EFFECTS OF RHYTHMIC CONTEXT ON TIME PERCEPTION IN INDIVIDUALS WITH PARKINSON DISEASE

Nathaniel S. Miller

A Dissertation

Submitted to the Graduate College of Bowling Green State University in partial fulfillment of The requirements for the degree of

DOCTOR OF PHILOSOPHY

December 2010

Committee:

J. Devin McAuley, Advisor

Dale Klopfer

Sheryl Coombs

Alexander Goberman Graduate Faculty Representative ii

ABSTRACT

J. Devin McAuley, Advisor

Parkinson disease (PD) is a neurodegenerative disorder of the basal ganglia (BG) that results in a significant loss of dopaminergic neurons. Previous studies have shown that individuals with PD show impairments in both the perception and production of duration, supporting the involvement of the BG and dopamine (DA) in perceptual and motor timing. One such DA-dependent timing impairment is gravitation in the remembered duration of an isolated (single) time interval toward the mean of a set of experienced time intervals (Malapani, et al., 1998). This dissertation extends research on time perception in PD to an investigation of the effects of rhythmic context on perceived duration. The basis for the project is a paradigm previously shown to produce large effects of rhythmic context on perceived duration in young adults (Barnes & Jones, 2000;

McAuley & Jones, 2003). In this paradigm, participants are asked to compare the duration of two empty time intervals marked by pairs of tones (a fixed standard interval followed by a variable comparison interval) with the instruction to ignore a preceding tone sequence (i.e., context rhythm). Previous studies have shown that participants are unable to ignore the context rhythm, as evident by relative duration judgments about the standard-comparison pair of intervals being more accurate when the tone marking the end of the standard interval is ‗on time,‘ relative to a periodic extrapolation of the context rhythm, than when the tone marking the end of the standard interval is ‗early‘ or ‗late.‘ The resulting ∩-shaped pattern of performance has been termed a temporal expectancy profile. Two experiments tested two hypotheses about the strength of the expectancy profiles in young adults, older adults, and individuals with PD. The first hypothesis is a period-correction hypothesis that posits that increasing the number of repetitions of the standard interval will eliminate (or at least weaken) the expectancy profile. The second iii hypothesis is a DA-mediated distortion hypothesis that posits that larger effects of rhythmic context will be observed when individuals with PD are tested off their DA-enhancing medication than when tested on medication. Moreover, the loss of DA in individuals with PD will result in larger effects of rhythmic context than observed with older adult controls or young adults.

Experiment 1 tested young adults in order to provide a baseline measure of performance and to provide an initial evaluation of the period-correction hypothesis, whereby effects of rhythmic context on perceived duration were predicted to be eliminated (or at least weakened) by increasing the number of repetitions of the standard interval. Experiment 2 compared the performance of individuals with PD (both on- and off-medication) to older adult controls and to the young adults tested in Experiment 1. In general, support for the period-correction hypothesis was found. Increasing the number of equal standard intervals reduced the effects of rhythmic context on perceived duration in all groups. Mixed support was found for the DA-mediated distortion hypothesis. Although no group differences were observed in the strength of expectancy profiles for the one-standard interval condition, increasing the number of equal standard intervals differentially affected individuals with PD on- and off-medication. Increasing the number of equal standard intervals served to weaken expectancy profiles for PD participants when they were on medication, but not when the same participants were off their medication.

Taken together, the present findings provide the first evidence for the potential involvement of

DA in period-correction processes.

iv

This dissertation is dedicated to the memory of Eva J. and John C. Miller. I am forever grateful for all that I have learned from you.

v

ACKNOWLEDGMENTS

Before I acknowledge everyone who has helped me throughout this dissertation, I must first thank my family for their love and support. Roger Miller and Kathy Carns, thank you for supporting me in all of my life endeavors. John and Eva Jean Miller, you have influenced my life in ways that I am only just beginning to realize. While I miss you more every day, I consider myself fortunate for all the great years we had together. Finally, thank you Michelle Renzetti for being my ‗sister‘ and putting up with me since 8th grade.

Special acknowledgement is also due to both Drs. J. Devin McAuley and Alexander

Goberman. Devin has provided invaluable mentoring and support throughout my graduate training, for which I am forever thankful. Alex took me under his wing, introduced me to the

Northwest Ohio Parkinson community, and taught me about conducting research on Parkinson disease. I am forever indebted to his continual support and willingness to listen and share ‗wild‘ testing stories. I must also thank my other committee members, Drs. Dale Klopfer and Sheryl

Coombs, for their ideas and comments that have shaped this dissertation. Finally, Drs. Laura

Dilley and Mary Hare must also be thanked for their advice and valuable comments throughout my graduate training.

I have been fortunate to meet so many great people in Bowling Green who have provided amazing levels of support during graduate school. Sarah Domoff, you have been my rock throughout everything. I appreciate your understanding, sacrifices, and undying support throughout this dissertation—I owe you big time. I must also thank Molly Henry, Kavita Desai,

Mo Wang, Guyla Davis, and Eileen Delaney for their continuing friendship. Thank you for your support and willingness to listen; I cannot imagine what graduate school would have been like had you six not been there for me. vi

I would also like to acknowledge both the Bowling Green State University and Michigan

State University branches of the Timing, Attention, and Perception Lab, who have contributed helpful comments and outstanding support during several research projects. Additionally, I must thank everyone who has helped with various stages of this dissertation: Jessica Zavadil, Shanta'

Coleman, Claire Hoover, Daniel Percival, Alicia Cornell and Kelly Kaltenbach–you have made my life much easier and the time spent organizing these datasets much shorter.

I must also extend appreciation to the participants of Experiment 2. Not only did they give me hours of their time, but also they were more than happy to have a relative stranger come into their home in the ‗wee‘ hours of the morning with several carloads of testing equipment. I am especially grateful to the members of the Parkinson Foundation of Northwest Ohio who have provided tireless assistance and support throughout this project and others.

