were employed. Consequently, it was unclear whether effects of moderate intensity exercise on A would translate into reduced smoking. Thus, Experiment 2 was conducted to assess whether ef fects of exercise on self reported

PAGE 37

37 T able 2 1. Smoking Motivation Scores across Time (Experiment 1) Factor Condition Baseline 0 min 10 min 20 min 30 min 40 min 50 min 60 min Positive Control Low Moderate 2 (3) Wilcoxon 69.8(4.7) 54.9(5.1) 59.3(6.5) 3.7 73.1(3.9) 46.0(6.2) 40.9(7.0) 14.9 C vs L* C vs M* 74.9(4.1) 49.4(5.9) 52.7(6.2) 2.3 C vs M* 79.4(3.4) 50.9(6.3) 51.4(6.9) 2.3 C vs M* 79.2(3.7) 55.0(6.5) 60.2( 6.1) 1.4 C vs M* 77.5(4.1) 59.7(6.0) 63.9(6.2) 4.1 80.8(3.5) 62.4(5.9) 66.4(5.6) 2.8 81.3(3.3) 64.7(6.1) 71.1(5.4) 3.4 Negative Control Low Moderate 2 (3) Wilcoxon 42.5(6.4) 27.8(4.7) 32.4(6.4) 1.6 41.8(5.9) 27.7(4.9) 24.2(5.5) 3.0 42. 8(5.2) 31.5(5.7) 27.4(5.2) 3.6 44.1(5.9) 29.4(5.8) 30.1(5.6) 3.2 42.4(6.8) 31.2(6.6) 34.5(6.3) 1.2 46.0(6.1) 35.1(6.4) 35.0(6.3) 3.0 47.8(6.5) 34.8(6.5) 36.5(6.4) 7.5 49.4(6.8) 37.8(5.9) 39.8(6.0) 3.4 Note. Scores represent averages and erro r = SEM. Friedman statistics ( 2 (3) ) and differences indicated by the Wilcoxon conditions that differed significantly from one another. Higher scores reflect higher smo king motivation levels.

PAGE 38

38 T able 2 2. Smoking Motivation Scores during Replication Sessions (Experiment 1) Factor Condition Baseline 0 min 10 min 20 min 30 min 40 min 50 min 60 min Positive Control Low Moderate 57.6(9.7) 44.4(8.5) 77.7(6. 4) 58.9(8.5) 19.7(7.3) 63.0(13.1) 61.7(8.4) 36.9(8.9) 68.6(12.9) 61.9(7.1) 48.8(12.2) 80.2(8.5) 63.6(7.3) 47.8(9.1) 81.4(7.7) 65.2(7.7) 50.4(10.9) 81.9(7.9) 65.1(7.6) 49.0(10.4) 86.7(4.3) 66.9(8.1) 54.7(12.2) 86.0(4.9) Negative Cont rol Low Moderate 29.2(8.5) 20.0(4.8) 52.5(14.2) 38.7(9.4) 12.1(5.0) 50.0(14.2) 35.1(11.4) 23.8(7.3) 56.3(17.3) 37.7(11.0) 18.8(5.4) 60.0(16.0) 40.5(10.6) 20.7(6.5) 59.4(15.4) 39.0(11.6) 23.3(6.7) 63.5(15.0) 38.0(10.8) 21.0(6.6) 57.0(1 5.7) 41.0(10.5) 24.0(9.3) 62.0(17.7) Note. Scores represent averages and error = SEM. Higher scores reflect higher smoking motivation levels.

PAGE 39

39 Figure 2 1. Average change in QSU positive (left) and QSU negativ e (right) scores at successive measurement times across different exercise intensities (Error = SEM). Asterisks indicate those conditions that differ significantly from one another.

PAGE 40

40 Figure 2 2. Individual participant changes in QSU positive (left) and QSU negative (right) score s after treatment (i.e., before versus immediately after exercise) across differing exercise intensities. Individual participant difference scores are ordered from left to right according to the magnitude of their decreases in smoking motivation (greatest decreases begin left).

PAGE 41

41 Figure 2 3. Correspondence between participant QSU positive (left) and QSU negative (right) scores relative to baseline between initial sessions versus replication sessions at control (top), low (middle), or moderate intensity exercise (bottom). Each set of connected data points represents an individual participant. Values closer to zero indicate that the magnitude of exercise effects on smoking motivation was similar across both sessions.

PAGE 42

42 CHAPTER 3 EXPERIMENT 2 Rationale Experiment 1 indicated th at moderate intensity exercise was associated with greater decreases in the positive reinforcement component of smoking motivation than low intensity exercise or inactivity. However, whether reductions in the anticipated positive reinforcement derived from smoking translated into subsequent reductions in smoking was not assessed. Thus, the purpose of Experiment 2 was to determine ad libitum smoking. Only a few previo us studies have measured both self reported smoking motivation and latency to ad libitum smoking (e.g., Faulkner, Arbour Nicitopoulos, & Hsin, 2010, Reeser, 1983; Thayer et al., 1993, Taylor & Katomeri, 2007). However, all of these studies employed single item smoking motivation measures. In addition, with the exception of Faulker et al. (2010), these studies employed measures of latency that required either direct observation by researchers, or self reports about smoking by participants. For example, follo wing either exercise or seated activities, participants in Reeser (1983) were observed by researchers who recorded the time that cigarettes were lit, the number of puffs taken, and the time that cigarettes were put out (researchers pretended to be fixing l aboratory equipment or entering data during these smoke in the laboratory, Thayer et al. (1993) and Taylor and Katomeri (2007) asked participants to report this information after they departed the lab. Specifically,

PAGE 43

43 participants in Taylor and Katomeri were instructed to text the experimenters when they initiated their first cigarette following either an exercise or control session. In contrast to the studies described above, Faulker et al. (2010) captured several objective indices of smoking using smoking topography equipment. In this work, participants smoked ad libitum following either exercise or seated activities, and the equipment from which they smoked collected smokin g topography data (e.g., puff number, puff volume, interpuff interval) passively. The results of this work showed that relative to seated activities. Although one might e xpect smoking cigarettes using this equipment to influence smoking behavior, research has shown that collecting topography measures in this way is an ecologically valid and reliable method to index conventional smoking ( Lee, Malson, Waters, Moolchan, & Pic kworth, 2003 ). Thus, Experiment 2 sought to examine the relationship between components of smoking motivation (as measured by the QSU) and objective indices of smoking. Method Participants Participants were recruited using the same methods described in Ex periment 1 (e.g., flyers, advertisements). Participants in Experiment 2 also underwent the same phone and in person screening, and inclusion and exclusion criteria were the same as they were in Experiment 1. The resulting group included 20 smokers (15M, 5F ). They had a mean age of 39.5 years ( SD = 13.1), had been smoking for 17.4 years ( SD = 10.6), smoked 15.6 cigarettes per day ( SD = 5.1), and had an FTND score of 4.7 ( SD = 1.72). Eight participants were Caucasian, seven were Black, and two were Hispanic

PAGE 44

44 ( three participants provided no information about their race). As in Experiment 1, participants were mailed a $40.00 check upon completing all four experiments. Apparatus and Materials Experimental sessions occurred in the same smoking rooms as Experiment 1. Plowshare topography and software (Baltimore, MD) were installed on both computers. A cigarette mouthpiece was connected to the Plowshare equipment via a hose and recorded the times at which participants smoked. Cigarettes were inserted, lit, and smoked using this mouthpiece. Prior to each session, the researcher demonstrated how to insert and smoke cigarettes using the equipment, and written instructions were displa yed on the monitor. Participants inserted their cigarette into the mouthpiece, lit it, experimenter). Experimental Design An ABAB within subjects design was employed, in which A represented non exercise days and B represented exercise days. Par ticipants were pseudo randomly assigned to one of two possible orders: ABAB (n = 10) or BABA (n = 10). Procedure minimum of one day in between sessions, and the time of sessions was kept as consistent as possible within participants. Sessions began as they did in Experiment 1 (e.g., smoke half of one cigarette, 1 hour smoking deprivation period). After the

PAGE 45

45 deprivation period, participants completed the QSU brief (Cox et al. 2001) , then underwent 20 minutes of either inactivity or moderate intensity exercise. Because Experiment 1 suggested that moderate intensity exercise more effectively decreased the positive reinforcement component of smoking motivation than low intensity exerc ise, all participants exercised at a moderate intensity in Experiment 2. As in Experiment 1, moderate intensity was defined in terms of heart rate (i.e., 40 60% the K arvonen method (Karvonen et al., 1957). All aspects of the exercise sessions in Experiment 2 were identical to Experiment 1 (e.g., 2 minute min warm up, 17 minutes of stationary biking). During control (inactivity) sessions, participants had access to heal th related magazines, a DVD, and laptops with Internet access. Heart rates were monitored during both exercise and inactivity. After exercise/inactivity, participants completed the QSU for a second time, and then underwent a 2 hour, ad libitum smoking per iod. The smoking period took place in the rooms described above (see Apparatus and Materials). Participants smoked freely during the 2 hour period, and were instructed simply to indicate when they were initiating and stopping cigarette smoking (i.e., by pr completing the study as opposed to their behavior during sessions, there was no clear incentive to cheat (e.g., smoking without pres suggested that participants correctly followed instructions (although no formal inter observer agreement calculations were conduc ted based on these videos).

PAGE 46

46 Consequently, ad libitum smoking sessions were not videotaped in the present experiment. Participants departed the laboratory after 2 hours elapsed. Regardless of session type, all sessions proceeded according to the schematic b elow: Smoke 1 hr QSU Brief (2 3 min) Exercise/Control (20 min) QSU Brief Ad lib smoke (2 hr) Data Analysis Statistical analyses were conducted using SPSS (version 21). The outcomes cies (in minutes) to prevented the assessment of smoking topography data other than latency (which was rather than collected passively). Preliminary analyses using one way analyses of Whitney U tests for delays to smoke (not normally distributed) (signifi cance criterion = p < .05) indicated that outcomes did not differ as a function of session order (i.e., ABAB versus BABA), session number, or gender, thus these variables were not used as covariates in subsequent analyses. Because outcomes did not differ across sessions of the same type, outcomes were collapsed across replications of each session. Paired samples t tests (significance criterion = p < .05) were conducted to assess differences in QSU positive/QSU negative before and after exercise/control se ssions, and to assess differences in pre post change scores. As in Experiment 1, we also calculated the percentage of individual participants whose data paths replicated effects of exercise on smoking motivation that were observed at the group level.

PAGE 47

47 Both between and within participant analyses were conducted to assess effects signed ranks tests were conducted to assess differences in latency as a function of treatment (i. e., exercise versus inactivity, significance criterion= p < .05), and visual the reliability of exercise effects on latencies within participants, and (2) individual dif ferences in the magnitude of exercise effects on latency to smoke. Finally, mediation analyses were conducted using a bootstrapping approach (Preacher & Hayes, 2004; Janse Van Rensburg et al., 2013) to assess whether QSU positive and/or QSU negative score s following exercise/inactivity mediated the relationship between treatment and the latency to smoke. This approach examines 1, c ) of treatment is significantly different from zero (Pr eacher & Hayes, 2004, as described by Fjeldsoe, (Figure 3 1, Panel A, c controlling for proposed mediat ors (Figure 3 1, Panel B, ). Thus, results of the analysis estimate: (1) the effect of treatment (exercise versus inactivity; X) on changes in the potential mediator (M; coefficient; Figure 3 1); (2) the effect of change in the potential mediator (M) on changes in the dependent variable (Y) after controlling for treatment (X; coefficient; Figure 3 1); (3) the product of the and and (4) asymmetric 95% percentile confidence intervals for the coefficient based on bootstrapping methods (5,000 samples taken). Statistically

PAGE 48

48 significant indirect effects were indicated by confidence intervals that did not include zero. Results Manip ulation Checks We first validated the effectiveness of our HR manipulation. A paired samples t M = 125.0, SD = 15.5) were M = 74.7, SD = 7.1), t (38) = 18.0, p < .001. The moderate intensity criterion was achieved on 34 out of 40 (85%) exercise fell above (N = 2) or below (N = 4) the 40 60% HRR range by an average of 11.3 bpm ( SD = 8.9 bpm). Smoking Motivation Figure 3 exercise/inactivity. Statistical analyses indicated that QSU positive did not differ before exercise ( M = 61.6, SD = 28.7) and inac tivity ( M = 58.0, SD = 28.2), but was significantly lower after exercise ( M = 49.9, SD = 33.1) relative to inactivity ( M = 66.8, SD = 26.2), t (39) = 4.0, p < .001. Similarly, QSU negative did not differ before exercise ( M = 38.9, SD = 28.1) versus inactivi ty ( M = 38.6, SD = 27.2), but was significantly lower after exercise ( M = 34.1, SD = 29.8) relative to inactivity ( M = 44.5, SD = 27.2), t (39) = 2.7, p < .01. Paired samples t tests on changes in smoking motivation relative to baseline indicated that pre p ost differences in QSU positive on exercise days ( M = 12.1, SD = 24.8) were significantly different from pre post differences in QSU positive on non exercise days ( M = 8.8, SD = 24.4), t (38) = 4.9, p < .001. Pre post differences in

PAGE 49

49 QSU negative on exercis e days ( M = 5.3, SD = 24.0) also differed significantly from pre post differences on non exercise days ( M = 5.8, SD = 9.5), t (38) = 3.1, p < .01. immediately after exercise /inactivity) replicated the effects that were observed at the group level. Specifically, decreases in QSU positive after moderate intensity exercise were replicated across 27 of 40 (67.5%) exercise sessions (40 sessions because there were 20 participants, and 2 exercise sessions per participant). Although effects of exercise on QSU negative were very small at the group level (i.e., a 5.3 point decrease on a 100 point visual analog scale), exercise induced decreases in smoking motivation on this component we re observed across 24 of 40 (60%) exercise sessions. Latency to Smoke Participant latencies (in minutes) to ad libitum smoking are shown in Figure 3 3. Latencies were significantly longer after exercise ( M = 20.9, SD = 33.0) than inactivity ( M = 4.0, SD = 8.8), p < .05. Figure 3 after exercise versus inactivity. These data indicate substantial individual differences in effects of moderate intensity exercise on latencies to smoke. Thus, the few exercise s essions on which participants completely abstained from smoking (N = 3) presumably contributed to the significant effect of exercise on latency observed at the group level. Mediation Analysis The regression coefficients and significance values that emerge d from mediation analyses are shown in Figure 3 5. Mediation analyses revealed that the effect size for indirect effect through QSU s = 1.6, 21.4), and the size of the indirect effect through QSU negative was 7.1, .7).