vii

TABLE OF CONTENTS CHAPTER 1: INTRODUCTION ...... 1

CHAPTER 2: BACKGROUND ON PARKINSON DISEASE ...... 9

Description and Classification ...... 9

Pathophysiology of PD ...... 13

Treatments for PD ...... 16

Pharmacological Therapies ...... 17

Levodopa...... 17

DA agonists...... 19

Catechol-O-methyl transferase inhibitors...... 20

Monoamine oxidase B inhibitors...... 20

Amantadine...... 20

Anticholinergics...... 21

Pharmacological therapies for non-motor symptoms of PD...... 21

Ablative Surgeries ...... 22

Deep Brain Stimulation...... 23

CHAPTER 3: PERCEPTION AND PRODUCTION OF TIME INTERVALS—THEORY,

DATA, AND NEURAL BASES ...... 28

Theories of Timing ...... 28

Interval Theories ...... 28

Entrainment Theories ...... 32

Theory of the Neural Bases of Timing ...... 34

Data ...... 35 viii

Critical Issues in the Timing Literature ...... 36

Psychophysics of time...... 36

Timescale...... 37

General measures of time perception and production...... 38

Time perception and production tasks...... 39

Production of Time Intervals ...... 42

Perception of Time Intervals ...... 48

Neural Bases of Duration Perception and Production ...... 55

Critical Issues in Determining the Neural Bases of Timing ...... 57

Timescale...... 57

Time perception and production tasks...... 59

Neural Mechanisms Recruited for the Production of Time Intervals ...... 61

Neural Mechanisms Recruited for the Perception of Time Intervals ...... 64

CHAPTER 4: EFFECTS OF AGING AND PD ON THE PERCEPTION AND PRODUCTION

OF DURATION...... 69

Aging and Timing ...... 69

Theoretical Perspectives on Aging and Timing ...... 70

Data of Aging and Timing ...... 72

Production of time intervals...... 72

Perception of time intervals...... 78

PD and Timing ...... 83

Theoretical Perspectives on Timing in PD ...... 83

Data of Timing in PD ...... 84 ix

Production of time intervals...... 84

Perception of time intervals...... 96

Disease Severity...... 99

Heterogeneity of performance in PD...... 99

Role of DA in timing...... 100

Effects of deep brain stimulation on timing...... 102

CHAPTER 5: EXPERIMENT 1—EFFECT OF RHYTHMIC CONTEXT ON DURATION

DISCRIMINATION IN YOUNG ADULTS ...... 105

Experiment Overview ...... 105

Method ...... 105

Design ...... 105

Participants ...... 106

Equipment ...... 106

Stimuli ...... 106

Procedure ...... 107

No-context condition...... 107

With-rhythmic-context condition...... 110

Predictions...... 111

Data Analysis ...... 112

Results ...... 112

CHAPTER 6: EXPERIMENT 2—EFFECT OF RHYTHMIC CONTEXT ON DURATION

DISCRIMINATION IN OLDER ADULTS AND INDIVIDUALS WITH PD ...... 116

Experiment Overview ...... 116 x

Method ...... 117

Design ...... 117

Participants ...... 117

Equipment ...... 119

Stimuli ...... 119

Procedure ...... 120

Predictions...... 123

No-context condition...... 123

With-rhythmic-context condition...... 124

Data Analysis ...... 125

Results ...... 126

Comparison of Individuals with PD to Older-Adult Controls and Young Adults

...... 126

Comparison of Individuals with PD On- and Off-medication ...... 131

PD severity ratings...... 133

PD characteristics and duration discrimination...... 134

Comparison of Spontaneous Motor Tempo in Individuals with PD to Older Adult

Controls ...... 135

CHAPTER 7: GENERAL DISCUSSION ...... 137

APPENDIX A: REFERENCES ...... 150

APPENDIX B: FOOTNOTES ...... 183

APPENDIX C: DESCRIPTION OF SCREENING MEASURES ...... 187 xi

Dementia Rating Scale (DRS-2) ...... 187

Morningness-Eveningness Questionnaire (MEQ) ...... 187

Center for Epidemiologic Studies Depression scale (CES-D) ...... 188

Unified Parkinson Disease Rating Scale (UPDRS) ...... 188

Self-Reports of PD Symptom Severity and Medication Effectiveness ...... 188

xii

LIST OF FIGURES/TABLES

Figure/Table Page

1 Schematic of the duration-discrimination with rhythmic context task ...... 4

2 Schematic of the quadratic- and flat-expectancy profiles ...... 6

3 Schematic of a proposed model of the basal ganglia in normal motor control and

Parkinsonism……………………………………………………………………….. 14

4 Schematic of levodopa/carbidopa effectiveness over time ...... 18

5 Schematic of target sites for ablative and deep-brain stimulation surgeries ...... 27

6 Schematic of Scalar Expectancy Theory ...... 29

7 Schematic of a prototypical entrainment theory ...... 34

8 Schematic of the isolated-interval discrimination and tempo-discrimination tasks .. 50

9 Schematic of the tone sequences for the no-context condition ...... 108

10 Schematic of the tone sequences for the with-context-rhythm condition ...... 109

11 Proportion of correct responses (PC) as a function of context condition for the

one- and five-standard interval sequences ...... 113

12 PC as a function of standard ending for the one- and five-standard interval

sequences…...... 114

1 Participant characteristics for individuals with PD ...... 118

2 Participant characteristics for older adult controls ...... 119

13 Diagram of testing epochs for individuals with PD and older adult controls ...... 122

14 PC as a function of context condition and group ...... 127

15 PC as a function of standard ending and group ...... 130

xiii

16 PC as a function of context condition, number of standard intervals, and

medication state for individuals with PD ...... 131

17 PC as a function of standard ending, number of standard intervals, and

medication state for individuals with PD ...... 132

1

CHAPTER 1: INTRODUCTION

The correct performance of any behavior requires that the behavior is performed in the right manner, and critically, at the right time. Both human and non-human animals must time their behavior across a wide range of timescales. For example, catching a ball requires precise timing in the range of milliseconds to determine the time of contact for both the ball and one‘s hand. Similarly, the timing of longer durations, spanning seconds – minutes, is required for knowing how much time we have to cross the street before the ‗walk‘ signal changes and cooking the perfect omelet.

A central question in the research literature on timing is how such events are timed. Many researchers agree that some form of a ‗clock‘ is used to time these events. One basic question that can be asked about this clock is the accuracy with which it times. Specifically, does the subjective clock accurately measure objective time? Answers to this question provide some insight into the nature of the clock.

Studies of time perception reveal that in many instances the perceived duration of an event closely matches its actual physical duration (Allan, 1979; Getty, 1975; Stevens, 1957).

Nonetheless, the human time sense is also subject to systematic distortions; underestimating the duration of an event can lead to the sense that ―time flies when you are having fun‖ while overestimating the duration of an event can lead to the sense that ―a watched pot never boils.‖ A number of different factors have been found to affect the accuracy with which people make temporal judgments about isolated events, such as a stoplight, and temporally extended sequences, such as those found in music and speech. Some factors that have been shown to produce systematic effects on temporal judgments include the level of attention to duration (see

2

Lustig, 2003 for a review), levels of stress or relaxation (Boltz, 1994), and the general temporal context that the judgment is being made in (Jones & McAuley, 2005; McAuley & Miller, 2007).

Distortions in time perception are also common in a number of different age-related neurological disorders and diseases, including Parkinson disease (PD), Alzheimer‘s disease, and dementia (Harrington, Haaland, & Knight, 1998; Hibbard, Migliaccio, Goldstone, & Lhamon,

1975; Malapani, et al., 1998; Nichelli, Venneri, Molinari, Tavani, & Grafman, 1993; Papagno,

Allegra, & Cardaci, 2004), but the nature and scope of these impairments are not well understood. The focus of this dissertation is on time perception in PD.

PD is a neurodegenerative disorder that results in a severe loss of dopaminergic cells in the substantia nigra, or more generally, within the basal ganglia (BG). While PD is characterized by motor symptoms, such as slowness of movement and tremor, a variety of non-motor issues, such as impairments in time perception, have also been reported (Grahn & Brett, 2005;

Harrington, Haaland, & Hermanowicz, 1998). The non-motor issues of PD are, comparatively, less well understood, but have also been attributed to the disordered BG or dopaminergic system

(Bodis-Wollner, 2003; Malapani, et al., 1998; Pastor, Artieda, Jahanshahi, & Obeso, 1992;

Pastor, Jahanshahi, Artieda, & Obeso, 1992; Richard, Schiffer, & Kurlan, 1996; Stein, Heuser,

Juncos, & Uhde, 1990).

The goals of this dissertation are twofold. The first goal of this dissertation is to better understand the mechanisms involved in time perception, specifically the BG and dopamine (DA)

(Grahn & Brett, 2009; Malapani, et al., 1998; Pastor, Artieda, et al., 1992; Pastor, Jahanshahi, et al., 1992). The second goal of this dissertation is to better understand potential impairments in time perception in individuals with PD. At least two potential outcomes stem from understanding these impairments in PD. First, they may provide insight into how current, rhythm-based,

3 therapies that are used to treat PD-related gait problems can be improved. Second, they may serve as an early perceptual marker or a diagnostic tool for PD. The perspective taken in this dissertation is that both goals are complementary and can inform current knowledge of both the neural bases of timing and auditory impairments in PD.

Several studies have provided evidence of timing impairments in PD, but support for such an impairment is mixed (Grahn & Brett, 2009; Harrington, Haaland, & Hermanowicz,

1998; Ivry & Keele, 1989; Malapani, et al., 1998; Riesen & Schnider, 2001; Smith, Harper,

Gittings, & Abernethy, 2007; Wearden, et al., 2008). Impairments have been reported for both isolated intervals and rhythmic sequences of time intervals (Grahn & Brett, 2009; Harrington,

Haaland, & Hermanowicz, 1998; Malapani, et al., 1998; Riesen & Schnider, 2001; Smith, et al.,

2007; Wearden, et al., 2008) and occur across a range of durations that span milliseconds – seconds (Harrington, Haaland, & Hermanowicz, 1998; Malapani, et al., 1998; Riesen &

Schnider, 2001; Smith, et al., 2007). While some evidence suggests that DA-enhancing medications, which are taken to treat the motor symptoms of PD, ameliorate time production impairments (Elsinger, et al., 2003; Malapani, et al., 1998; O'Boyle, Freeman, & Cody, 1996;

Pastor, Jahanshahi, et al., 1992), it is less clear whether DA-enhancing medications ameliorate time perception impairments. Only two studies have investigated the effect of DA-enhancing medications on time perception in individuals with PD, reporting no effect of medication state on these impairments (Harrington, Haaland, & Hermanowicz, 1998; Wearden, et al., 2008).