PAGE 50

50 Thus, the direct effect of treatment, and the indirect effect of treatment on delay through QSU positive, were significantly different from ze ro ( ps < .05). The indirect effect of treatment on delay through QSU negative was not significantly different from zero ( p > .05). Therefore, exercise decreased anticipated positive reinforcement associated with smoking, which in turn increased participant Discussion The results of Experiment 2 indicated that moderate intensity exercise decreased self reports of both positive and negative reinforcement aspects of smoking motivation, smoking, though there was substantial variability in the magnitude of these effects. In addition, the mediation analysis revealed that exercise uniquely influenced smoking motivation based on anticipated positive consequences of smoking, which in turn incr eased latencies to smoke. The effects of exercise on smoking motivation that were observed in Experiment 2 were consistent with those observed in Experiment 1. Specifically, both experiments revealed significant pre post decreases in QSU positive for moder ate intensity exercise but not for inactivity, though exercise induced decreases in QSU positive were replicated in 90.5% of sessions in Experiment 1 and 67.5% of sessions in Experiment 2. Thus, the significant effect of exercise on QSU positive observed a t the group level in Experiment 2 was presumably due in part to strong effects of exercise on this component among a small pool of individuals. Although differences in QSU negative change scores were only significant in Experiment 2, the magnitude of this decrease (i.e., 5.3) was comparable to the exercise induced decrease in QSU negative observed in Experiment 1 ( 8.2). Thus, the finding that effects of exercise on QSU negative were

PAGE 51

51 only significant in Experiment 2 was likely due to the inclusion of twice as many data points in the between participant analyses. Also similar to Experiment 1, baseline QSU negative scores were lower than baseline QSU positive scores in Experiment 2. Effects of moderate itum smoking are consistent with those observed in previous research (e.g., Faulkner, Arbour Nicotopoulos, & Hsin, 2010; Reeser, 1983; Taylor & Katomeri, 2007; Thayer et al., 1993). However, the results of Experiment 2 also expand previous work in numerous ways. First, whereas previous research has primarily manipulated exercise intensity via verbal instruction (e.g., telling participants to walk briskly, as if they were late for an appointment, Taylor & Katomeri, 2007) and measured heart rate as a byproduc t, participants in Experiment 2 were given an explicit target heart rate to achieve. Consequently, whereas the mean HR exhibited by participants in Taylor and Katomeri was 106 bpm ( SD = 13.3) fallen in the low rather than moderate intensity range the mean HR exhibited by participants in Experiment 2 was 125.0 bpm ( SD = on 85% of exercise sessions. These data support the validity of our protocol for manipulating exercise i ntensity in the laboratory, and suggest that operationalizing intensity in terms of HR may permit greater experimental control over exercise intensity relative to verbal instruction. Experiment 2 also expands previous research by investigating individual differences in effects of exercise on smoking. Although the large standard deviations in suggestive that such variability was present (e.g., exercise: M = 83.7, SD = 61.1;

PAGE 52

52 c ontrol: M = 26.6, SD = 34.4), individual participant data was not discussed in that present results, it seems feasible that individual differences may have been present albeit undetected due to the exclusive use of between participant analyses. Finally, Experiment 2 also expands previous work by evaluating whether effects of exercise on smoking are mediated through positive and negative reinforcement components of smokin g motivation. Although Faulkner et al. (2010) assessed the relationship between post exercise reductions in smoking motivation and latency to smoke, our inclusion of a multidimensional measure of smoking motivation and a formal mediation analysis extends this work by suggesting that the mediational role of craving is unique to anticipated positive reinforcement derived from smoking. Of course, this relationship holds true only for those participants who were responsive to exercise effects on latency to smo ke, and the variables associated with responsiveness to exercise in the present work were unclear. Specifically, despite individual differences in effects of exercise on smoking motivation and the latency to ad libitum smoking, preliminary analyses did not indicate differences in these outcomes as a function of smoking history (e.g., FTND score, age at which one started smoking). In sum, Experiment 2 demonstrated that modera te intensity exercise reduced smoking by decreasing the anticipated, positive consequences of smoking, though individuals differed substantially in the magnitude of this effect. Interpreted in terms of temporal discounting, some smokers may have been unre sponsive to exercise because exercise decreased the value of smoking (i.e., by decreasing A in Equation 1 1), but it

PAGE 53

53 did not increase the value of abstaining. Thus, the value of smoking even after being diminished by exercise may have continued to exceed t he value of the likely delayed consequences associated with abstaining. Importantly, interpreting the results of Experiment 2 solely in terms of a temporal discounting framework overlooks another, more proximal contributor to smoking in this experiment. S pecifically, in the context of Experiment 2, there were no contrived reinforcers (e.g., money, social praise) and few if any natural reinforcers for abstaining (e.g., the 2 hour ad libitum smoking period was presumably too short to experience delayed healt h benefits associated with abstaining). Consequently, one purpose of Experiment 3 was to assess whether a combined manipulation to simultaneously decrease the value of smoking (exercise) and increase the value of abstaining (a laboratory based CM model, in which contrived reinforcers are delivered contingent on abstinence) would be more effective than exercise or CM alone. Additionally, although we have focused primarily on just two treatment implications derived from contextualizing cigarette smoking in te rms of temporal discounting (i.e., manipulations to decrease A and/or decrease D may reduce smoking), we have acknowledged that there is a third possibility. Specifically, a manipulation that decreases the rate at which smokers devalue delayed reinforcers (i.e., k values) may also decrease smoking. Because Experiment 2 only evaluated whether effects of exercise on smoking were mediated through self reported smoking motivation ( A ), a second purpose of Experiment 3 was to evaluate whether effects of exercise on smoking were mediated through A and/or k.

PAGE 54

54 X = Treatment (exercise or inactivity) Y = Latency (in minutes) to ad libitum smoking M = Post exercise/post inactivity scores for QSU positive and QSU negative (tested separately) c Figure 3 1. Hypothesized mediation model for effects of exercise on latency to ad libitum smoking. path path c Independent variable (X) Dependent variable (Y) Panel A Dependent variable (Y) Independent variable (X) Panel B Potential Mediator (M)

PAGE 55

55 Figure 3 2. Mean QSU positive (left) and QSU negative (right) scores before and after exercise (dark gray) and inactivity (light gray) (Error = SD). Asterisks indicate statistical significance.

PAGE 56

56 Figure 3 3. Mean latency to ad libitum smoking (in min utes) during exercise and control that particular treatment. The asterisk indicates that control latencies differed significantly from exercise latencies.

PAGE 57

57 Figure 3 4. Individual participant latencies (in minutes) to ad libitum smoking during control sessions (left) and exercise sessions (right). Latencies on individual sessions are ordered from shortest to longest (from left to right).

PAGE 58

58 Figure 3 5. Regression coefficients and significance levels (* = p < .05, ** = p < .01, *** = p < .001) associated with mediation analyses (Preacher & Hayes, 2004). 9.9* .1 .5*** 10.4 16.9** Treatment QSU positive QSU negative Latency (in min) to smoke

PAGE 59

59 CHAPTER 4 EXPERIMENT 3 Rationale As was noted previously, a temporal discounting framework implies that the choice to smoke a cigarette or abstain may be influenced by either smoking motivation ( A k ). k indexes the rate at which delayed reinforcers decline in value as a function of the delay until their receipt, with higher k k will steeply discount the delayed consequences that are likely to result from abstaini ng (e.g., better health, saving money), and therefore choose to smoke in most choice situations. The area under the discounting curve (AUC) is another way to index discounting rates. Because k is the slope of the discounting curve, k and AUC exhibit a stro ng, negative relation with one another. Thus, impulsive individuals have high k values and low AUC values (vice versa for self controlled individuals). Recently, researchers have begun to argue that because the units on k (i.e., the inverse of time) are n ot intuitive, a preferable way to index discounting is using ED50 values, where ED50 = the delay that effectively discounts the value of the delayed reinforcer by 50% (Yoon & Higgins, 2008). Research showing that exercise diminishes brain activity in areas associated with motivation and reward upon exposure to smoking cues (Janse Van Rensburg et al., 2009), as well as Experiment 1 and 2 data indicating that exercise decreases positive reinforcement aspects of A, suggest that exercise diminishes the immediat e, reinforcing value of smoking. If so, exercise may also decrease the extent to which the likely delayed consequences associated with abstinence are devalued relative to the

PAGE 60

60 immediate reinforcement associated with smoking (i.e., exercise may decrease k i n Equation 1 1). This, however, is a separate mechanism through which exercise may influence smoking, and may or may not mediate the relation between exercise and smoking in addition to effects of exercise on A. Although discounting tends to be fairly stab le over time (Beck & Triplett, 2009; Kirby, 2009; Ohmura, Takahashi, Kiramura, & Wehr, 2006), previous research has shown that reinforcer values can be manipulated experimentally to promote self controlled choice (e.g., Bickel, Yi, Landes, Hill, & Baxter, 2011; Black & Rosen, 2011; Giordano et al., 2002). Further, some of the brain areas in which Janse Van Rensburg et al. (2009) observed diminished responsiveness to smoking images after exercise (e.g., striatum, orbitofrontal cortex) are the same areas tha t are active when an individual chooses smaller, immediate reinforcers in delay discounting tasks (Mcclure, Laibson, Loewenstein, & Cohen, 2004). Thus, effects of exercise on the brain areas that promote impulsive choice might represent the neural underpin nings for effects of exercise on k. Consequently, one purpose of Experiment 3 was to evaluate whether effects of exercise on discounting rates represent a behavioral mechanism in addition to effects of exercise on A through which exercise decreases smokin g. A second purpose of Experiment 3 was to evaluate whether effects of exercise on smoking are enhanced when it is combined with a manipulation to increase the value of abstinence. As discussed previously, CM allows individuals to earn motivational incen tives (e.g., vouchers exchangeable for goods or services) contingent on meeting objectively verifiable goals (e.g., smoking abstinence, Higgins et al., 2002; JABA , 2008; Lussier et al., 2006). Thus, CM promotes abstinence by providing alternative, non dru g

PAGE 61

61 reinforcers at delays that are substantially shorter than the delays that typically ensue before one contacts the natural reinforcers that are likely to accompany drug abstinence. In other words, CM offers alternative commodities at comparatively short delays contingent on abstinence (i.e., CM decreases the D associated with each alternative commodity). Importantly, conceptualizing CM as a treatment that decreases D is in some sense an oversimplification. That is, in addition to decreasing D , CM also in volves a new contingency. Just as exercise might reduce smoking singly via its effects on A and/or k , CM might reduce smoking singly via its effects on D . In other words, effects of exercise on smoking might be due solely to decreases in smoking motivati on and/or temporal discounting rates, and effects of CM on smoking might be due solely to increases in the value of abstinence. Like exercise then, CM alone addresses only one contributor to smoking abstinence. Thus, even after the value of abstinence is enhanced by decreasing the D smoking commodities , the immediate, reinforcing value of smoking may continue to exceed the value of abstaining. From a temporal discounting perspective, this is why some smokers do not achieve drug free periods during CM (Dallery & Raiff, 2007; Iguchi et al., 1996; Roll et al., 1996; Silverman et al., 1996). Because exercise and CM target both the value of smoking and abstinence, respectively, combining them should more effectively reduce smoking than either component on its own. Support for this hypothesis has been garnered from research conducted in naturalistic settings. For example, Weinstock, Barry, and Petry (2008) showed that individuals who engaged in exercise related activities while undergoi ng

PAGE 62

62 intensive outpatient CM achieved longer durations of abstinence than individuals who did not complete exercise related activities. Thus, the primary purpose of Experiment 3 was to assess the effectiveness of a combined, exercise plus CM approach to red ucing smoking relative to its independent components (i.e, exercise and CM alone). A secondary purpose was to evaluate whether effects of exercise on smoking were mediated through A and/or k. Method Participants Recruitment methods and inclusion/exclusion criteria were identical to those in the previous experiments. The resulting participants were 20 smokers (13M, 7F). They had a mean age of 41.4 years ( SD = 15.1), had been smoking for 18.5 years ( SD = 6.2), smoked 17.3 cigarettes per day ( SD = 9.4), and h ad an FTND score of 4.8 ( SD = 2.1). earnings were dependent on their in session behavior (described below), but could vary between $10.00 (minimum) and $20.00 (maximum). The y were paid in the form of a check upon completing all four experiments. Apparatus and Materials Experiment sessions occurred in the same smoking rooms as the previous experiments. The same Plowshare topography and software (Baltimore, MD) described in E xperiment 2 were used, but participants did not smoke ad libitum in the present experiment. Instead, the equipment was used to collect topography data passively while participants underwent the CM or CM control procedure (see Procedure). The CM (or contr ol) procedure was run using Microsoft Visual Studio ® 10.0, on a laptop placed directly in front of the desktop on which the Plowshare software was installed.