However, a number of factors that might influence the effects of DA-enhancing medication on time perception have yet to be investigated.

Context Rhythm Standard Interval Comparison Interval (IGNORE) Standard

Ending

Context IOI Pre-standard Gap Pause 2x Context IOI

Time

Figure 1. Schematic of the duration-discrimination with rhythmic context paradigm. Circles represent tones. Listeners are presented with an isochronous tone sequence, termed the context rhythm, that precedes a standard-comparison interval pair. The key manipulation is the final tone that marks the duration of the standard interval. The standard ending is varied in time to occur early, on time, or late relative to the context rhythm. Listeners are instructed to ignore the context rhythm and judge the duration of the

4

comparison interval relative to the standard interval, responding ‗shorter,‘ ‗same,‘ or ‗longer.‘

5

This dissertation investigates the perception of time in individuals with PD by examining the effects of rhythmic context on the perceived duration of a time interval. The basis for this study is a paradigm previously shown to produce large effects of rhythmic context on perceived duration in young adults (Barnes & Jones, 2000; McAuley & Jones, 2003). In this paradigm, participants are asked to compare the duration of two time intervals marked by pairs of tones (a fixed standard time interval followed by a variable comparison time interval) with the instruction to ignore a preceding tone sequence (i.e., context rhythm). The critical manipulation in this task is the onset of the final tone that marks the duration of the standard interval (Figure 1). The onset of this tone is varied in time so that the standard ending occurs either ‗early,‘ ‗on time,‘ or ‗late‘ relative to a periodic extrapolation of the context rhythm.

If listeners are able to ignore the context rhythm when making duration judgments between the standard and comparison, then equivalent proportion of correct responses (PC) should obtain for the standard ending conditions; this pattern of results has been termed a flat expectancy profile (Figure 2). However, studies using this task report a consistent ∩-shaped pattern of correct responses associated with early, on-time, and late standard endings (Barnes &

Jones, 2000; Large & Jones, 1999; McAuley & Jones, 2003). This pattern of performance has been termed a temporal expectancy profile. The expectancy profile suggests that the context rhythm leads participants to generate expectancies about when the standard ending should occur, with unexpectedly timed endings resulting in systematic distortions in the perceived duration of the standard interval.

This dissertation will test two hypotheses regarding the effects of rhythmic context on perceived duration. The period-correction hypothesis posits that the effects of rhythmic context on perceived duration should be eliminated (or at least weakened) with multiple repetitions of a

6 standard interval. Multiple standards intervals were predicted to afford a more accurate percept of the standard interval, thereby eliminating the effects of rhythmic context. While ∩-shaped expectancy profiles are observed with one standard interval, increasing the number of repetitions of a standard interval should tend to flatten expectancy profiles.

Proportion Correct

Early On Time Late Standard Ending

Figure 2. Schematic of an expectancy profile. Solid bars represent the ∩-shaped temporal expectancy profile. This profile is marked by higher PC for standard intervals that end on-time, relative to a periodic extrapolation of the context rhythm, and lower PC for early or late standard endings (Barnes & Jones, 2000; McAuley & Jones, 2003). The dotted line represents a flat expectancy profile with equal PC for early, on-time and late standard endings, which suggests that the context rhythm does not have an effect on perceived duration.

Based on the proposed role of DA in time perception, the DA-mediated distortion hypothesis posits that the effects of rhythmic context on perceived duration are modulated by

DA. This hypothesis predicts stronger expectancy profiles for individuals with PD when tested off DA-enhancing medication, compared to on medication. Moreover, due to the loss of dopaminergic cells in PD, stronger expectancy profiles are predicted for individuals with PD, compared to older adult controls and young adults.

7

The remainder of this dissertation is organized as follows. Chapter 2 provides additional background information on PD. An overview of PD is provided, along with descriptions of PD- severity scales, and non-motor issues associated with PD. The pathophysiology of PD and current treatments are also discussed. The purpose of this chapter is to provide the reader with background knowledge of PD that will assist with the interpretation of the data discussed in

Chapters 4, 6, and 7.

Chapter 3 provides a review of the general timing literature. Included in this review is an overview of influential theories of timing, followed by an evaluation of these theories based on empirical data spanning the perception and production of isolated time intervals and sequences of time intervals. Finally, relevant research on the neural bases of timing is reviewed.

Chapter 4 reviews normal age-related changes that occur in the perception and production of time intervals and distinguishes those changes from timing impairments associated with PD.

Chapter 5 describes an experiment that provided both a baseline measure of performance and an initial evaluation of the period-correction hypothesis in young adults. One- and five- standard interval sequences, with either no-context rhythm or rhythmic context, were presented.

The no-context condition served as a baseline measure of duration discrimination. The rhythmic context condition served to test the period-correction hypothesis by examining the effects of rhythmic context on perceived duration when one- and five-standard intervals were presented.

The effects of rhythmic context were predicted to be eliminated (or at least weakened) by increasing repetitions of the standard interval, whereby flattening the expectancy profile.

Chapter 6 describes an experiment that compared the performance of individuals with PD

(both on- and off-medication) to older adult controls and the young adults tested in Experiment 1.

These comparisons provided a test of the period-correction hypothesis in individuals with PD

8 and older adult controls. Again, the effects of rhythmic context were predicted to be eliminated

(or weakened) by increasing repetitions of the standard interval. Additionally, the experiment provided a test of the DA-mediated distortion hypothesis, which predicted greater effects of rhythmic context on perceived duration, as reflected by the expectancy profile, in individuals with PD when off medication, compared to when on medication. Moreover, individuals with PD were predicted to show greater effects of rhythmic context on perceived duration for both one- and five-standard interval sequences, compared to older adult controls and young adults.

Finally, Chapter 7 provides a general discussion of the findings of the two experiments.

Specifically, the tests of the period-correction hypothesis and the DA-mediated distortion hypothesis are discussed. Following is a discussion of how these findings fit within the broader literature of timing.

9

CHAPTER 2: BACKGROUND ON PARKINSON DISEASE

Parkinson disease (PD) is the most well understood and researched movement disorder, yet we are still far from a complete understanding of this complex disease. Significant progress has been made in the treatment of PD motor symptoms, though the mechanisms underlying treatment success are often not well understood. New treatments for PD, such as deep brain stimulation (DBS), highlight our lack of understanding about the basal ganglia (BG), as the efficacy of this treatment conflicts with current models of the BG. Continued research on PD and the BG is necessary for both a better understanding of the disease and the role of the BG in behavior, as relatively recent research has shown that the BG may play a stronger role in cognition than originally thought.

The purpose of this chapter is twofold. First, the chapter is meant to act as an introduction to PD and the major issues, both clinical and theoretical, relevant to understanding the disease.

Second, the chapter provides in-depth descriptions of PD-related topics discussed throughout this dissertation. The chapter begins with an introduction to PD and a description of the main scales used to classify PD severity. Following is a description of the pathophysiology of the disease and current treatments for PD.

Description and Classification

Parkinsonism is a family of movement disorders characterized by impaired motor control, muscle rigidity, and tremor. Occasionally, these symptoms can occur due to head trauma, vascular problems, or through drug use.1 However, in the majority of cases, the specific cause of the disordered movement is unknown. These latter cases fall into the sub-family of Parkinsonism disorders known as idiopathic PD. This dissertation, and the following review, considers only idiopathic PD.

10

Idiopathic PD is a neurodegenerative disorder that affects a collection of brain nuclei collectively referred to as the BG. The disease is characterized by motor symptoms such as akinesia (difficulty initiating movement), bradykinesia (slowness of movement), muscle rigidity and resting tremor. Additionally, problems with balance and walking are common in this population (Bloem, Hausdorff, Visser, & Giladi, 2004; Lang & Lozano, 1998a; Morris, Iansek,

Matyas, & Summers, 1996). The motor symptoms of PD are due to a severe loss of dopaminergic cells in the substantia nigra (SN), located within the BG. The extent of the cell loss is large, with some estimates suggesting that approximately 80% of dopaminergic cells are lost before motor symptoms are observed (Fearnley & Lees, 1991; Tissingh, et al., 1998).2

Dopaminergic cell loss within the SN affects communication with the striatum, resulting in the disordered movement of individuals with PD (Bergman & Deuschl, 2002; Lang & Lozano,

1998a).