PAGE 63

63 tton was green when participants had the opportunity to smoke. To initiate until the next smoking opportunity. During this 30 second interval, participants were ins tructed to take one puff only. Participants who chose to smoke did so using the Plowshare topography equipment. Participants who chose not to smoke did not need to indicate this choice, but rather sat passively or engaged in other activities during the CM or CM control sessions. Experimental Design A within subjects design was employed, in which participants returned to the laboratory four times for four different experiment sessions: (A) exercise plus CM, (B) non exercise plus CM control, (C) exercise plu s CM control, and (D) non exercise plus CM. A Latin Square design was used to pseudo randomly assign participants to each of the following orders: ABCD, BCDA, CDAB, DABC (N = 5 per order). Procedure Regardless of session type, all sessions began as descr ibed in Experiments 1 and 2 (i.e., smoke half a cigarette, smoking deprivation period). After the 1 hour smoking deprivation period, participants completed the QSU Brief (Cox et al., 2001) and a computerized, monetary discounting task (Dallery & Raiff, 200 7). In this task, participants made a series of hypothetical choices about receiving a small sum of money that was immediately available (e.g., $75.00), or a larger sum that was available after a delay (e.g., $100.00 in two weeks). Impulsive choice refers to choosing the smaller, immediate sum, whereas self controlled choice refers to choosing the larger, delayed sum (Rachlin, Raineri, & Cross, 1991; Reynolds, 2006). The smaller, sooner

PAGE 64

64 a point of subjective equality (i.e., indifference point), at which point they repeated the series of choices at a new delay. Participants repeated these choices across eight different delays, which were presented in a random order, and the indifference p oints at each delay were used to calculate k, AUC, and ED50 (see Data Analysis). Sample choices in this task, and the rate of discounting calculated based on these choices, are shown in the Appendix. After completing the QSU and discounting task, partici pants underwent 20 minutes of exercise or inactivity. Exercise sessions were identical to the moderate intensity sessions described in Experiments 1 and 2 (i.e., 20 minutes of stationary biking at 40 exerci se sessions were also identical to those described in the previous experiments (i.e., 20 minutes of seated activities including computer use, reading, etc.). Immediately after exercise or inactivity, participants completed the QSU and monetary discounting task a second time. Last, they underwent the procedures associated with their particular CM assignment for that day. A session schematic is provided below: Smoke 1 hr Measures (15 min) Exercise/Control (20 min) Measures CM/Control (40 min) On CM days, participants underwent a computerized procedure that asked them to make a choice at 30 second intervals: take a puff on a cigarette or earn an increasing sum of money (Dallery & Raiff, 2007). The first interval that participants chose not to smo ke, they earned $0.015. This increased by $0.005 for every successive interval that participants chose not to smoke. Choosing to take a puff reset the monetary sum at its

PAGE 65

65 baseline value. Participants made choices at 30 second intervals for a 10 minute b lock, after which they had a 20 minute break, and then repeated the CM procedure three times (with 20 minute breaks in between successive blocks). Participants could earn a maximum of $1.25 per block ($5.00 per session) on CM days, thus the CM condition i n the present study replicated the low magnitude condition in Dallery and Raiff (2007). These particular amounts generated adequate variability in the percentage of participants who smoked versus abstained in Dallery and Raiff, therein preventing either fl oor or ceiling effects. In addition, the escalating schedule of reinforcement used in this procedure has been shown to promote prolonged periods of smoking abstinence in several clinical studies (Petry & Simcic, 2002; Romanowich & Lamb, 2010). On CM contro l days, participants underwent a similar, computerized procedure in which they chose between taking a puff on a cigarette or remaining abstinent. However, this procedure awarded participants $0.0625 every 30 seconds regardless of whether they smoked. The amount received per block was thus equal to the maximum sum that could be earned per block during CM ($1.25). By equating earnings across the two conditions but delivering money non contingently on CM control days, we were able to isolate the effects that escalating pay contingent on abstention (i.e., CM) had on smoking. In addition, the CM control condition approximates the control conditions used in clinical interventions, in which reinforcement is contingent solely on providing biological samples from wh ich smoking can be verified (e.g., breath CO samples, Dallery et al., 2007), irrespective of whether these samples indicate that an individual smoked. After the CM or CM control procedure, participants were permitted to leave the laboratory.

PAGE 66

66 Unlike both of the previous experiments, as well as Dallery and Raiff (2007), participants did not receive additional compensation for their participation beyond what they earned during CM or CM control sessions. Upon completing all four sessions, a check was mailed session earnings. This amount could range from $10.00 (if they smoked every 30 seconds on both CM days), to a maximum of $20.00 (if they abstained during all opportunities to smoke on both CM days ). On average, participants earned $4.19 ( SD = $1.14) during CM sessions (mean total earnings for all four sessions = $18.38). Data Analysis All statistical analyses were conducted using SPSS (version 22). Dependent asures of smoking (i.e., latency to smoke, total puff number, puff volume), subjective measures of smoking motivation (i.e., QSU positive, QSU k, AUC, session was not completed (i.e., the participant departed the laboratory before the CM control portion on her non exercise plus CM control session) there was an N of 20 for analyses in which smoking motivation and discounting rates were the outcomes, and an N of 19 for analyses in which objective measures of smoking were the outcomes. Prior to evaluating whether outcomes differed as a function of condition, one way t tests were conducted to determine whether outcomes differed based on demographic variables (e.g., race, gender, income), session order, or session number (Kruskal Wallis tests and Mann Whitney U tests were used for variables that were not normally distributed). Because

PAGE 67

67 outcomes did not di ffer based on these variables, no covariates were used in subsequent analyses (Janse van Rensburg et al., 2013). Objective measures of smoking Because none of the objective measures of smoking were normally distributed, Friedman tests were conducted to de tect differences in latency to smoke, puff number, and puff volume as a function of condition (significance criterion = p < .05), and post hoc, Wilcoxon signed rank follow up tests were conducted to determine which conditions differed from one another. Su bjective measures of smoking motivation baseline for each session. QSU positive and QSU nega tive change scores were collapsed across the two exercise/inactivity sessions, and paired samples t tests were conducted to determine whether QSU positive/QSU negative change scores differed as a function of exercise versus inactivity (significance criteri on p < .05). Monetary discounting task systematic data were indifference points across successive delays do not exceed the previous indifference indifference points could not exceed previous indifference points by > $20.00), and (b) the indifference point at the shortest delay m ust exceed the indifference point at the indifference points at 1 week had to be $10.00 greater than their indifference points at

PAGE 68

68 10 years). Discounting data from each admi nistration time (i.e., one at intake, plus before and after exercise/inactivity across 4 sessions = 9 administrations total per participant) were evaluated independently according to these criteria, and only non systematic administrations were excluded fro m subsequent analyses. In total, 33.8% of the discounting data were dropped. The discounted value of delayed reinforcement in the monetary discounting task was indexed using k, AUC, and ED50. Equation 1 was fitted to the eight indifference points via le ast squares regression, and an estimate of k was obtained (Mazur, 1987). As explained by Morrison, Madden, Odum, Friedel, & Twohig (2014), area under the discounting curve (AUC; Myerson, Green, & Warusawitharana, 2001) is the area under the indifference p oints across the range of delays (i.e., 1 week, 2 weeks, 1 month, 4 months, 8 months, 1 year, 5 years, 10 years). AUC values range from 0 1, where a 0 would indicate exclusive preference for the smaller, immediate sum, and 1 would indicate exclusive prefe rence for the larger, delayed sum. Finally, the ED50 (i.e., the delay at which the large sum is subjectively equivalent to 50% of its value) is simply the reciprocal of k , or 1/ k (Yoon & Higgins, 2008). Although one advantage of using AUC rather than k to index discounting is that it is typically normally distributed, neither k nor AUC values were normally distributed in the present research, and k remained skewed even after applying a square root transformation. Rather than conduct parametric statistics using the square root of AUC values as the outcome, which involves units that are difficult to interpret, ED50 values (units = days) were used as the outcome in statistical analyses of the temporal discounting data. Data were collapsed across the two exer cise/inactivity sessions, and

PAGE 69

69 Mann Whitney U tests were conducted to assess differences in ED50 values before and after exercise versus inactivity (significance criterion = p < .05). In addition, visual analysis was employed to evaluate effects of exercis e versus inactivity on individual Interobserver agreement (IOA) Because participants completed the CM or control procedures independently in the experiment rooms, it was possible that they could smoke freely without indicat ing without using the Plowshare equipment). The incentive for doing this would be that it would allow participants to maximize their earnings on CM days without hav ing to abstain from smoking. Thus, measuring puff number in terms of the number of times s moking topography data. Specifically, all puff volumes > 15 milliliters (mL) were occur due to movement artifacts (e.g., bringing the mouthpiece up to the lips), and other researchers have employed similar criteria to dissociate true versus false puffs (e.g., McKee, Krishnan , 2006). In addition to indexing puffs in terms of the smoking byproducts described above, 56 blocks (i.e., approximately tw o, 10 minute blocks per participant) were randomly selected to be video recoded. Although the informed consent stated that participants may be video recorded on some sessions, they were not explicitly told which sessions were being recorded. The purpose o f recording participants using a minimally obtrusive method was to minimize reactivity. Two independent observers then viewed the video

PAGE 70

70 samples and counted the number of puffs (including lighting puffs) per block during CM or CM control. Given that there were three indices of puff number (i.e., clicks, volumes, and objectively observed puffs), the following measures of IOA were conducted: 1. sessions. To be conservative, participant summed to form a variable of total puffs. IOA for each videotaped block was calculated by dividing the smaller number of total puffs by the larger number of total puffs, and multiplying by 100. 2. IOA between the two byproducts of smoking. Agreement was defined as cases in which the number of clicks and the number of puffs > 15 mL were discrepant by < 2 (e.g., 2 clicks and 4 puffs was scored as an agreement). Disagreement was defined as any cases in which the number of clicks and the number of puffs were discrepant by > 2 (e.g., 2 clicks and 5 puffs). IOA was defined as: (number of agreements) / (total number of agreements plus disagreements) X 100. 3. IOA between video observed puffs and each byproduct. Recall that t wo observers viewed videos, thus the total puff number recorded was not identical in all cases. In cases where the total puff number differed, the average number of puffs recorded by the observers was taken (rounded to the nearest whole number), and this number was used in the following IOA calculations: a) An agreement between video observed puffs and the number of clicks was scored in all cases where the two numbers were discrepant by < 2. Disagreements were scored when the number of observed puffs and the number of clicks were discrepant by > 2. IOA was defined as: (number of agreements) / (total number of agreements plus disagreements) X 100. b) An agreement between video observed puffs and the number of puffs recorded via Plowshare was scored in all cases where the two numbers were discrepant by < 2. Disagreements were scored when the number of observed puffs and Plowshare indicated puffs were discrepant by > 2. IOA was defined as: (number of agreements) / (total number of agreements plus disagreeme nts) X 100. Although permitting discrepancies of < 2 in these calculations was arbitrary, some leniency was required given that (1) pilot data suggested that the smoking s, and (2) some participants were not positioned optimally in their videos, making it difficult to determine whether they were smoking at times.

PAGE 71

71 Results Manipulation Checks We first validated the effectiveness of our heart rate (HR) manipulation. Because preliminary analyses using one between exercise plus CM versus exercise plus CM collapse d across exercise/inactivity sessions. Paired samples t tests indicated that intensity exercise ( M = 119.1, SD = 17.1) than inactivity ( M = 77.5, SD = 11.7), t (35) = 17.1, p < .001. The moderate intensity cr iterion was achieved on 26 out of 40 (65.0%) exercise sessions. On those the target 40 60% range by an average of 9.1 bpm ( SD = 7.3 bpm). Objective Measures of Smoking Latency to smoke in Figure 4 1. The Friedman test indicated significant differences in latency across 2 (3) = 12.9, p < .01. Specifically, latencies for exercise plus CM sessions ( M = 26.1, SD = 17.3) were greater than latencies for both non exercise plus control ( M = 13.8, SD = 17.5), Z = 88.0, and exercise plus control sessions ( M = 13.7, SD = 17.5), Z = 16.0, ps < . 05, but not non exercise plus CM sessions ( Mdn = 27.0, SD = 16.2) , p > .05. Similarly, latencies for non exercise plus CM sessions were greater than latencies for both non exercise plus control, Z = 12.0, and exercise plus control sessions, Z = 9.0, ps < . 01. Thus, there was a main effect of CM condition on latency to smoke, in that

PAGE 72

72 latencies were longer on CM sessions than CM control sessions, irrespective of exercise. Individual participant latencies to smoke are shown in Figure 4 2. Because participants underwent four, 10 minute blocks of CM or CM control, a latency of 40 minutes indicates that no smoking occurred. These data suggest that the main effect of CM that was observed at the group level was due to the fact that a greater number of participants abstained from smoking during CM relative to CM control. Speci fically, complete abstinence from smoking occurred on 10 CM control sessions, whereas complete abstinence occurred on 22 CM sessions. Fourteen of 20 participants (70%) exhibited longer latencies during exercise plus CM relative to exercise plus CM control , and fourteen of 19 participants (73.7%) exhibited longer latencies during inactivity plus CM relative to inactivity plus CM control. Puff number Figure 4 3. The Friedman test indicated significant differences in total puff number 2 (3) = 23.6, p < .001. Specifically, total puffs during exercise plus CM ( M = 4.1, SD = 7.2) were less than total puffs during both non exercise plus control ( M = 14.4, SD = 13.0), Z = 6.0, and exercise plus control ( M = 13.7, SD = 13.8), Z = 15.9, ps < . 01, but not non exercise plus CM ( M = 4.9, SD = 6.6), p > .05. Similarly, total puffs during non exercise plus CM were less than total puffs during both non exercise plus control, Z = 7.5, and exercise plus control, Z = 104.0, ps < . 01. Thus, there was a main effect of CM condition on puff number, in that participants took fewer puffs during CM relative to CM control, irrespective of exercise.

PAGE 73

73 across sessions are shown in Figure 4 4. As with the latency data, these data suggest that the main effect of CM that was observed at the group level was due to the fact that a greater number of participants completely abstained from smoking during CM re lative to CM control. Thirteen of 20 participants (65.0%) took fewer puffs during exercise plus CM relative to exercise plus CM control, and twelve of 19 participants (63.2%) took fewer puffs during inactivity plus CM relative to inactivity plus CM control . Puff volume compensated for smoking less during CM by taking deeper inhalations. Individual averaged, and a Friedman test was conducted using this value as the outcome. The 2 (3) = 1.6, p > 4 5. A visual analysis of these data indicated minimal differences in volume across conditions, with the exception that there were no very high volumes on any non exercise plus CM sessions. The reason for this is that no topography data were collected for the participant who reliably had the highest volumes on all other sessions. Subjective Measures of Smoking Motivation 6. The magnitude of change in QSU positive differed significantly as a function of ex ercise versus inactivity, t (39) = 3.7, p = .001. Specifically, participants demonstrated increases in QSU positive after inactivity ( M = 7.6, SD = 14.7) and decreases in QSU positive after exercise ( M = 10.7, SD = 24.0). The magnitude of change in QSU n egative also differed significantly

PAGE 74

74 as a function of exercise versus inactivity, t (39) = 2.8, p = .01. As with QSU positive, participants demonstrated increases in QSU negative after inactivity ( M = 3.5, SD = 11.3) and decreases in QSU negative after exerc ise ( M = 5.9, SD = 14.5). positive/QSU negative scores relative to baseline for exercise/inactivity are shown in Figure 4 7. These data indicate substantial individual differences in effects of exercise on smoking motivatio n. Nonetheless, inactivity induced increases in QSU positive and QSU negative that were observed at the group level were replicated on 70% and 62.5% of non exercise sessions, respectively. Exercise induced decreases in QSU positive and QSU negative that were observed at the group level were replicated on 70% and 60 % of exercise sessions, respectively. Monetary Discounting Task Table 4 1 displays median values for all indices of discounting before and after exercise/inactivity. Mann Whitney U tests indi cated that ED50 values (in days) did not differ before exercise ( Mdn = 106.8, SD = 3,654.1) versus inactivity ( Mdn = 162.1, SD = 16,526.7), or after exercise ( Mdn = 141.2, SD = 6,308.8) versus inactivity ( Mdn = 112.0, SD = 10,048.8), all ps > .05. Figure 4 8 shows the distribution of AUC values across conditions. AUC was used as the outcome in these plots because the highly skewed nature of both k and ED50 (the reciprocal of k ) produced boxplots that were difficult to interpret. If exercise decreased impu lsivity relative to inactivity, this would be reflected in upward shifts in the boxplots (i.e., increased AUC values) after exercise relative to before exercise, whereas either no change or downward shifts in the boxplots would be observed for inactivity. Visual analysis of the boxplots before and after inactivity suggested minimal to no change in AUC values. Although median AUC values did not appear to differ

PAGE 75

75 substantially before and after exercise, the 75 th percentile was higher after exercise relative to before, suggesting that some portion of participants may have exhibited increased AUC values (i.e., lower discounting rates) after exercise relative to before. The above impressions were confirmed by a visual analysis of individual alues (Figure 4 9). The distributions look nearly identical before and discounting. Although there was more clustering at lower AUC values before exercise relative to a fter, there were also more systematic datasets before exercise (N = 26) relative to after (N = 21). In other words, because only the data from non systematic administrations were removed, a greater number of datasets were retained (and therefore a greate r number of AUC values were plotted) from the sample of discounting data collected before exercise relative to after. Thus, the appearance of small effects of exercise on discounting observed in Figure 4 8 may be due to a small number of participants who e xhibited higher AUC values after exercise. Interobserver Agreement (IOA) The first index of IOA calculated was between two independent observers who lighting puffs and regular puffs per block (which were added together to form a summary score of total puffs). IOA for total puffs was calculated for each block, and these percentages were averaged to yield an overall IOA of 90%. We also calculated IOA between the two byp roducts of smoking (i.e., number of mouse clicks and number of puffs indicated by the topography equipment). IOA about instances of smoking according to these measures was 86.3%.

PAGE 76

76 Finally, IOA was calculated between video observed puffs and each of the tw o byproducts. IOA between video observed puffs and the number of clicks was 92.3%. IOA between video observed puffs and Plowshare detected puffs was 77.8%. Given the comparatively low IOA yielded by this latter calculation, we attempted to examine the so urce of the discrepancy. One possibility was that identifying puff volumes of at least 15 mL as true puffs was too liberal of a criterion (e.g., McKee et al. 2006 excluded all puffs < 20 mL). If this were the case, then a majority of disagreements would b e due to Plowshare indicated puffs exceeding video observed puffs. However, Plowshare indicated puffs exceeded video observed puffs in only 58.3% of disagreements between these two indices of puff number. These data suggest that 15 mL was an appropriate c riterion, and that errors made by the Plowhare topography equipment in detecting true puffs were unsystematic (i.e., the equipment was not reliably oversensitive or under sensitive). Correlates of Treatment Effects r (continuous variables) and S variables) correlations were conducted to determine whether any demographic responsiveness to CM. Given no differences between CM with or with out exercise, correlations were conducted with the data collapsed across CM and CM control conditions. Results indicated that participants with higher resting heart rates took more puffs during CM, r = .35, and participants who discounted the value of dela yed reinforcement less steeply (i.e., less impulsive participants) exhibited longer latencies to smoke. Specifically, participants with higher AUC values (lower k values) exhibited longer latencies, r = .40 (for k, r = .40), ps < .05. Participants who dis played greater

PAGE 77

77 decreases in QSU positive also exhibited longer latencies to smoke, r = .39, p < .05. None of these relationships were observed for CM control sessions. On CM control sessions, participants who self reported abstaining from cigarettes for g reater durations at some point in their smoking history exhibited longer latencies, r = .36, p < .05. In addition, greater increases in QSU positive, r = .40, and QSU negative, r = .36, ps < .05, were associated with taking more puffs. None of these relati onships were observed for CM sessions. Discussion The primary purpose of Experiment 3 was to determine whether a combined, exercise plus CM approach to reducing smoking was more effective than its independent components. The results indicated that exercis e plus CM and inactivity plus CM were similarly effective at reducing smoking, whereas exercise plus CM control and inactivity plus CM control were similarly ineffective. That is, despite reducing smoking motivation relative to inactivity, exercise had no influence on any objective measures of smoking. Thus, the results of this experiment suggested that a combined, exercise plus CM manipulation was no more effective at reducing smoking than CM alone, and that exercise alone reduced anticipated reinforcement associated with smoking, but not smoking itself. The finding that CM decreased smoking for some, but not all smokers, is consistent with previous, laboratory based research (e.g., Dallery & Raiff, 2007, Mueller et al., 2009). In addition, the present res shallower discounting at intake (i.e., higher AUC values) was associated with longer latencies to smoke during CM. The results also expand previous research by evaluating whether the addition of exercise wo uld enhance the effectiveness of CM. Although our

PAGE 78

78 predictions about the utility of combining exercise and CM were not confirmed, this dissertation represents to our knowledge the first laboratory based assessment of a combined, exercise plus CM approach t o reducing smoking. The findings that baseline smoking motivation scores were lower on QSU negative than QSU positive, and that exercise induced deceases in QSU positive were greater than decreases in QSU negative, are consistent with the results of Ex periments 1 and 2. Having replicated this finding across all three experiments, results of this dissertation further support the notion that craving is a multidimensional construct that includes both positive and negative reinforcement components (Davies e t al., 2000; Eissenberg et al., 1999; Toll et al., 2006). As discussed in the previous experiments, greater effects of exercise on QSU positive relative to QSU negative may be due to floor effects (i.e., low baseline QSU negative scores prevented equivalen t, exercise induced decrements) and/or a smoking deprivation period of insufficient length (i.e., one hour may not produce elevations on this component). Based on research indicating that QSU negative items discriminate nicotine dependence significantly b etter than QSU positive items (Germeroth, Wray, Gass, & Tiffany, 2013), it is also possible that QSU negative captures more stable aspects of smoking behavior, and is therefore less likely to be influenced by experimental manipulations. In any case, the fa ct that effects of exercise were stronger for QSU positive across all three experiments suggests that researchers interested in evaluating the effect of experimental manipulations on smoking motivation ( A ) should administer multidimensional craving measure s rather than single item measures. Other behavioral

PAGE 79

79 manipulations may also vary in the extent to which they decrease anticipated positive versus negative reinforcement associated with smoking. In addition to the fact that QSU negative may be less prone t o influence by experimental manipulations, it may also be less predictive of objective measures of smoking in the human laboratory. This assertion is supported by Experiment 2 in which QSU positive uniquely mediated the relation between exercise and smokin g as well as Experiment 3 data indicating that exercise induced decreases in QSU positive were uniquely associated with responsiveness to CM. Additionally, in a laboratory based study that was similar to Experiment 3, Bisaga, Padilla, Garawi, Sullivan, and Haney (2007) showed that, given the choice between smoking and an alternative monetary reinforcer of differing magnitudes, only QSU positive items predicted the likelihood of smoking when higher magnitude reinforcers were available. Based on these results , Bisaga et al. (2007) concluded that anticipated positive consequences of smoking may be a better predictor of behavior in the human laboratory when smoking and alternative reinforcers are simultaneously available. Interestingly, effects of exercise on t he latency to smoke that have been reported in previous research (e.g., Reeser, 1983; Taylor & Katomeri, 2007), and in Experiment 2, were not replicated in Experiment 3. One contributor to this replication failure may be l below their targets on a greater portion of sessions in Experiment 3 (65%) relative to Experiment 2 (85%). Indeed, previous research (e.g., Daniel et al., 2004) as well as the results from Experiment 1 indicate that low intensity exercise may have less influence on smoking related variables than moderate intensity exercise. It should be noted, however, that Daniel et al., as well as a

PAGE 80

80 substantial portion of other studies on exercise and smoking (e.g., Everson et al., 2008; Scerbo et al., 2010; Taylor et al., 2005, 2006; Ussher et al., 2009), have administered craving measures without measuring smoking objectively. Because the results of Experiment 3 suggest that there may be a dissociation between objective measures of smoking and subjective measures of smoking motivation, future research should consider measuring both, as smoking motivation is not necessarily an accurate proxy for smoking. Additional contributors to the null effect of exercise on smoking observed in the present experiment are difference s between the present sample and those used in both Experiment 2, and other previous work. Although characteristics reflective of Experiment 3 participants smoked slightly more cigarettes per day (17.3 versus 15.6), and reported that they had been smokers for slightly longer (18.5 versus 17.4 years), than Experiment 2 participants. It seems unlikely that these very small differences contributed to the null effects of exercise on smoking, but perhaps the combination of to some participants being unresponsive to exercise. It will be important for future research to identify those variables that predict responsiveness to exercise based cessation interventions. Unlike exercise, some research has investigated those variables that are associated with responsiveness to CM (e.g., temporal discounting rates in Dallery & Raiff, 2007). The present experim ent also reported a relationship between resting heart rates at intake and the number of total puffs during CM. Although resting heart rate is

PAGE 81

81 resting heart rates (Groppe lli, Giorgi, Omboni, Parati, & Mancia, 1992), thus the present research is broadly consistent with previous work in which heavier smokers have been shown to be less responsive to CM (e.g., Dallery et al., 2007; see Lamb, Kirby, Morral, Galbicka, & Iguchi, 2004, for a Discussion). The results of Experiment 3 are also effectiveness of CM (Littlejohn, 2006; McCay et al., 2005; Rash, Olmstead, & Petry, 2008). Although correlates of responsiveness to CM have been investigated to a greater extent than correlates of responsiveness to exercise, continued efforts to identify predictors of responsiveness to CM are likely to be highly impactful. Such research will provide a foundation f or developing and prescribing treatments involving CM based on Although previous research has shown that experimental manipulations can influence k (e.g., working memory training in Bickel et al., 2011; money ma nagement training in Black & Rosen, 2011; mindfulness based acceptance practice in Odum et al., One reason for this discrepancy may be that the manipulations employed i n previous work differed in nature from exercise. Specifically, the manipulations mentioned above prefrontal brain areas the same areas that are activated when individuals cho ose larger, delayed reinforcers in temporal discounting tasks (e.g., McClure et al., 2004). Although a behavior analytic view does not acknowledge the extent to which these tasks engage specific brain areas as a cause of their effects on discounting, neur al

PAGE 82

82 variables can serve as motivating operations, antecedent stimuli, and reinforcers in a functional analysis of operant behavior (Thompson, 2007). Thus, one possibility is that manipulations that activate prefrontal brain areas serve as abolishing operati ons that immediately available sum of money in the temporal discounting task). In contrast, rather than activate prefrontal brain areas, exercise has been shown to reduce a stimuli (e.g., Jans Van Rensburg et al., 2009, 2011). These are the same areas that are activated when individuals choose smaller, immediate reinforcers in temporal discounting task s (McClure et al., 2004). Although one might expect increasing brain activity in those areas involved in self controlled choice to exert comparable effects on discounting as decreasing activity in those areas involved in impulsive choice, it is possible th at manipulations of the latter sort (e.g., exercise) are less robust. In other words, perhaps manipulations that decrease activity in reward and motivation regions are weaker abolishing operations than manipulations that increase prefrontal activity, thus the subsequent devaluation of smaller, immediate reinforcers is insufficient to promote self controlled choice. The fact that indices of discounting changed in the predicted direction as a function of exercise, but were not statistically significant, may s upport this possibility. In sum, Experiment 3 was conducted to determine (1) whether a combined, exercise plus CM manipulation was more effective at reducing smoking than its independent components, and (2) whether effects of exercise on smoking were medi ated through smoking motivation ( A k ).