Disordered movement is the most recognizable characteristic of PD, but an array of non- motor impairments, such as depression/anxiety, cognitive impairments, and perceptual impairments have also been reported. Comparatively, these impairments have received little attention in the research literature. Depression and anxiety are common among individuals with

PD, with some estimates suggesting that approximately 50% of individuals with PD suffer from depression and approximately 40% suffer from some form of anxiety (Dubois & Pillon, 1997;

Starkstein, et al., 1989). Comorbidity of both depression and anxiety have also been reported for this population (Henderson, Kurlan, Kersun, & Como, 1992). Cognitive impairments, such as dementia, attentional impairments, impairments in visuo-spatial functioning, verbal fluency, and executive functioning, are also frequently reported in individuals with PD (Antal, Bandini, Keri,

& Bodis-Wollner, 1998; Bodis-Wollner, 2003; Bohnen, Minoshima, Giordani, Frey, & Kuhl,

11

1999; Brown & Marsden, 1998; Dubois & Pillon, 1997). More relevant to this dissertation are the perceptual impairments in olfaction, vision, and audition observed in individuals with PD.

The most common and well understood perceptual deficit in PD is hyposmia, or the reduced ability to smell or detect odors. Hyposmia is estimated to present in 70-90% of individuals with PD, often presenting before the motor symptoms of the disease (Hawkes,

Shephard, & Daniel, 1997; Ponsen, et al., 2004). Several visual and visuospatial processing impairments have also been reported in PD, including impairments in spatial contrast sensitivity, mental rotation, and object detection (Bodis-Wollner & Jo, 2006; Laatu, Revonsuo, Pihko,

Portin, & Rinne, 2004). Finally, several studies have reported impairments in the perception or production of time intervals in PD (see Chapter 4 for a review). These impairments are observed for both isolated time intervals and sequences of time intervals (Grahn & Brett, 2005;

Harrington, Haaland, & Hermanowicz, 1998; Laskowska, et al., 2007) and observed across durations spanning milliseconds – seconds (Jones, Malone, Dirnberger, Edwards, & Jahanshahi,

2008; Lange, Tucha, Steup, Gsell, & Naumann, 1995). Currently, the reasons for the non-motor impairments of PD are not well understood. Often, these impairments are attributed to the damage of either the BG or DA-transmitter systems that result from the disease (Bodis-Wollner,

2003; Malapani, et al., 1998; Pastor, Artieda, et al., 1992; Pastor, Jahanshahi, et al., 1992;

Richard, et al., 1996; Stein, et al., 1990).

Due to the neurodegenerative nature of PD, neurologists have been concerned with how to best track the progression, or severity, of PD. The two primary scales used to track the severity of PD are the Hoehn and Yahr scale and the Unified Parkinson Disease Rating Scale (UPDRS).

The Hoehn and Yahr (1967) scale classifies PD severity using a five-stage system. Stages 1 and

2 are considered mild PD. In the mild stages, individuals with PD report symptoms as

12

‗inconvenient,‘ symptoms switch from unilateral to bilateral, and observable problems with posture and gait occur. Stage 3 is considered moderate PD. In the moderate stage, individuals have noticeably slowed body movements, balance impairments, and medication is less effective for controlling motor symptoms. Finally, Stages 4 and 5 are considered advanced PD. In advanced PD, individuals show muscle rigidity and bradykinesia. Additionally, balance becomes so unstable that individuals have difficulty standing or walking. Once symptoms have progressed to the later stages, constant care and supervision are often required for individuals with PD. The

Hoehn and Yahr scale is still used by some to classify PD, but it is a gross classification system that overlooks many critical benchmarks in the progression of PD. Moreover, the scale fails to account for the heterogeneity of PD, as symptoms do not necessarily present in the order specified by the scale.

Comparatively, the UPDRS uses a finer classification system than the Hoehn and Yahr scale (Fahn & Elton, 1987). Individuals with PD are assessed on three categories, 1) Mentation,

Behavior, and Mood, 2) Activities of Daily Living, and 3) Motor Examination (Appendix C provides a more detailed description of the motor examination of the UPDRS). Individuals are rated on a 0- through 4-point scale, or through yes/no responses, and a score is given for each category. Total scores on the UPDRS range from 0 (no disability) – 199 (total disability). Since the UPDRS quantitatively assesses PD severity over a wide range of symptoms, it is a useful tool to both track changes in disease severity and to compare treatment efficacy in PD. Due to the finer classification the UPDRS provides, this rating scale was used to assess PD severity in the experiment described in Chapter 6.

13

Pathophysiology of PD

The BG are a collection of four nuclei: the 1) SN, 2) globus pallidus (GP), 3) striatum, and 4) subthalamic nucleus (STN). These nuclei send ascending projections to the cortex, via the thalamus, which in turn send descending projections to the BG, creating a basal ganglia- thalamocortical loop. The BG regulate voluntary movement. Damage to this brain area, through either PD, Huntington‘s disease, or stroke, results in disordered movement characterized by 1) involuntary movements (e.g., tremor), 2) changes in posture or muscle tone, and 3) paucity or slowness of movement (Kandel, Schwartz, & Jessell, 2000). Additionally, growing evidence suggests that the BG play a role in perception and cognition. While the role of the BG in perception and cognition is less understood, compared to movement, some research has suggested that the BG influences these processes through either direct connections with the cerebral cortices or by modulation of the cerebral cortices through other brain areas (Middleton

& Strick, 2000a, 2000b).

Nuclei of the BG work in coordination to control movement, perception, and cognition.

The striatum, which is divided into the caudate nucleus and the putamen, receives descending input from the cortex. The striatum projects to the GP and the SN, which output, through different pathways, to the STN and GP. The STN and GP send ascending projections, via the thalamus, to the premotor, motor, and somatosensory cortices (Starr, Vitek, & Bakay, 1998).

Figure 3a shows a model of the basal ganglia-thalamocortical loop (Albin, Young, & Penney,

1989; Delong, 1990).

a) Cortex

Striatum GPe STN

Thalamus

GPi/SNr

b) Cortex

Striatum GPe STN

Thalamus

GPi/SNr

Figure 3: A proposed model of the basal ganglia in a) normal motor control and b) Parkinsonism (Albin, Young, & Penney, 1989; Delong, 1990). For the sake of clarity, this model is not complete. Grey arrows represent excitatory connections and black arrows represent inhibitory connections. For Parkinsonism, bold arrows represent increased activity and dotted arrows represent decreased activity. Subthalamic nucleus (STN); globus pallidus external (GPe); globus pallidus internal (GPi); substantia nigra pars reticulata (SNr).

14

15

Two major pathways exist within the basal ganglia-thalamocortical loop: the direct and indirect pathways. The role of the direct pathway is to facilitate movement, while the role of the indirect pathway is to inhibit movement. A balance of activity within both pathways is necessary for smooth, controlled movements. Both pathways project from the striatum. The direct pathway projects to the internal segment of the GP (GPi), and thalamus. In this pathway, the striatum receives excitatory input, through the neurotransmitter glutamine, from the SN pars compacta

(SNpc). Once activated, the striatum inhibits activation in the GPi through γ-Aminobutyric acid

(GABA). Less inhibition of the GPi results in less GABAergic inhibition of the thalamus, or disinhibition of the thalamus. Thalamic disinhibition result in stronger excitation of the cortex. In other words, the direct pathway facilitates movement through thalamic disinhibition.

The indirect pathway projects to the GP external (GPe), STN, and GPi. Through the indirect pathway, the striatum receives inhibitory, GABAergic, input from the SNpc. Activation of the striatum inhibits the GPe, which decreases activity in this area. The decreased activity of the GPe results in disinhibition of the STN. Excitatory projections from the STN increase activity in the GPi, which result in greater inhibition of the thalamus. Thalamic inhibition results in weaker excitatory projections to the cortex. In other words, the indirect pathway inhibits movement through thalamic inhibition.

An imbalance of activity in either the direct or indirect pathway results in disordered movement. In regard to PD, severe dopaminergic cell degeneration in SNpc affects DA input to the striatum, thereby increasing activity in the indirect pathway and decreasing activity in the direct pathway (Figure 3b). Through the direct pathway, the striatum receives decreased excitatory activation from the SNpc, which results in less inhibition of the GPi and greater inhibition of the thalamus. Through the indirect pathway, the decrease of inhibitory activity from

16

the SNpc to the striatum results in disinhibition of the GPe. Disinhibition of the GPe leads to less inhibition of the STN, and hence greater excitation of the GPi. The combined input from the both direct and indirect pathways into the GPi results in thalamic inhibition, which is responsible for the difficulties with movement initiation and slowed movement observed in PD.