PAGE 83

83 The results indicated that the combined approach was no more effective than CM alone, and that exercise alone was ineffective. Given that there was no effect of exercise on smoking, an assessment of whether A and/or k mediated this effect was not possible. Once future research has identified predictors of responsiveness to exercise based approaches to reducing smoking, the variables that mediate this effect could be investiga ted among a larger sample that includes a higher portion of responsive smokers. Identifying the behavioral mechanisms through which exercise effects are mediated may elucidate some variables to target in clinical, exercise based cessation interventions.

PAGE 84

84 Table 4 1. Median scores on the three indices of temporal discounting before and after exercise versus inactivity Inactivity Exercise Before After Before After K {days 1 ) .00617 (.03) .00596 (.84) .00938 (.02) .00708 (.02) AUC .29 (.29) .28 (.27) .30 (.26) .36 (.31) ED50 (days) 162.1(16,526.7) 112.0 (10,048.8) 106.8 (3,654.1) 141.2 (6,303.8) Note. Numbers in parentheses represent standard deviations

PAGE 85

85 Figure 4 1. Mean latencies to smoke (in minutes) as a fun ction of session type (Error = SEM).

PAGE 86

86 Figure 4 2. Individual participant latencies (in minutes) to smoke. Latencies of 40 min Indicate that the participant abstained from smoking for the entire session.

PAGE 87

87 Figure 4 3. Mean number of puffs per block across conditions (Error= SD).

PAGE 88

88 Figure 4 an individual participant. Note: Puffs are ordered from lowest number of puffs to highest number of puffs per condition (absence of bars indicates no smoking).

PAGE 89

89 Figure 4 5. point represents an average of puff volumes across all four b locks of a particular condition.

PAGE 90

90 Figure 4 positive (dark gray) and QSU negative (light gray) scores before a nd after inactivity (left) and exercise (right) (Error = SEM). Bottom: Change in QSU positive and QSU negative scores relative to baseline as a function of exercise (light gray) versus inactivity (dark gray) (Error = SD). ** = p < .01, *** = p < .001.

PAGE 91

91 Figure 4 (right). Each bar represents an individual participant, and values are ordered from the largest decreases (left) to largest increases (righ t) (40 per graph because each participant underwent 2 exercise sessions and 2 inactivity sessions). Pre Post Treatment Change in VAS Rating Participant Inactivity Exercise

PAGE 92

92 Figure 4 8. Box plots showing the distribution of AUC values as a function of condition. Error bars range from the 10 th to 90 th percentiles, boxes from the 25 th to 75 th percentiles, and the horizontal lines in the center of each box represent the condition medians.

PAGE 93

93 Figure 4 exercise (right).

PAGE 94

94 CHAPTER 5 GENERAL DISCUSSION Based on a conceptualizati on of cigarette smoking in terms of temporal discounting, the overarching aim of present research was to test the effectiveness of a combined (exercise plus CM) approach to cigarette smoking relative to its independent components. En route to achieving thi s aim, three experiments were conducted: Experiment 1 demonstrated that moderate intensity exercise more effectively reduced self reports of the anticipated positive consequences of smoking relative to low intensity exercise or inactivity. Experiment 2 ind icated that the effect of moderate intensity exercise on anticipated positive reinforcement mediated the relation between exercise whether the combination of exercise plus C M more effectively reduced smoking than exercise or CM alone. Results suggested that the combination was no more effective than CM alone, and exercise alone decreased smoking motivation but had no influence on smoking itself. Although exercise induced decr eases in QSU positive and QSU negative did not translate into reduced smoking in Experiment 3, an argument for implementing the combined intervention in a naturalistic setting can still be made, as smoking motivation has been shown to powerfully predict re lapse at proximal follow ups in clinical interventions (Allen et al., 2008; Cofta Woerpel et al., 2011; Heffer, Lee, Artaega, & Anthenelli, 2010; Powell, Dawkins, West, Powell, & Picking, 2010). Because the present research evaluated only the acute effects of exercise in a laboratory setting, it would be hasty to conclude that exercise is unlikely to be a beneficial adjunct to CM in a long term, outpatient smoking cessation intervention. In fact, I will argue subsequently that

PAGE 95

95 for various reasons, continued investigations of the efficacy of exercise based cessation interventions are likely to be highly worthwhile. Aside from the fact that smoking motivation and smoking have been shown to be related in naturalistic settings thus exercise plus CM may be effect ive in these settings there are several additional benefits that may be uniquely afforded by a combined intervention. For example, unlike other approaches to treatment (e.g., NRT, cognitive behavioral therapy), exercise reduces the risk of various smoking related diseases in both current and former smokers (e.g., Type 2 diabetes, systemic inflammation, arterial stiffness, dyslipidemia, low bone mineral density; Linke, Ciccolo, Ussher, & Marcus, 2013; Shaw, Shaw, & Brown, 2011; Singh, Jani, John, Singh, & Jo seley, 2011). Even in the absence of promoting complete abstinence, reducing smoking during an exercise based intervention has been shown to reduce inflammation (e.g., white blood cells) and increase cardiovascular fitness (e.g., maximum oxygen consumption [V O2 max]), and these effects are enhanced among participants who are most adherent to exercise recommendations (e.g., decrease in total cholesterol and ratio of total cholesterol:high density lipoprotein cholesterol; Korhonen et al., 2011). Even among th ose who do not reduce smoking, initiating a regular exercise regimen has been shown to reduce the risk of developing chronic obstructive pulmonary disease (COPD, Garcia Aymerich, Lange, Benet, Schnohr, & Anto, 2007). In addition to combatting smoking relat ed diseases, another unique benefit of exercise relative to other treatments is that exercise may function as a substitute reinforcer for smoking. Substitute reinforcers are behaviors or products that replace and decrease the likelihood of smoking (Green & Fisher, 2000). Although they are

PAGE 96

96 physically dissimilar to smoking, the behaviors may share some characteristics with smoking (e.g., both exercise and smoking are used for weight management, Byrne & Byrne, 1993). The possibility that exercise may function as a reinforcing substitute for smoking is supported by neuroimaging research indicating that exercise activates brain reward circuitry (e.g., mesolimbic regions) and releases neurotransmitters purportedly responsible for the subjective experience of smok ing reward (e.g., epinephrine, serotonin, dopamine; Bortz et al., 1981; Boecker, Sprenger, & Spilker, 2008). A treatment that provides a substitute for smoking is important, as research conducted by Goelz et al. (2014) showed that, among individuals enroll ed in a counseling plus nicotine patch treatment, abstinence at week 8 was associated with having significantly higher substitute reinforcers and significantly fewer complementary reinforcers (i.e., behaviors or stimuli that have been associated with smoki ng). One final benefit uniquely afforded by exercise interventions is that they might quitting smoking (Pomerleau et al., 2001; Cooper et al., 2006). In Pomerleau and Kerth (1996 ), for example, 75% of women and 35% of men reported that they were unwilling to gain more than 5 pounds (2.3 kg) during a quit attempt. In women under 25 years of age, 57% were unwilling to gain any weight at all. Since weight concerned individuals often terminate their quit attempts very early (or do not attempt to quit at all), interventions that postpone weight gain beyond the first few fragile days of cessation (e.g., exercise) may increase the number of weight concerned smokers who make a successful q uit attempt (Pomerleau, Pomerleau, Namenek, & Merhinger, 2000).

PAGE 97

97 Although a review by Ussher, Taylor, and Faulkner (2012) suggested that the evidence to support the efficacy of exercise based cessation interventions was limited, in that review was defined in terms of whether complete abstinence was achieved. However, research has shown that increasing exercise while continuing to smoke at a lower rate nonetheless offers health benefits (e.g., Korhonen et al., 2011), and that achi eving substantial smoking reductions may ultimately facilitate quitting (Gartner, Hall, & McNeil, 2010). These data suggest that perhaps new conceptions of treatment success are warranted for the 19% of individuals who continue to smoke despite a national leveling off of smoking rates. In other words, dismissing exercise interventions as ineffective because they do not promote cessation at proximal follow up points may be short sighted, as those individuals who merely reduce may quit smoking at a later date . According to Ussher et al. (2012), one major contributor to the limited support for the efficacy of exercise based cessation interventions is that adherence to exercise prescriptions is often low. However, data from several randomized clinical trials sug gests that among adherent participants, exercise is an effective treatment (or adjunct to treatment) to reduce smoking (e.g., Marcus et al.,2005; Prochaska et al., 2008; see Kaczynski, Manske, Mannell, & Grewal, 2008, for a review). Thus, one important dir ection for future research will be increasing adherence to exercise based cessation interventions. One method for doing so may be via the use of CM. Specifically, an intervention that offers participants incentives for achieving both exercise and smoking g oals may overcome the problem of adherence in exercise based cessation interventions. CM has been used to improve numerous health behaviors (e.g.,

PAGE 98

98 reduce alcohol intake, Alessi & Petry, 2013; increase glucose monitoring among diabetics, Raiff & Dallery, 20 11) including exercise (Kurti & Dallery, 2014) in naturalistic settings. In addition, recall that Weinstock et al. (2008) showed that participants who engaged in exercise related activities during an outpatient CM intervention to reduce drug use had better treatment outcomes than participants who did not exercise. Thus, the efficacy of a clinical intervention in which motivational incentives are provided contingent on meeting both exercise and smoking goals seems promising. Although Experiment 2 indicated that the effect of exercise on smoking was mediated through self reports of the anticipated, positive consequences of smoking, the lack of exercise effects on objective measures of smoking in Experiment 3 prevented us from conducting the mediation analysis necessary to replicate this finding. Related to this, we were also unable to assess whether discounting rates ( k ) mediated the relation between exercise and smoking. Although the failure to observe significant effects of exercise on discounting suggests t hat k is not likely to mediate this relation, effects of exercise on discounting may have been achieved with more intense exercise. For example, Bothe et al. (2013) showed that men who ran on a treadmill for 30 minutes at a moderate to high intensity were less sensitive to monetary rewards (i.e., exhibited less activation in brain reward areas in the presence of stimuli that signaled monetary gains) relative to men who performed placebo exercise (i.e., stretching). These data may suggest that higher intensi immediate sums of money in the discounting task and therein decrease impulsive

PAGE 99

99 choice. Thus, future assessments of the conditions under which exercise may influence discounting (e.g., exerci se intensity) will be worthwhile. In addition to assessing effects of exercise on temporal discounting, future research should evaluate effects of exercise on other types of discounting (e.g., probability, discounting of real as opposed to hypothetical re wards). Although the present work contextualized cigarette smoking in terms of temporal discounting because this framework directly lends itself to several treatment strategies, the framework is oversimplified in several ways. First, the reinforcers associ ated with abstaining (e.g., health, saving money) are not only delayed: They are also probabilistic. Although some researchers argue that the two are related (e.g., both can be described by a hyperbolic curve, Shead & Hodgins, 2009; rates of delay and prob ability discounting are correlated, Myerson et al., 2003), some manipulations have been shown to have different effects on delay versus probability discounting rates (e.g., smoking abstinence, Yi & Landes, 2012). Thus, it may be worthwhile for future resea rch to evaluate effects of exercise on discounting tasks that incorporate both delay and probability components. In addition to the fact that our theoretical framework for approaching cigarette smoking underemphasizes that the reinforcers associated with abstaining are both delayed and probabilistic, a second issue is that participants do not actually experience the delays associated with their choices (i.e., the consequences are not delivered). Dallery and Raiff (2007) state that, although several studies have indicated no outcomes (Johnson & Bickel, 2002; Lagorio & Madden, 2005), verbally mediated

PAGE 100

100 hypothetical choices may be less sensitive to certain manipulations than b ehavior based choices. Therefore, future research assessing effects of exercise manipulations on discounting may consider incorporating behavior based measures of impulsive choice (e.g., Cherek & Lane, 2000), in which behavior contacts the delays and conse quences associated with each choice. Another important direction for future research will be determining the predictive validity of laboratory based evaluations of potential smoking cessation treatments. Although laboratory based smoking studies reflect th e predictors of smoking abstinence identified in full scale clinical trials (e.g., latency to smoke after an abstinence period, Shadel et al., 2011; smoking reinforcement, Perkins et al., 2006; MacKillop et al., 2008), no research to date has assessed whet her smoking behavior in the laboratory predicts treatment outcome in a subsequent clinical intervention. However, given the substantial differences in laboratory versus naturalistic settings, it seems feasible that results might differ. For example, recent research conducted by Deiches, Baker, Lanza, and Piper (2013) suggested that antecedents to lapse during the first eight weeks of a quit attempt differed across individuals, where antecedents comprised particular symptoms, locations, social features, and activities associated with a given context. In addition, reported motivators for quitting consistently include the monetary cost of smoking, health concerns, and social concerns (McCaul et al., 2006), all of which are rarely modeled in labora tory based smoking studies. Because the laboratory rarely contains the momentary or situational factors that may influence smoking in naturalistic settings, research is needed to evaluate whether irrespective of these differences

PAGE 101

101 between the two contexts l aboratory studies have predictive validity for treatment outcomes in naturalistic settings. Importantly, despite unanswered questions about the predictive validity of the present results, laboratory based treatment assessments such as this one offer numero us benefits. For example, laboratory studies are efficient, cost and time effective, likely to reduce attrition, and they allow researchers to conduct parametric manipulations and experimentally control key variables (e.g., length of smoking deprivation, e xercise intensity; Chow, 1995; McKee, 2009; Kamboj et al., 2012; Shadel et al., 2011; Sweitzer et al., 2012). Additionally, researchers attempting to index the efficacy of smoking cessation pharmacotherapies in the laboratory emphasize that objective meas ures of smoking (e.g., latency to ad libitum smoking) may provide a reported craving or withdrawal, McKee, 2009; McKee, Weinberger, Shi, Tatrault, & Coppola, 2012; Perkins et al., 2008; Perkins, Stitzer, & Lerman, 2006; Perkins et al., 2010; Perkins & Lerman, 2011). Laboratory based studies have also shown that a programmed lapse after temporary abstinence is associated with a shorter delay to relapse in a naturalistic setti ng (Chornock, Stitzer, Gross, & Leischow, 1992; Juliano, Donny, Hoursmuller, & Stitzer, 2006; Shadel et al., 2011). Thus, although results of the present work cannot promise that a combined, exercise plus CM intervention will be highly effective in clinica l settings, the experimental control achieved in all three experiments, and the use of objective indices of smoking in two of the three experiments, represent a small step towards indexing clinical outcomes in laboratory based settings.