It should be noted that while this model is the best working model of BG circuitry that we currently have, it is not complete. This model has served as a useful tool for understanding the pathophysiology of movement disorders associated with the BG and the development of drug therapies. However, one of the biggest challenges to this model is that it does not explain why

DBS or ablative surgeries alleviate some of the motor symptoms of PD. Specifically, this model predicts that both DBS and ablative surgeries should worsen PD symptoms. Four potential explanations for why the this model of BG circuitry inaccurately predicts worsening of PD symptoms with DBS or ablative surgeries include that: 1) the model only considers neuronal discharge rates, but not neuronal firing patterns, 2) the circuit may behave differently in a chronic DA-deprived state, 3) potentially undiscovered subcircuits within certain nuclei are responsible for PD symptoms, and 4) motor circuits outside of the basal ganglia-thalamocortical loop play a role in movement disorders (Lang & Lozano, 1998b; Starr, et al., 1998).

Treatments for PD

Several treatments have been developed to control the motor symptoms of PD by balancing the activity of both the direct and indirect pathways in the basal ganglia- thalamocortical loop. Treatments include pharmacological therapies, ablative surgeries

(pallidotomy and thalamotomy), and DBS. Pharmacological therapies are the first line of defense against the motor symptoms of PD. The main strategy employed by these therapies is to enhance

DA activity in the striatum by either: 1) increasing the amount of DA in the system, 2) directly

17 stimulating DA receptors, or 3) inhibiting the uptake or metabolism of DA. However, if an individual‘s responsiveness to a medication significantly wanes or if their motor symptoms are too severe, then surgical treatments might be considered.

Pharmacological Therapies

Levodopa. The most common and effective pharmacological therapy for PD is levodopa. Since levodopa was first used effectively in 1967 to treat PD, this drug has become the gold standard by which all PD treatments are measured (Fahn, 2008). Currently, levodopa is usually administered with other concomitant medications, such as carbidopa or carbidopa/entacapone, to improve efficacy (Lang & Lozano, 1998b; Schapira, et al., 2006). Both regular release and controlled release versions of these drugs are prescribed. In certain situations, these drug combinations can also be administered transdermally or intravenously (Schapira, et al., 2006). This review of levodopa/carbidopa will focus on the regular release version of levodopa/carbidopa, as it is well researched. However, generally speaking, most of the information discussed also relates to the other formulations of levodopa (unless otherwise specified).

Levodopa is a DA replacement therapy. Carbidopa is often combined with levodopa to stop the conversion of levodopa into DA until it has passed the blood-brain barrier (Lang &

Lozano, 1998b). After oral intake, levodopa/carbidopa achieves maximum concentration within approximately 15 – 45 minutes and has a half-life of approximately two hours (Goetz, Koller,

Poewe, Rascol, & Sampaio, 2002; Nutt & Fellman, 1984). Controlled release versions of the drug usually reach maximum concentration within two hours, but produce a constant elevation of levodopa for 3 – 4 hours longer than regular levodopa (see Simuni & Hurtig, 2002). The main success of levodopa is that lower dosages are required to manage motor symptoms and fewer

18 side effects are experienced, compared to other PD medications. However, if levodopa levels are too high, individuals may experience nausea, vomiting, hypotension, and hallucinations (Goetz, et al., 2002).

2 - 5 year Honeymoon------Wearing off effect------ON-OFF Effect Figure 4: Schematic of levodopa/carbidopa effectiveness over time. Initial response to the drug is strong for a period of 2 – 5 years, or the ‗honeymoon‘ period. After this period, individuals with PD may notice a ‗wearing off‘ effect, due to the medication being effective for a shorter period of time. Wearing off is usually tied to medication schedule. As the window of effectiveness for levodopa/carbidopa narrows, individuals with PD may notice times when their PD medications are effective (ON states) and other times when their PD-Medications are not effective (OFF states). After long-term use of levodopa/carbidopa, individuals with PD may also experience dyskinesias.

Although levodopa/carbidopa is considered the gold standard of PD treatment, some complications are associated with extended use (Figure 4). Initially, individuals taking the drug report a ‗honeymoon‘ period of approximately 2-5 years, during which their motor symptoms are consistently managed (Lang & Lozano, 1998b). After the honeymoon period, levodopa/carbidopa effectiveness decreases over time, resulting in motor side effects or

19 fluctuations in medication effectiveness. Many report ‗wearing off‘ effects, due to the drug being effective for shorter periods of time. Wearing off typically occurs before a scheduled dose of PD medications and individuals can perceive changes in movement and rigidity (Lang & Lozano,

1998b). Later in levodopa treatment, individuals may experience ON states, when their medication is effective, followed by gradual or sudden occurring OFF states, when their medication is no longer effective (Goberman & Coelho, 2002). The ON-OFF effect is not tied to the medication schedule, and OFF states may spontaneously improve without additional medication (Goberman & Coelho, 2002). Finally, long-term use of levodopa may result in rapid, involuntary jerking movements throughout the body, or dyskinesias (Obeso, Olanow, & Nutt,

2000). Dyskinesias occur in the ON state, usually when levodopa/carbidopa has reached maximum concentration (Obeso, et al., 2000).

DA agonists. Unlike levodopa, DA agonists (e.g., bromocriptine, pergolide, cabergoline, pramipexole, ropinirole, rotigitine) directly stimulate DA receptors (Lees, 2005; Schapira, et al.,

2006). Although the mechanisms of DA agonists are not completely understood, it is widely believed that DA agonists stimulate the indirect pathway through D2-like receptors (Lees, 2005;

Schapira, et al., 2006).

DA agonists provide a number of clinical benefits. The primary clinical benefit is that

DA agonists have lower incidences of dyskinesia, compared to levodopa. Furthermore, DA agonists are believed to have a longer elimination half-life than levodopa (Goetz, et al., 2002).

The longer half-life is believed, by some, to create a constant pattern of stimulation on the DA receptors, as opposed to the ‗bursts‘ of stimulation levodopa provides; providing a possible explanation of why dyskinesias are not common with the use of DA agonists (Lees, 2005;

Schapira, et al., 2006). However, a number of problems exist with the use of DA agonists. First,

20 they take longer to reach effective doses, compared to levodopa (Lees, 2005; Schapira, et al.,

2006). Second, serious side effects are associated with the use of DA agonists, such as nausea, vomiting, dizziness, drowsiness, confusion, psychosis, and hallucinations (Lees, 2005; Schapira, et al., 2006).

Catechol-O-methyl transferase inhibitors. Catechol-O-methyl transferase (COMT) is an enzyme that metabolizes levodopa. COMT inhibitors (e.g., entacapone, tolcapone) are used in combination with levodopa to limit levodopa metabolization, thereby increasing the availability of levodopa (Goetz, et al., 2002). Entacapone is used to inhibit COMT in the peripheral nervous system, while tocapone inhibits COMT in the central nervous system (Goetz, et al., 2002).

Adverse reactions to COMT inhibitors are primarily due to increase in the availability of levodopa, so COMT inhibitors share the same side effects as levodopa (Goetz, et al., 2002).

Monoamine oxidase B inhibitors. Monoamine oxidase B (MAO-B) inhibitors (e.g., selegiline, rasagiline) are used to block the metabolism of DA by the MAO-B enzyme. In the treatment of PD, MAO-B inhibitors are used to either increase endogenous levels of DA or are combined with levodopa therapy to increase levodopa levels (Goetz, et al., 2002; Lees, 2005).

Therapeutically, MAO-B inhibitors may be initially prescribed after a PD diagnosis in an attempt to postpone levodopa therapy. Or, MAO-B inhibitors may be prescribed in combination with levodopa to limit dyskinesias (Lees, 2005).

Amantadine. Amantadine can be used either alone, or in combination with levodopa,

DA agonists, and anticholinergics (Lees, 2005). While the mechanisms of action of amantadine are not well understood, one proposal is that it enhances the release of DA and blocks DA uptake

(Lees, 2005). More recently, there has been a resurgence of interest in the use of amantadine to

21 treat PD, since it might reduce the motor fluctuations and dyskinesias that occur after prolonged use of levodopa (Lees, 2005).