PAGE 102

102 Although the asses sment of a combined, behavioral approach to reducing smoking represents a substantial contribution to existing literature, this work had several limitations. First, there were substantially fewer females than males in all three experiments, which raises qu estions about the generalizability of our findings. Although none of our experiments indicated differences between males and females for any outcomes, it is possible that significant differences would have emerged had we enrolled equal numbers of males and females. Future research with a larger sample of females should be conducted to clarify questions about the generalizability of the present findings. A second limitation of this work was that 33.8% of the discounting data in Experiment 3 were dropped on account of not being systematic. One contributor to this may be that the task presented delays in a random order, which may have been confusing for participants. Participants completed the task at their intake session, thus they had one opportunity to prac tice to the task, but they were not provided with any feedback about their choices. Although the purpose of the randomly presented delays was to increase the likelihood of observing effects of the exercise manipulation, research in which delays were presen ted in a fixed order (i.e., from shortest to longest) has identified a smaller portion of the data as non systematic (e.g., 20% in Beck & Triplet, 2009). Unfortunately, presenting delays in an ascending order may also contribute to artificially elevated di scounting rates relative to presenting delays in a descending order (Hardisty, Thompson, Krantz, & Weber, 2013; Reid et al., 2009). Another way to promote data retention may be to use a simpler discounting task (e.g., the Monetary Choice Questionnaire, Ki rby, Petry, & Bickel, 1999), which requires

PAGE 103

103 participants to make a series of binary choices between two different amounts. However, simpler tasks necessarily have lower resolution, and have smaller effect size associations with health behaviors (MacKillop & Kahler, 1990). In the absence of consensus about an optimal discounting task, perhaps providing participants with more detailed instructions and requiring them to practice the task while providing feedback about their choices would have increased the por tion of systematic data. One final limitation of this work was that, in all three experiments, some portion of participants did not achieve their target, moderate intensity heart rates. As noted previously, research has suggested that low intensity exerci se is less effective at influencing smoking related variables than moderate intensity exercise (e.g., Daniel et below their targets in Experiment 3 relative to Exper iment 2, for example, may explain why moderate intensity exercise increased latencies to smoke in Experiment 2 but not in Experiment 3. Future research in which participants receive a small incentive for exercising within their target HR range for the full duration of the exercise session may increase the likelihood that they sustain their target HR goals. Inspired by a temporal discounting approach to the behavior of cigarette smoking, this dissertation represents the first assessment of a combined, exerc ise plus CM approach to smoking reduction. Although exercise reliably decreased smoking motivation across all three experiments and increased latencies to smoke in Experiment 2, it was no more effective at reducing smoking in Experiment 3 than CM alone. No netheless, the benefits uniquely offered by exercise based cessation interventions (e.g., combating smoking related disease and fears about weight gain), as well as the

PAGE 104

104 promising data that have emerged from clinical trials in which participants adhered to exercise prescriptions, suggest that the development of a clinical, exercise plus CM intervention to reduce smoking remains worthwhile. Moreover, it should be emphasized that smoking and inactivity often co occur, and can have additive, negative impacts on health (Garber et al., 2011). Because physical activity interventions are capable of both (1) promoting abstinence, and (2) preventing or delaying the onset of smoking related diseases even among individuals who do not achieve abstinence (Garber et al., 2 011; Garcia Aymerich et al., 2007), exercise based cessation interventions stand poised to contribute substantially to improving public health.

PAGE 105

105 APPENDIX HYPOTHETICAL TEMPORAL DISCOUNTING DATA The data in Table A 1 represent hypothetical raw values that c ould be collected from a typical, temporal discounting procedure. The columns represent the delays (in days) to receipt of a larger monetary sum ($100 in this case). For each delay, the large sum remains constant at $100, and the amount of a smaller, immed iately available sum (leftmost column) varies systematically across an ascending sequence (some experiments include both an ascending and descending sequence, and Experiment 3 used a random sequence). Hypothetical responses on each trial are indicated with a D (delayed sum selected) or an I (immediate sum selected). The amount at which participants ceased choosing the delayed sum and chose the immediate seem (i.e., the hig hlighted with an asterisk. For example, when $100.00 was delayed by 7 days, the hypothetical participant below forewent the immediate sum until he or she was offered $95.00 now, at which point a preference reversal occurred. When $100.00 was delayed by 14 then $80.00 now, respectively, and so on for the rest of the delays. Determination of k values and AUC values are displayed in Figure A 1 and Figure A 2, respect ively.

PAGE 106

106 Table A 1. Hypothetical indifference points from a typical, temporal discounting procedure. ____________________________________________________________________ Delay to large reward (in days) Amount of immediate reward 7 14 30 120 240 365 18 25 3650 ____________ 1 D D D D D D D D 2 D D D D D D D D 3 D D D D D D D D 4 D D D D D D D D 5 D D D D D D D I* 10 D D D D D D I* I 15 D D D D D D I I 20 D D D D D D I I 25 D D D D D I* I I 30 D D D D D I I I 35 D D D D D I I I 40 D D D D D I I I 45 D D D D I* I I I 50 D D D D I I I I 55 D D D D I I I I 60 D D D D I I I I 65 D D D I* I I I I 70 D D D I I I I I 75 D D D I I I I I 80 D D I* I I I I I 85 D D I I I I I I 90 D I* I I I I I I 95 I* I I I I I I I 100 I I I I I I I I ____________________________________________________________________________

PAGE 107

107 Figure A 1. Indifference points across successive delays. After graphing the indifference points, the data are fit with Equation 1 1 using least squares regression (LSR) to yield estimates of the slope of the curve ( k ). LSR is a procedure in which k is adjusted in an iterative fashion until the resultant k value is one that minimizes the sum of squared residuals (i.e., minimizes the sum of squared distances between each indifference point and the curve). When the data in Table A 1 were fit with Equation 1 1 in Exce l, the curve above resulted, for which k = .00581. 0 10 20 30 40 50 60 70 80 90 100 0 500 1000 1500 2000 2500 3000 3500 Present Value ($) Delay (weeks)

PAGE 108

108 Figure A 2. Calculation of the area under the discounting curve (AUC). To visualize how AUC is calculated, vertical lines have been drawn below that extend from each indifference point to the X ax is. AUC is calculated by summing the area contained by the adjacent trapezoids created between successive indifference points, and AUC values can range from 0 to 1. AUC for the present, hypothetical dataset = .1628 0 10 20 30 40 50 60 70 80 90 100 0 500 1000 1500 2000 2500 3000 3500 Present Value ($) Delay (weeks)

PAGE 109

109 LIST OF REFERENCES Ainslie, G. (1974). Impulsive control in pigeons. Journal of the Experimental Analysis of Behavior, 21, 485 489. doi: 10.1901/jeab.1974.21 485 Allen, S.S., Bade, T., Hatsukami, D., & Center, B. (2008). Craving, withdrawal, and smoking urges on days immediately prior to smokin g relapse. Nicotine & Tobacco Research, 10, 35 45. American College of Sports Medicine. Physical Activity and Public Health Guidelines. United States, 2007. Retrieved August 20, 2011 from http://www.acsm.org/ Arbour Nic itopoulos, K.P., Faulkner, G.E., Hsin, A., & Selby, P. (2011). A pilot study examining the acute effects of exercise on cigarette cravings and affect among individuals with serious mental illness. Mental Health and Physical Activity, 4, 89 94, doi: 10.1016 /j.mhpa.2011.06.001 Baker, F., Bickel, W.K., & Johnson, M.W. (2003). Delay discounting in current and never smokers: Similarities and differences across commodity, sign, and magnitude. Journal of Abnormal Psychology, 112, 382 392. doi: 10.1037/0021 843X.1 12.3.382 Beck, R.C., & Triplett, M.F. (2009). Test retest reliability of a group administered paper pencil measure of delay discounting. Experimental and Clinical Psychopharmacology, 17, 345 355. doi: 10.1037/a0017078 Bickel, W.K., Yi, R., Landes, R. D., Hill, P.F., & Baxter, C. (2011). Remember the future: Working memory training decreases delay discounting among stimulant users. Biological Psychiatry, 69, 260 265. Bisaga, A., Padilla, M., Garawi, F., Sullivan, M.A., & Haney, M. (2007). Effects of a lternative reinforcer and craving on the choice to smoke cigarettes in the laboratory. Human Psychopharmacology, 22, 41 47. Black, A.C., & Rosen, M.I. (2011). A money management based substance use treatment increases valuation of future rewards. Addictiv e Behaviors, 36, 125 128. Bock, B.C., Marcus, B.H., King, T.K., Borrelli, B., & Roberts, M.R. (1999). Exercise effects on withdrawal and mood among women attempting smoking cessation. Addictive Behaviors, 24, 399 410. doi: 10.1016/S0306 4603(98)00088 4 Bo mechanisms in the human brain. Cerebral Cortex, 18, 2523 2531.

PAGE 110

110 Borjeson, S., Hursti, T.J., Peterson, C., Fredikson, M., Furst, C.J., Avall Lundqvist, E., & Steinbeck, G. (1997) . Similarities and differences in assessing nausea on a verbal category scale and a visual analogue scale. Cancer Nursing, 20, 260 266. Bortz, W.M., Angwin, P., Mefford, I.N., Boarder, M.R., Noyce, N., & Barchas, J.D. (1981). Catecholamines, dopamine, a nd endorphin levels during extreme exercise. New England Journal of Medicine, 305, 466 467. Bothe, N., Zschucke, E., Dimeo, F., Heinz, A., Wustenberg, T., & Strohle, A. (2013). Acute exercise influences reward processing in highly trained and untrained me n. Medicine & Science in Sports & Exercise, Byrne, A., & Byrne, D.G. (1993). The effect of exercise on depression, anxiety, and other mood states: A review. Journal of Psychosomatic Research, 37, 565 574. Cappelleri, J.C., Bushmakin, A.G., Baker, C.L., Merikle, E., Olufade, A.O., & Gilbert, D.G. (2007). Multivariate framework of the brief questionnaire of smoking urges. Drug and Alcohol Dependence, 90, 234 242. doi: 10.1016/j.drugalcdep.2007.04.002 Centers for Disease Control and Prevention (CDC), 2012 . Current cigarette smoking among adults United States, 2011. MMWR 61, 889 894. Centers for Disease Control and Prevention (CDC), 2011. Quitting smoking among adults United States, 2001 2010. MMWR 60, 1513 1519. Cherek, D.R. & Lane, S.D. (2001). Acute ef fects of D fenfluramine on simultaneous measures of aggressive escape and impulsive responses of adult males with and without a history of conduct disorder. Psychopharmacology, 157, 221 227. Chow, S. (1995). In defense of experimental data in relativistic milieu. New Ideas in Psychology, 15, 259 276. Cofta Woerpel, L., McClure, J.B., Li, Y., Urbauer, D., Cinciripini, P.M., & Wetter, D.W. (2011). Early cessation success or failure among women attempting to quit smoking: Trajectories and volatility of urg e and negative mood during the first post cessation week. Journal of Abnormal Psychology, 120, 596 606. Cox, L.S., Tiffany, S.T., & Christen, A.G. (2001). Evaluation of the brief questionnaire of smoking urges (QSU brief) in laboratory and clinical sett ings. Nicotine and Tobacco Research, 3 , 7 16. doi: 10.1080/14622200124218 Critchfield, T.S., & Kollins, S.H. (2001). Temporal discounting: Basic research and the analysis of socially important behavior. Journal of Applied Behavior Analysis, 34, 101 122 . doi: 10.1901/jaba.2001.34 101

PAGE 111

111 Dallery, J., Glenn, I.M., & Raiff, B.R. (2007). An internet based abstinence reinforcement treatment for cigarette smoking. Drug and Alcohol Dependence, 86 , 230 238. doi: 0.1016/j.drugalcdep.2006.06.013 Dallery, J., & Raif f, B.R. (2007). Delay discounting predicts cigarette smoking in a laboratory model of abstinence reinforcement. Psychopharmacology, 190 , 485 496. doi: 10.1007/s00213 006 0627 5. Daniel, J.Z., Cropley, M., & Fife Schaw, C. (2007). Acute exercise effects on smoking withdrawal symptoms and desire to smoke are not related to expectation. Psychopharmacology, 195, 125 129. doi: 10.1007/s00213 007 0889 6 Daniel, J., Cropley, M., Ussher, M., & West, R. (2004). Acute effects of a short bout of moderate versus lig ht intensity exercise versus inactivity on tobacco withdrawal symptoms in sedentary smokers. Psychopharmacology , 174, 320 326. doi: 1007/s00213 003 1762 x Davies, G.M., Willner, P., Morgan, M.J. (2000). Smoking related cues elicit craving in factor structure of the Questionnaire of Smoking Urges. Psychopharmacology, 152, 332 342. doi: 10.1007/s002130000526 Deiches, J.F., Baker, T.B., Lanza, S., & Piper, M.E. (2013). Early lapses in a cessation attempt: Lapse contexts, cessation success, and predictors of early lapse. Nicotine & Tobacco Research, 15, 1883 1891. effects of tobacco cigarettes: Subject rated and physiological measures. Nicotine and Tobacco Research, 1, 317 324. doi: 10.1080/14622299050011441 Elibero, A., Janse Van Rensburg, K., & Drobes, D.J. (2011). Acute effects of aerobic exercise and hatha yoga on craving to smoke. Nicotine and Tobacco Resear ch. Published online first 17 August 2011. doi: 10.1093/ntr/ntr163 Everson, E.S., Daley, A.J., & Ussher, M. (2008). The effects of moderate and vigorous exercise on desire to smoke, withdrawal symptoms and mood in abstaining young adult smokers. Mental H ealth and Physical Activity, 1, 26 31. doi: 10.1016/j.mhpa.2008.06.001 Faulkner, G.E., Arbour Nicotopoulos, K.P., & Hsin, A. (2010). Cutting down one puff at a time: The acute effects of exercise on smoking behaviour. Journal of Smoking Cessation, 3, 136 141. doi 10.1375/jsc.5.2.141 Fjeldsoe, B.S., Miller, Y.D., & Marshall, A.L. (2013). Social cognitive mediators of the effect of the MobileMums intervention on physical activity. Health Psychology, 32, 729 738. doi: 10.1037/a0027548

PAGE 112

112 Garcia Aymerich, J ., Lange, P., Benet, M., Schnohr, P., & Antó, J. M. (2007). Regular physical activity modifies smoking related lung function decline and reduces risk of chronic obstructive pulmonary disease: a population based cohort study. American Journal of Respirator y and Critical Care Medicine , 175 , 458 463. Garber, C.E., Blissmer, B., Deschenes, M.R., Franklin, B.A., Lamonte, M.J., Lee, I.M., Nieman, D.C., & Swain, D.P. (2011). Quantity and quality of exercise for developing and maintaining cardiorespiratory, musc uloskeletal, and neuromotor fitness in apparently healthy adults: Guidelines for prescribing exercise. Medicine and Science in Sports and Exercise, 43, 1334 1359. Gartner, C., Hall, W., & McNeill, A. (2010). Harm reduction policies for tobacco. In T. Rhod es, & D. Henrich (Eds.), Harm reduction: Evidence, impacts and challenges (pp. 255 273). Germeroth, L.J., Wray, J.M., Gass, J., & Tiffany, S.T. (2013). Diagnostic utility of craving in predicting nicotine dependence: Impact of craving content and item sta bility. Nicotine & Tobacco Research, 15, 1988 1995. Giordano, L.A., Bickel, W.K., Loewenstein, G., Jacobs, E.A., Marsch, L., & Badger, G.J. (2002). Mild opioid deprivation increases the degree that opioid dependent outpatients discount delayed heroin and money. Psychopharmacology, 163, 174 182. Goelz, P.M., Audrain McGovern, J., Hitsman, B., Leone, F.T., Veluz Wilkins, A., changes in alternative reinforcers and short term smoking ces sation. Drug and Alcohol Dependence. Accessed online first at: http://www.sciencedirect.com/science/article/pii/S0376871614000623 Green, L. (1982). Self control behavior in animals . In V.L. Smith (Ed.), Research in Experimental Economics (Vol. 2, pp. 129 150).Greenwich, CT: JAI Press. Green, L., & Fisher, E.B.J. (2000). Economic substitutability: Some implications for health behavior. In: Bickel, W.K., Vuchinich, R.E. (Eds.), Ref raming Health Behavior Change with Behavior Economics, Mahwah, NJ, pp. 115 144. Groppelli, A., Giorgi, D. M., Omboni, S., Parati, G., & Mancia, G. (1992). Persistent blood pressure increase induced by heavy smoking. Journal of hypertension , 10 , 495 499. H afen, B.Q., & Hoeger, W.W.K. (1994). Wellness: Guidelines for a Healthy Lifestyle. Morton Publishing Co.: Englewood, CO. Haney, M. & Spealman, R. (2007). Controversies in translational research: Drug self administration. Psychopharmacology, 199, 403 419 .

PAGE 113

11 3 Hardisty, D.J., Thompson, K.F., Krantz, D.H., & Weber, E.U. (2013). How to measure time preferences: An experimental comparison of three methods. Judgment & Decision Making, 8, 236 249. Heatherton, T.F., Kozlowski, L.T., Frecker, R.C., & Fagerstrom, K. (1991). The Fagerstrom test for nicotine dependence: A revision of the Fagerstrom Tolerance Questionnaire. British Journal of Addiction, 86, 1119 1127. doi: 10.1111/j.1360 0443.1991.tb01879.x Heffner, J.L., Lee, T.C., Arteaga, C., & Anthenelli, R.M. (2010 ). Predictors of post Treatment relapse to smoking in successful quitters: Pooled data from two phase III varenicline trials. Drug and Alcohol Dependence, 109, 120 125. Hendricks, P.S., Ditre, J.W., Drobes, D.J., & Brandon, T.H. (2006). The early time course of Smoking withdrawal effects. Psychopharmacology, 187, 385 396. doi: 10.1007/s00213 006 0429 9 Higgins, S.T., Alessi, S.M., & Dantona, R.L. (2002). Voucher based incentives: A substance abuse treatment innovation. Addictive Behaviors, 27, 887 910. doi: 10.1016/S0306 4603(02)00297 6 Higgins, S.T., Heil, S.H., & Lussier, J. (2004). Clinical implications of reinforcement as a determinant of substance use disorders. Annual Review of Psychology, 55 , 431 46. doi:10.1146/annurev.psych.55.090902.1420 33 Iguchi, M.Y., Lamb, R.J., Belding, M.A., Platt, J.J., Husband, S.D., & Morral, A.R. (1996). Contingent reinforcement of group participation versus abstinence in a methadone maintenance program. Experimental and Clinical Psychopharmacology, 4, 315 321. doi 10.1037/1064 1297.4.3.315 Jaeschke, R., Singer, J., & Guyatt, G.H. (1990). A comparison of seven point and visual analogue scales: Data from a randomized trial. Controlled Clinical Trials, 11, 43 51, doi: 10.1016/0197 2456(90)90031 V Janse Van Rensbu rg, K., Elibero, A., Kilpatrick, M., & Drobes, D.J. (2013). Impact of aerobic exercise intensity on craving and reactivity to smoking cues. Experimental and Clinical Psychopharmacology, 21, 196 203. doi: 10.1037/a0032768 Janse Van Rensburg, K., & Taylor, A.H. (2008). The effects of acute exercise on cognitive functioning and cigarette cravings during temporary abstinence from smoking. Human Psychopharmacology, 23, 193 199. doi: 10.1002/hup.925 Janse Van Rensburg, K., Taylor, A., Hodgson, T., & Benattaya llah, A. (2009). Acute exercise modulates cigarette cravings and brain activation in response to smoking related images: An fMRI study. Psychopharmacology, 203, 589 598. doi: 10.1007/s00213 008 1405 3

PAGE 114

114 Janse Van Rensburg, K., Taylor, A., Hodgson, T., & B enattayallah, A. (2011). The effects of exercise on cigarette cravings and brain activation in response to smoking related images. Psychopharmacology. doi: 10.1007/s00213 011 2610 z Johnson, M.W., & Bickel, W.K. (2002). Within subject comparison of real a nd hypothetical money rewards in delay discounting. Journal of the Experimental Analysis of Behavior, 77, 129 146. Journal of Applied Behavior Analysis (2008) . The Special Issue on the Behavior Analysis and Treatment of Drug Addiction. Can be accessed on line at: http://seab.envmed.rochester.edu/jaba contingencies.html Kaczynski, A.T., Manske, S.R., Mannell, R.C., & Grewal, K. (2008). Smoking and physical activity: A systematic revi ew. American Journal of Health Behavior, 32, 93 110. Kamboj, S.K., Joye, A.,Das, R.K., Gibson, A., Morgan, C., & Curran H. (2012). Cue exposure and response prevention with heavy smokers: A laboratory based randomized placebo controlled trial examining t he effects of D cycloserine on cue reactivity and attentional bias. Psychopharmacology, 221, 273 284. Karvonen, M.J., Kentala, E., & Mustala, O (1957). The effects of training on heart rate: A longitudinal study. Annales Medicinae Experimentalis et Biolog iae, 35, 307 315. Katomeri, M. (2009). Acute effects of self paced walking on smoking withdrawal and cravings. PhD dissertation, University of Plymouth, UK. Retrieved from http: //search.proquest.com/docview/898757575?accountid=10920 K irby, K.N. (2009). One year temporal stability of delay discount rates. Psychonomic Bulletin and Review, 16, 3, 457 462. Kirby, K.N., Petry, N.M., & Bickel, W.K. (1999). Heroin addicts have high er discount rates for delayed rewards than non drug using controls. Journal of Experimental Psychology, 128, 78 87. doi: 10.1037/0096 3445.128.1.78 Korhonen, T., Goodwin, A., Miesmaa, P., Dupuis, E. A., & Kinnunen, T. (2011). Smoking cessation program wi th exercise improves cardiovascular disease biomarkers in sedentary women. Journal of Women's Health , 20 (7), 1051 1064. Lagorio, C.H., & Madden, G.J. (2005). Delay discounting of real and hypothetical rewards III: Steady state assessments, forced choice t rials, and all real rewards. Behavioual Processes, 69, 173 187. Lamb, R. J., Kirby, K. C., Morral, A. R., Galbicka, G., & Iguchi, M. Y. (2004). Improving contingency management programs for addiction. Addictive Behaviors , 29 , 507 523.

PAGE 115

115 Laraway, S., Snycers ki, S., Michael, J., & Poling, A. (2003). Motivating operations and terms to describe them: Some further refinements. Journal of Applied Behavior Analysis, 36, 407 414.doi: 10.1901/jaba.2003.36 407 Lee, E.M., Malson, J.L., Waters, A.J., Moolchan, E.T., & Pickworth, W.B. (2003). Smoking topography: Reliability and validity in dependent smokers. Nicotine and Tobacco Research, 5, 673 679. Linke, S.E., Ciccolo, J.T., Ussher, M., & Marcus, B.H. (2013). Exercise based smoking cessation interventions among wom en. 9, 69 84. Littlejohn, C. (2006). Does socio economic status influence the acceptability of, attendance for, and outcome of, screening and brief interventions for alcohol misuse: A review. Alcohol Alcohol, 41 , 540 545. Loewenstein, G. ( 1996). Out of control: Visceral influences on behavior. Organizational Behavior and Human Decision Processes, 65, 272 292, doi: 10.1006/obhd.1996.0028 Lussier, J.P., Heil, S.H., Mongeon, J.A., Badger, G.J., & Higgins, S.T. (2006). A meta analysis of vouc her based reinforcement therapy for substance use disorders. Addiction, 101, 192 203. doi: 10.1111/j.1360 0443.2006.01311.x MacKillop, J., Brown, C.L., Stojek, M.K., Murphy, C.M., Sweet, L., & Niaura, R.S. (2012). Behavioral economic analysis of withdraw al and cue elicited craving for tobacco: An initial investigation. Nicotine and Tobacco Research, 14, 1426 1434. doi: 10.1093/ntr/nts006 MacKillop, J. & Kahler, C.W. (2009). Delayed reward discounting predicts treatment response for heavy drinkers receiv ing smoking cessation treatment. Drug & Alcohol Dependence, 104, 197 203. MacKillop, J., Menges, D.P., McGeary, J., & Lisman, S.A. (2007). Effects of craving and DRD4 VNTR genotype on the relative value of alcohol: an initial human laboratory study. Behav ioral and Brain Functions, 3, 1 12. doi: 10.1186/1744 9081 3 11 MacKillop, J., Murphy, J.G., Ray, L., Eisenberg, D.T.A., Lisman, S.A., Lum, J.K., Wilson, D.S. (2008). Further validation of a cigarette purchase task for assessing the relative reinforcing e fficacy of nicotine in smokers. Experimental and Clinical Psychopharmacology, 16, 57 65. Madden, G.J., & Bickel, W.K. (2009). Impulsivity: The behavioral and neurological science of discounting. American Psychological Association. doi: 10.1037/12069 000

PAGE 116

116 & Abrams, D.B. (2005). The efficacy of moderate intensity exercise as an aid for smoking cessation in women: A randomized controlled trial. Nicotine & Tobacco Research, 6, 871 880. Mazur, J.E. (1987). An adjusting delay procedure for studying delayed reinforcement: The effect of delay and of intervening events on reinforcement value. In M.L. Commons, J.E. Mazur, & J. Anthony (Eds.), Quantitative analyses of behavior (pp. 55 73). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. McCaul, K.D., Hockemeyer, J.R., Johnson, R.J., Zetocha, K., Quinlan, K., & Glasgow, R.E. (2006). Motivation to quit using cigarettes: A review. Addictive Behaviors, 31, 42 56. McClure , S. M., Laibson, D. I., Loewenstein, G., & Cohen, J. D. (2004). Separate neural systems value immediate and delayed monetary rewards. Science , 306 , 5695, 503 507. McKay, J.R., Foltz, C., Stephens, R.C., Leahy, P.J., Crowley, E.M., & Kissin, W. (2005). P redictors of alcohol and crack cocaine use outcomes over a 3 year follow up in treatment seekers. Journal of Substance Abuse Treatment, 28 , S73 82. McKee, S.A. (2009). Developing human laboratory models of smoking lapse behavior for medication screening. Addiction Biology , 14 , 99 107. McKee, S.A., Krishnan effect of alcohol on smoking lapse behavior. Psychopharmacology, 189, 201 210. doi: 10.1007/s00213 006 0551 8 McKee, S.A., Weinberg er, A.H., Shi, J., Tetrault, J., & Coppola, S. (2012). Developing and validating a human laboratory model to screen medications for smoking cessation. Nicotine and Tobacco Research, 14, 1362 1371. doi: 10.1093/ntr/nts090 Michael, J. (1993). Establishing o perations. The Behavior Analyst, 16, 191 206. Morrison, K.L., Madden, G.J., Odum, A.L., Friedel, J.E., & Twohig, M.P. (2014). Altering impulsive decision making with an acceptance based procedure. Behavior Therapy. Available online first at: http://dx.doi.org/10.1016/j.beth.2014.01.001 Mueller, E.T., Landes, R.D., Kowal, B.P., Yi, R., Stitzer, M.L., Burnett, C.A. et al. (2009). Delay of smoking gratification as a laboratory model of relapse: Effects of incentives for not smoking, and relationship with measures of executive function. Behavioural Pharmacology, 20, 461 473. doi: 10.1097/FBP.0b013e3283305ec7