Anticholinergics. Although cholinergic neurons are rarely mentioned in models of BG circuitry, they are present in the striatum and synapse onto the direct and indirect pathways. The role of striatal cholinergic interneurons is to inhibit the direct pathway and excite the indirect pathway. Thus, less DA activity in the striatum, due to cell loss, results in more excitation of the indirect pathway and increased inhibition to the motor cortex. The behavioral results of this increased inhibition of the motor cortex are the movement problems (i.e., slowed movement) observed in PD. One pharmacological approach to treating the motor symptoms of PD has been to reduce the activity of the cholinergic interneurons of the striatum through anticholinergic medications (e.g., benzhexol, benztropine, orphenadrine, procyclidine).3

While the exact mechanism of action of anticholinergic medications is still not well understood, it is hypothesized that anticholinergic medications balance the production of DA and acetylcholine activity in the striatum (Lees, 2005). Additionally, some anticholinergics (e.g., benztropine) block DA uptake (Goetz, et al., 2002). While some improvement in the motor symptoms of PD have been reported with anticholinergic medications (Katzenschlager, Sampaio,

Costa, & Lees, 2003), these medications have also been shown to impair cognitive functioning

(Goetz, et al., 2002).

Pharmacological therapies for non-motor symptoms of PD. Often, the non-motor issues of PD are unresponsive to DA-enhancing treatments (Truong, Bhidayasiri, & Wolters,

2008). Such issues include: cognitive impairment/dementia, psychiatric disorders, autonomic disorders, and sleep disorders. Medications normally prescribed to treat these issues in the general population are prescribed for individuals with PD. However, doctors must be cautious in

22 which drugs they prescribe to individuals with PD, as there is the potential for dangerous drug interactions. While a review of these medications is beyond the scope of this dissertation, interested readers are referred to Truong, Bhidayasiri and Wolters (2008).

In conclusion, a variety of pharmacological therapies can be used to treat PD. These drug classes can be used alone (monotherapeutically) or in combination. The appropriate therapeutic approach is dependent on how symptoms present and individual reaction to a particular drug.

Hence, the appropriate pharmacological treatment is dependent upon the individual and disease stage; prescribed pharmacological treatments are very heterogeneous within PD. Finally, DA- enhancing medications are effective for treating the motor symptoms of PD, but not many of the non-motor issues. Therefore, it is common to see a variety of non-DA-enhancing medications also prescribed to treat these aspects of PD.

The preceding review is provided to familiarize readers with the different pharmacological therapies used to treat the motor symptoms of PD and to describe the mechanisms of action for these therapies. This information is relevant for the experiment described in Chapter 6, as the DA-mediated distortion hypothesis generates predictions regarding the effects of DA-enhancing medications on the perceived duration of a standard interval in the presence of rhythmic context. Furthermore, these pharmacological therapies are discussed to provide the reader with some background information on the medications participants were taking when tested on medication.

Ablative Surgeries

Ablative surgeries for the treatment of PD began in the 1940‘s. These early surgeries were conducted before levodopa treatment and, unfortunately, before the establishment of a strong scientific basis for their use (Starr, et al., 1998). During the 1960‘s and 1970‘s, the use of

23 surgery waned with the availability of levodopa (Starr, et al., 1998). However, a resurgence of ablative surgeries occurred once the levodopa honeymoon period was discovered. Thankfully, this resurgence brought new target sites and improved ablative techniques (Starr, et al., 1998).

Today, ablative surgeries are rare, due to the development of DBS. DBS provides similar results as ablative surgeries and is, technically, reversible (Starr, et al., 1998).

The primary target sites for ablative surgeries are the GPi (pallidotomy) and the ventral intermediate nucleus (VIM; thalamotomy). However, some research has also investigated the clinical effects of STN lesions (Lozano, Dostrovsky, Chen, & Ashby, 2002; Starr, et al., 1998).

The choice of target site is dependent upon the motor symptoms that present. Thalamotomy is effective for the treatment of PD tremor, but has little effect on other motor symptoms.

Pallidotomy is effective for most motor symptoms of PD. Figure 5 shows the target site and physiological outcomes of ablative surgeries. The ultimate goal of these surgeries is to eliminate abnormal patterns of activity from the BG motor circuit. Specifically, ablative surgeries attempt to balance the activity between the direct and indirect pathways to increase ascending activity between the thalamus and cerebral cortex.

Deep Brain Stimulation

A more recent treatment for the motor symptoms of PD is DBS. The general procedure implants high-frequency electrodes into a target structure, dependent upon the symptoms an individual presents. Once the electrodes are determined effective, a pulse generator for the electrodes is subcutaneously implanted into the pectoral pouch. Electrode stimulation can be continually adjusted through the pulse generator, based on the response of the individual, until the optimal pulse frequency, amplitude, and duration parameters are found that limit the motor symptoms of PD (Kringelbach, Owen, & Aziz, 2007).

24

Although DBS is gaining acceptance as a treatment for a variety of pharmacological treatment-resistant disorders (e.g., pain, depression), its initial success was in the treatment of PD and essential tremor (Kringelbach, et al., 2007). DBS for PD is performed when medications have a short-lived effect on motor symptoms, cause extreme dyskinesias, or cause severe side effects (Benabid, 2003; Perlmutter & Mink, 2006). Moreover, individuals must be a good candidate for general surgery and show no signs of cognitive impairment, as DBS can exacerbate these impairments (Perlmutter & Mink, 2006).

Target sites for DBS in PD are the VIM, GPi, or STN. VIM DBS is becoming less common (Perlmutter & Mink, 2006). While VIM DBS is effective at reducing tremor in the limbs, it has little effect on the other motor symptoms of PD. The GPi is a more common target site for DBS, as it reduces most of the motor symptoms of PD, along with reducing levodopa- induced dyskinesias and the sensory hallucinations associated with other target sites (Loher,

Burgunder, Weber, Sommerhaider, & Krauss, 2002; Perlmutter & Mink, 2006). Unfortunately,

GPi DBS rarely leads to a reduction in DA-enhancing medications, so this target area is not beneficial for individuals that experience severe medication side effects (Perlmutter & Mink,

2006). Since STN DBS reduces most of the motor symptoms of PD, and often allows for reductions in DA-enhancing medications, this is becoming the preferred target site.

The effectiveness of DBS on PD motor symptoms is on par with DA-enhancing medications alone (Perlmutter & Mink, 2006). However, there are a few additional benefits of

DBS. First, DBS can reduce the medications used to treat PD, and in turn, the side effects associated with medications. Second, DBS reduces PD symptom severity during OFF periods

(Jaggi, et al., 2004). DBS-related benefits for PD are known to last for at least four years; however, the long-term effects of DBS are still unknown (Rodriguez-Girones & Kacelnik, 2001;

25

Visser-Vandewalle, Temel, van der Linden, Ackermans, & Beuls, 2004). There is some evidence to suggest that certain impairments in cognitive processing may occur with DBS, such as delayed spatial recall or response inhibition (Hershey, et al., 2004) and DBS has been reported to cause hallucinations in certain individuals (Diederich, Alesch, & Goetz, 2000). However, other studies have found that DBS may help to alleviate some of the non-motor issues of PD, reporting that

DBS may improve executive functioning in PD (Hershey, et al., 2004) and may act like antidepressant medications for some individuals (Takeshita, et al., 2005).

DBS has shown great promise as a treatment option for severe cases of PD, but little is known about the mechanisms of action for this treatment. Initially, DBS was thought to only be clinically effective with high-frequency stimulation, 130 – 187 Hz, of the target area

(Kringelbach, et al., 2007). However, recent findings suggest that low-frequency stimulation, 5 –

10 Hz, in the STN and PPN might also be clinically effective for some motor symptoms of PD

(Kringelbach, et al., 2007; Montgomery & Gale, 2008).