PAGE 117

117 Murphy, J.G., MacKillop, J., Tidey, J.W., Brazil, L.A., & Colby, S.M. (2011). Validity o f a demand curve measure of nicotine reinforcement with adolescent smokers. Drug and Alcohol Dependence, 113, 207 214. Myerson, J., Green, L., Hanson, J.S., Holt, D.D., & Estle, S.J. (2003). Discounting delayed and probabilistic rewards: Processes and tr aits. Journal of Economics and Psychology, 24, 619 635. Myerson, J., Green, L., & Warusawitharana, M. (2001). Area under the curve as a measure of discounting. Journal of the Experimental Analysis of Behavior, 76, 235 243. Ohmura, Y., Takahashi, T., Ki tamura, N., & Wehr, P. (2006). Three month stability of delay and probability discounting measures. Experimental and Clinical Psychopharmacology, 14, 318 328. doi: 0.1037/1064 1297.14.3.318 Packer, R.P., Howell, D.N., McPherson, S., & Roll, J.M. (2012 ). Investigating reinforcer magnitude and reinforcer delay: A contingency management analog study. Experimental and Clinical Psychopharmacology, 1 6. doi: 10.1037/a0027802 Perkins, K.A., Grobe, J., & Fonte, C. (1997). Influence of acute smoking exposure on the subsequent reinforcing value of smoking. Experimental and Clinical Psychopharmacology, 5, 277 285. doi: 10.1037/1064 1297.5.3.277 Perkins, K.A., Lerman, C., Stitzer, M.L., Fonte, C.A., Briski, J.L., Scott, J.A., & Chengappa, K.N.R. (2008). Develop ment of procedures for early screening of smoking cessation medications in humans. Nature, 84, 216 221. doi: 10.1038/clpt2008.30 Petry, N.M., & Simcic Jr., F. (2002). Recent advances in the dissemination of contingency management techniques: Clinical and research perspectives. Journal of Substance Abuse & Treatment, 23, 81 86. Perkins, K.A., Stitzer, M., & Lerman, C. (2006). Medication screening for smoking cessation: A proposal for new methodologies. Psychopharmacology, 184, 628 636. Pomerleau, C.S., & K urth, C.L. (1996). Willingness of female smokers to tolerate postcessation weight gain. Journal of Substance Abuse, 8, 371 378. Pomerleau, C. S., Pomerleau, O. F., Namenek, R. J., & Mehringer, A. M. (2000). Short term weight gain in abstaining women smok ers. Journal of Substance Abuse Treatment , 18 , 339 342. Pomerleau, O.F., Seherzer, H.H., Grunberg, N.E., Pomerleau, C.S., Judge, J., Fertig, J.B., & Burleson, J. (1987). The effects of acute exercise on subsequent cigarette smoking. Journal of Behavioral Medicine, 10, 1 11, doi: 10.1007/BF00846420

PAGE 118

118 Powell, J., Dawkins, L., West, R., Powell, J., & Pickering, A. (2010). Relapse to smoking during unaided cessation: Clinical, cognitive, and motivational predictors. Psychopharmacology (Berlin), 272, 537 549. P reacher, K.J., & Hayes, A.F. (2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods, Instruments, & Computers, 36, 717 731. doi: 10.3758/BF03206553 Prochaska, J.J., Hall, S.H., Humfleet, G., Munoz, R.F., Reus, V., Gorecki, J., & Hu, D. (2008). Physical activity as a strategy for maintaining tobacco abstinence: A randomized trial. Preventive Medicine, 47, 215 220. Rachlin, H., Raineri, A., & Cross, D. (1991). Subjective probability and d elay. Journal of the Experimental Analysis of Behavior, 55, 233 244. doi: 10.1901/jeab.1991.55 233 Radosevich, D.M., Wetzler, H., & Wilson, S.M. (1994). Health status questionnaire (HSQ) 2.0: Scoring comparisons and reference data. Bloomington, MN : He alth Outcomes Institute. Rash, C. J., Olmstead, T. A., & Petry, N. M. (2009). Income does not affect response to contingency management treatments among community substance abuse treatment seekers. Drug and Alcohol Dependence , 104, 249 253. Reid, A.K., Sh ahan, T.A., Grace, R.C., Robies, E., Vargas, P.A., & Bejerano, R. (2009). Within subject differences in degree of delay discounting as a function of order of presentation of hypothetical cash rewards. Behavioural Processes, 81, 260 263. Reeser, K.A. (198 3). The effects of repeated aerobic and non aerobic exercise on Reynolds, B. (2006). A review of delay discounting research with humans: Relations to drug use and gambling. Behavioural Pharmacology, 17, 651 667. doi: 10.1097/FBP.0b013e3280115f99 Roberts, V., Maddison, R., Simpson, C., Bullen, C., & Prapavessis, H. (2012). The acute effects of exercise on cigarette cravings, withdrawal symptoms, affect, and smoking behaviour: Systematic review update and meta analysis. Psychopharmacology, 222, 1 15. doi: 10.1007/s00213 012 2731 z Roll, J.M., Higgins, S.T., & Badger, G.J. (1996). An experimental comparison of three different schedules of reinforcement of drug abstinence using cigarette s moking as an exemplar. Journal of Applied Behavior Analysis, 29 , 495 505. doi 10.1901/jaba.1996.29 495 . Romanowich, P., & Lamb, R.J. (2010). Effects of escalating and descending schedules of incentives on cigarette smoking in smokers without plans to quit . Journal of Applied Behavior Analysis, 43, 357 367.

PAGE 119

119 Rosenberg, H. (2009). Clinical and laboratory assessment of the subjective experience of drug craving. Clinical Psychology Review, 29, 519 534. Rosenthal, T.L., Edwards, N.B., Ackerman, B.J., Knott, D.H ., & Rosenthal, R.H. (1990). Substance abuse patterns reveal contrasting personality traits. Journal of Substance Abuse, 2, 255 263. doi: 10.1016/S0899 3289(05)80060 4 Sayette, M.A., Martin, C.S., Wertz, J.M., Shiffman, S., & Perrott, M.A. (2001). A mul ti dimensional analysis of cue elicited craving in heavy smokers and tobacco chippers. Addiction, 96, 1419 1432. doi: 10.1046/j.1360 0443.2001.961014196.x Scerbo, F., Faulker, G., Taylor, A., & Thomas, S. (2010). Effects of exercise on craving to smoke: The role of exercise intensity and cortisol. Journal of Sports and Science, 28, 11 19. doi: 10.1080/02640410903390089 Schuh, K.J., & Stitzer, M.L. (1995). Desire to smoke during spaced intervals. Psychopharmacology, 120, 289 295. doi: 10.1007/BF02311176 Shadel, W.G., Martino, S.C., Setodji, C., Cervone, D., Witkiewitz, K., Beckjord, E.B., Scharf, D., & Shih, R. (2011). Lapse induced surges in craving influence relapse in adult smokers: An experimental investigation. Health Psychology, 30, 588 596. Shaw, I., Shaw, B.S., & Brown, G.A. (2011). Concurrent training and pulmonary function in smokers. International Journal of Sports Medicine, 32, 776 780. Shead, N.W., & Hodgins, D.C. (2009). Probability discounting of gains and losses: Implications for risk attitudes and impulsivity. Journal of the Experimental Analysis of Behavior, 71, 121 143. Sigmon, S.C., Lamb, R.J., & Dallery, J. (2008). Contingency management for reducing tobacco use. In: Higgins, S.T., Silverman, K., & Heil, S.H. (Eds.) Contingency M anagement in the Treatment of Substance Use Disorders: A Science based Treatment Innovation. New York: The Guilford Press: pp. 99 119. abstinence through voucher based reinfo rcement therapy. Drug and Alcohol Dependence, 41, 157 165. Doi: 0.1016/0376 8716(96)01246 X Singh, V.P., Jani, H., John, V., Singh, P., & Joseley, T. (2011). Effects of upper body resistance training on pulmonary functions in sedentary male smokers. Lung India, 28, 169 173. Stitzer, M.L., & Bigelow, G.E. (1984). Contingent reinforcement for carbon monoxide reduction: within subject effects of pay amount. Journal of Applied Behavior Analysis, 17 , 477 483. doi: 0.1901/jaba.1984.17 477

PAGE 120

120 Sweitzer, M.M., Deling er, R.L., & Donny, E.C. (2012). Dependence and withdrawal Induced craving predict abstinence in an incentive based model of smoking relapse. Nicotine and Tobacco Research, Published online first April 17, 2012. Taylor, A.H., & Katomeri, M. (2007). Walking reduces cue elicited cigarette cravings and withdrawal, and delays ad libitum smoking. Nicotine and Tobacco Research , 9, 1183 1190. doi:10.1080/14622200701648896 Taylor, A.H., Katomeri, M., & Ussher, M. (2005). Acute effects of self paced walking on urg es to smoke during temporary smoking abstinence. Psychopharmacology, 181, 1 7, doi: 10.1007/s00213 005 2216 4 Taylor, A.H., Katomeri, M., & Ussher, M. (2006). Effects of walking on cigarette cravings circumplex model. Journal of Sport and Exercise Psychology, 28, 18 31. Thayer, R., Peters, D., Takahashki, P., & Birkhead Flight, A. (1993). Mood and behavior (smoking and sugar snacking) following moderate exercise: A partial test of self regulation t heory. Personality and Individual Differences, 14 , 97 104. doi: 10.1016/0191 8869(93)90178 6 Thompson, T. T. (2007). Relations among functional systems in behavior analysis. Journal of the Experimental Analysis of Behavior , 87, 423 440. 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. Toll, B.A., Katulak, N.A., & McKee, S.A. (2006). Investigating the structure of the Questionnaire of Smoking U rges Brief (QSU Brief). Addictive Behaviors, 31, 1231 1239. doi: 10.1016/j.addbeh.2005.09.008 Toll, B.A., McKee, S.A., Krishnan factor structure of the Questionnaire on Smoking Urges. Psychological Ass essment, 16, 391 395. Ussher, M., Cropley, M., Playle, S., Mohidin, R., & West, R. (2009). Effect of isometric exercise and body scanning on cigarette cravings and withdrawal symptoms. Addiction, 104, 1251 1257, doi: 10.1111/j.1360 0443.2009.02605.x Ussh er, M. H., Taylor, A., & Faulkner, G. (2012). Exercise interventions for smoking cessation. Cochrane Database Syst Rev , 1 . van Laerhoven, H. & van der Zaag Loonen, H.J. (2004). A comparison of Likert scale and visual analogue scales as response options i Acta Paediatrica, 93, 830 835, doi: 10.1111/j.1651 2227.2004.tb03026.x

PAGE 121

121 Weinstock, J., Barry, D., & Petry, N.M. (2008). Exercise related activities are associated with positive outcome in contingency management treatment for substance use disorders. Addictive Behaviors, 2008, 1072 1075. doi: 10.1016/j.addbeh.2008.03.011 Williams, D.M., Dunsiger, S., Whitley, J.A., Ussher, M.H., Ciccolo, J.T., & Jennings, E.G. (2011). Acute effects of moderate intensity aerobic exercise on affective withdrawal symptoms and cravings among women smokers. Addictive Behaviors, 36, 894 897, doi: 10.1016/j.addbeh.2011.04.001 Willner, P., Hardman, S., & Eaton, G. (1995). Subjective and behavioural evaluation of cigarette cravings. Psychopharmacol ogy, 118, 171 177. Yi, R., & Landes, R.D. (2012). Temporal and probability discounting by cigarette smokers following acute smoking abstinence. Nicotine and Tobacco Research, 14, 547 558. Yoon, J.H., & Higgins, S.T. (2008). Turning k on its head: Comments on use of an ED50 in delay discounting research. Drug and Alcohol Dependence, 95, 169 172.

PAGE 122

122 BIOGRAPHICAL SKETCH A llison Kurti graduated from Wake Forest University in 2008 with a B.A. in psychology. Her undergraduate honors thesis, conducted under th e mentorship of Dr. Robert C. Beck, assessed the effect of subliminal affective primes on temporal discounting. Following graduation, she earned an M.S. in experimental psychology at Villanova University in 2010, where she primarily studied effects of dru gs on time ventured to the University of Florida to pursue her Ph.D. in psychology (concentration in ogy laboratory. Between Villanova and the University of Florida, Allison earned several academic and Thesis Award, the College of Liberal Arts & Sciences Dissertation Resear ch Award, the Behavior Analysis Research Award, and the Behavior Analysis Gerber Award. She earned her Ph.D. in August 2014, after which she continued to conduct research in behavior analysis and behavioral pharmacology as a National Institute of Health p ostdoctoral fellow at the University of Vermont, under the mentorship of Dr. Stephen Higgins.