Moreover, the proposed mechanisms of action in DBS are controversial. Since DBS and ablative surgeries produce similar results, DBS was hypothesized to inhibit neuronal transmission in a manner similar to lesioning target areas (Benabid, Chabardes, Mitrofanis, &

Pollak, 2009). Specifically, DBS was proposed to inhibit the overactivity of the GPi by decreasing the firing rate. However, critics of this hypothesis suggest that DBS increases the firing rate of the GPi, reporting activity in efferent nuclei due, presumably, to excitation of the

DBS target (Kringelbach, et al., 2007). Currently, the most accepted hypothesis is that DBS creates excitation in the target area. Alternative hypotheses exist, but have limited supporting evidence. One such hypothesis is that DBS might work in treating the motor symptoms of PD by regulating the firing pattern of the target area, but not the firing rate. The firing pattern

26 hypothesis proposes that PD results in irregular firing of certain BG nuclei and that DBS causes the target areas to fire in a regular pattern (Montgomery, 2005). Another hypothesis is that DBS might work by changing the levels of activation throughout the BG circuit (system stimulation), as opposed to just activation of the target area (local stimulation; Kringelbach, et al., 2007;

Montgomery & Gale, 2007). While interesting, stronger supporting evidence is necessary for these hypotheses.

None of the participants in the experiments described in Chapter 6 had either ablative surgeries or DBS; however, a brief review of these treatments are included for the reader to provide background information for the DBS timing research discussed in Chapter 4.

Additionally, this information was provided to give a complete overview of treatments for PD.

Cortex

Striatum GPe STN

Thalamus

GPi/SNr

Ablative Surgery Target Site DBS Target Site

Figure 5: Target sites for ablative and DBS surgeries. Grey arrows represent excitatory connections and black arrows represent inhibitory connections. Bold arrows represent increased activity and dotted arrows represent decreased activity. The choice of target site is dependent on the symptoms presented in each case of PD. Regardless of target site, the end goal of ablative surgeries or DBS is to reduce the increased inhibition of the thalamus through the GPi/SNr, or to increase the excitatory connections between the thalamus and cortex; this will result in increased excitation of the cortex, via the thalamus. Subthalamic nucleus (STN); globus pallidus external (GPe); globus pallidus internal (GPi); substantia nigra pars reticulata (SNr)

27

28

CHAPTER 3: PERCEPTION AND PRODUCTION OF TIME INTERVALS—THEORY, DATA, AND NEURAL BASES

We are astutely aware of the duration of a stoplight on our commute to work and can easily tap along with a song on the radio. Presumably, these behaviors require an internal ‗clock‘ to measure time. A central theoretical issue in the field of timing is the nature of this clock.

Interval theories view the clock as a stopwatch or an hourglass that can be started, stopped, and reset arbitrarily. Alternatively, entrainment theories view the clock as an oscillator that entrains

(synchronizes) to temporal event regularities in our environment and generates expectancies about the occurrence of future events in time. These two perspectives on the clock make up the general theoretical approaches discussed in the timing literature.

The aim of this chapter is to provide an overview of the research on the perception and production of both isolated time intervals and sequences of time intervals. To this end, theoretical viewpoints from both interval and entrainment models of timing are discussed.

Following is a review of the perceptual and motor timing data for both isolated time intervals and sequences of time intervals in young adults. Finally, the relevant research on the neural bases of perceptual and motor timing is reviewed.

Theories of Timing

Interval Theories

Interval theories take an information processing approach to explain the timing mechanism. Models within this framework often include three separate stages: clock, memory, and decision (Church, 1984; Gibbon, Church, & Meck, 1984). Scalar Expectancy Theory (SET) is the most cited, and tested, interval theory (Gibbon, 1977). The success of SET is due, in part, to its flexibility. It has been used to test hypotheses of both human and non-human animal

29 timing, age-related changes in timing, and timing performance across a range of tasks. Figure 6 provides a schematic of SET.

Switch

Clock Stage Pacemaker Accumulator

Attention

Memory Stage Working Reference Memory Memory

Decision Stage Comparator

Response

Figure 6: A schematic of SET (Gibbon, 1977).

According to SET, the clock stage consists of pacemaker, an attentionally-mediated switch, and an accumulator. The pacemaker generates a continuous stream of pulses at a regular rate. Once released, pulses are sent to the accumulator through an attention-modulated switch.

30

Attention to the start of a to-be-timed event closes the switch and pulses flow into the accumulator. At the end of the event, the switch opens, stopping the collection of pulses. The pulse count collected in the accumulator is the encoded representation of event duration.

The memory stage of SET consists of working and reference memory. Once the switch opens, the pulse count in the accumulator is sent directly to working memory (Church, 1984;

Gibbon, et al., 1984). The duration code stored in working memory can either be used to make a comparative decision between the current duration and previously experienced durations (e.g.,

‗shorter‘ or ‗longer‘), or it can be sent to reference memory. Reference memory is theorized to be the long-term store of pulse counts. Within reference memory, a distribution of pulse counts are stored and retrieved during the decision stage.

The decision stage consists of the comparator (Allan, 1998; Church, 1984; Gibbon, et al.,

1984). As previously mentioned, pulse counts can be retrieved from either working or reference memory. Once retrieved, duration codes are compared, via a decision rule, and the appropriate response is generated.

The main assumption of SET is the scalar property, which proposes that variability in timing is proportional to the timed duration (Meck, 1996). In other words, timing variability is proposed to be scale-invariant and holds to Weber‘s law. According to SET, scalar variability can arise from any stage in the model (Meck, 1996). However, violations of the scalar property pose problems for the theory, as they are difficult to explain within the context of the model (see

Malapani & Rakitin, 2003).

While SET makes specific predictions regarding the timing of isolated intervals, it does not make specific predictions regarding the timing of sequences of intervals. Three interval- based theories have been proposed to explain the perception and production of sequences of time

31 intervals. The Wing and Kristofferson (W&K) model and the component (slope) analysis model generate predictions regarding the production of sequences of time intervals, while the multiple- look model generates predictions regarding the perception of sequences of time intervals.

According to the W & K model, the production of sequences of intervals, such as when tapping along to a metronome, is carried out through two component processes: 1) a central clock that emits pulses and sends commands to the effector system and 2) a motor delay that carries out the response. Equation 1 provides a mathematical representation of the model. At the start of the target time interval, the clock begins emitting pulses into the accumulator until the appropriate pulse count, or clock interval (Cn), for the target time is reached. A finger tap occurs after a delay in the motor implementation system (D). Therefore, the total produced duration (In) is the sum of the clock interval and the difference between the initiating (Dn-1) and terminating

(Dn) taps that mark the inter-tap interval (ITI). The W & K model proposes that total ITI variability arises from variability in the clock and motor system. Since clock and motor components are assumed to be independent, total tapping variance can be decomposed into separate estimates of clock variance and motor variance (Wing, 1980; Wing & Kristofferson,

1973).

(1)

The component (slope) analysis model has also been proposed to assess the production of sequences of time intervals (Ivry & Corcos, 1993; Ivry & Hazeltine, 1995). Like the W & K model, slope analysis decomposes variability into clock and motor components. However, consistent with Weber‘s law, this model hypothesizes that standard deviation increases with T.

Hence, the standard deviation of taps is regressed onto T. The slope of the line provides an estimate of clock variability and the intercept provides an estimate of motor variability.

32

Drake and Botte (1993) proposed the multiple-look model to extended interval theories of timing to the perception of sequences of time intervals, such as when the rate (tempo) between two sequences of time intervals must be compared. For example, suppose that listeners are presented with a tone sequence in which two standard intervals are followed by two comparison intervals and participants are asked to judge the tempo of the comparison sequence relative to the standard sequence. The multiple-look model proposes that each standard interval provides an independent, yet variable, estimate of the standard duration (a ‗look‘). In this example, the two looks at the standard interval are averaged to form an aggregate memory trace of the standard.

As the number of looks increase (e.g., four standard intervals), a more accurate, and less variable, estimate of the average duration of the standard interval will obtain, resulting in improved temporal discrimination.

Entrainment Theories

Unlike the information-processing approach taken by interval theories, entrainment theories provide a dynamical systems approach to timing (see Jones, 2004 for a general review of entrainment models; see also McAuley, 1995 for a more in-depth description of the model).

Generally speaking, this approach conceptualizes the clock as an endogenous oscillator that synchronizes with exogenous periodic events. Once synchronized, a change in duration is detected through a misalignment between the peak of the oscillator and the timed event (Jones,

1976; Jones, 2004; McAuley, 1995; McAuley & Jones, 2003)

Figure 7a provides a schematic of a prototypical entrainment theory. Entrainment theories are based on three main assumptions. The first assumption is that environmental events have a periodic (rhythmic) structure, referred to as a driving rhythm. The second assumption is that humans have self-sustaining endogenous oscillators that span a wide range of rates, referred to as

33 the driven rhythm.4 The third assumption is that humans connect and adapt with rhythms in the environment, through the process of entrainment. In other words, driven rhythms within the human adapt to the driving rhythms in the environment, resulting in entrainment. Once the oscillator is entrained, information about the relative timing of events is obtained from a misalignment between the driving and driven rhythms, referred to as temporal contrast (Figure

7b). For example, events that occur before the peak of the oscillator indicate shorter durations (-) and events that occur after the peak indicate longer durations (+).

Two adaptive processes that are critical for entrainment are period and phase correction

(Large & Jones, 1999; Mates, Muller, Radil, & Poppel, 1994; Semjen, Vorberg, & Schulze,

1998). The period-correction process adjusts the period of the oscillator to match, in time, with the rate of the driving rhythm. Phase correction adjusts the peak of the oscillator, so that it matches, in phase, with the onset of the driving rhythm. Comparatively, the process of phase correction are hypothesized to occur quickly, while period correction is hypothesized to be slow since it is dependent upon phase correction (Large & Jones, 1999).

One aim of this dissertation is to test a specific hypothesis regarding period correction.

Specifically, the period-correction hypothesis posits that repetitions of a standard interval should eliminate the effects of rhythmic context on perceived duration. According to entrainment models of timing, the effects of rhythmic context are due to the oscillator entraining to the context rhythm. Early- or late-standard endings occur unexpectedly, due to the periodic extrapolation of the context rhythm. The unexpected endings result in systematic distortions in the perception of the duration of the standard interval. However, multiple standard intervals should afford period correction to the standard interval, thereby eliminating the effect of

34 rhythmic context. Period correction to the standard interval is predicted to produce a flat (or at least weakened) expectancy profile. a) Driving Rhythm

Phase Period

Driv en Rhythm

Time b) Early Onset On-Time Onset Late Onset

- + - + - +

Figure 7: Schematic of a prototypical entrainment theory. a) In this example, a tone sequence (circles) is the driving rhythm. Over time, the driven rhythm (oscillator) entrains with the tone onsets of the driving rhythm. Entrainment occurs through the processes of period and phase correction. b) Examples of temporal contrast, for which events occur either early, on time, or late, relative to the peak of the oscillator. Initial negative temporal contrast indicates that the duration of the time intervals in the sequence are ‗shorter,‘ while initial positive temporal contrast indicates that the duration of the time intervals in the sequence is ‗longer.‘

Theory of the Neural Bases of Timing

Another ongoing theoretical debate in the timing literature concerns the neural bases of perceptual and motor timing. Several brain areas have been shown to play a role in timing, yet there is debate over the function each area serves. While several different proposals have been made in an attempt to untangle the different brain areas involved in perceptual and motor timing,

Lewis and Miall (2003b) offer the most complete hypothesis for the involvement of different brain areas in perceptual and motor timing. Two distinct timing systems are proposed that are differentially recruited for a given timing task, based on the following criteria: 1) the duration of

35 the to-be-timed interval(s), 2) the requirements of the task, and 3) the movement required to perform the task. The automatic timing system is hypothesized to be involved in the processing of subsecond durations, which are continuously measured, and defined by movement. Since automatic timing is proposed to operate for predictable or overlearned stimuli, this system is hypothesized to operate without ―attentional modulation‖ (Lewis & Miall, 2003b). Generally speaking, the automatic timing system is most likely recruited for tasks similar to those used to investigate perceptual and motor timing of sequences of time intervals. In contrast, the cognitively controlled timing system is hypothesized to be involved in the processing of suprasecond durations, which are not defined by movement, and occur as discrete events. This system is likely to be recruited for tasks used to investigate the perceptual and motor timing of isolated time intervals. Lewis and Miall propose that different brain areas serve the automatic and controlled timing systems. The automatic system is hypothesized to be served by motor and pre-motor circuits, while the cognitively controlled system is served by the prefrontal and parietal cortices. Other brain areas, such as the BG and cerebellum, are hypothesized to be involved in general timing processes used by both systems.

Data

The literature on the perception and production of the duration of time intervals is vast, with studies utilizing a wide range of different methodologies and generating a large amount of behavioral data. While a complete review of this data is beyond the scope of this dissertation,

Grondin (2001b, 2010) provides an excellent discussion of these topics and presents data across several different areas within psychology. Within experimental psychology, there is also an extensive research on non-human animal timing. Interested readers are directed to reviews by

Allen and Church (2002) and Crystal (2007), who discuss non-human animal data on the

36 perception and production of the duration of time intervals. This dissertation will provide an overview of the human behavioral data of perceptual and motor timing. Specifically, this review will focus on research concerned with short interval durations (< 2000-ms), as short-interval timing is investigated by the experiments reported in Chapters 5 and 6. Data from studies investigating longer durations are discussed, briefly, so that comparisons can be made between the timing data for young adults, older adults, and individuals with PD.

Critical Issues in the Timing Literature

Even within the literature on human timing, studies of perceptual and motor timing have utilized a variety of different tasks and timescales. Often, data from these studies have been mixed, making it difficult to reach broad conclusions about timing. Before discussing the behavioral data for the perceptual and motor timing of isolated intervals and sequences of intervals, I will discuss some of the critical issues in the timing literature. Consideration of these issues will help the reader prepare for the following review of data and alert him/her to the empirical issues that arise when conducting research on timing and certain factors that must be considered when interpreting this data. I have divided this discussion into the following issues:

1) psychophysics of time, 2) timescale, 3) general measures of time perception and production, and 4) time perception and production tasks.

Psychophysics of time. One overarching issue in the timing literature is the best way to conceptualize the experience of time. Few would argue against the claim that we have a ‗sense‘ of time, but what exactly is meant by this statement? Can timing be considered a sense in the same way we consider audition and vision? Or, is the sense of time abstracted from other sensory information? A portion of the research on timing has focused on the question of whether timing acts like a sense. These studies have taken a psychophysical approach to the study of time

37 perception. Psychophysics is concerned with the relationship between the physical properties of a stimulus and the psychological properties of the sensation. One basic psychophysical question asked by timing researchers is whether a relationship exists between subjective and objective time. While some debate exists over the best characterization of this relationship, objective and subjective time have an approximately linear relationship, which suggests that the subjective experience of time accurately reflects objective time (Allan, 1979; Getty, 1975; Stevens, 1957; but see Eisler, 1976).

Further research has investigated whether perceptual laws, specifically Weber‘s law, apply to timing. Weber‘s law states that the difference threshold, or the smallest change necessary between two stimuli to detect a difference, increases proportionally to the magnitude of the stimulus. In other words, the ratio (Weber ratio) between the difference threshold and stimulus magnitude is proposed to be constant.

Other research on timing has focused on a form of Weber‘s law that considers the relationship between a to-be-timed duration and timing variability. The scalar property of timing, predicted by SET, proposes that increases in the duration of a to-be-timed event should result in proportional increases in variability. In other words, the measure of coefficient of variation, or

CV = (standard deviation / mean duration), is predicted to be constant across all durations.

Timescale. A central issue in the research on perceptual and motor timing is the timescale under consideration. The majority of research on human and non-human animal timing has focused on durations in the range of milliseconds – minutes. Yet, several behavioral and neural activation differences are observed within this range of durations for perceptual and motor timing. Hence, this dissertation will focus on the timescale of durations between 300-ms – 2000- ms. A similar timescale division has been proposed by others (Drake & Botte, 1993; Fraisse,

38

1978; Getty, 1975; Grondin, 2010; Michon, 1964; Poppel, 2004). This timescale is of interest for several reasons. First, this timescale covers many relevant behaviors, such as speech, music, and motor coordination. Second, this timescale is rhythmically stable in humans (Drake & Botte,

1993; Fraisse, 1963; Grondin, 2001a). For sequences of equal time intervals, in which the onsets of stimuli (or the inter-onset interval, IOI) mark sequence rate, both extremely fast sequence rates (< approximately 300-ms IOIs) and slow sequence rates (> approximately 2000-ms IOIs) appear to be timed differently than rates within this range (Drake & Botte, 1993; Grondin,

2001a; Madison, 2001). Sequences consisting of equal 2000-ms IOIs are no longer heard as a cohesive pattern, which led to Fraisse‘s (1963) proposal that 2000-ms might be the upper boundary of the psychological present. While rhythm is still perceivable for sequences of equal time intervals with IOIs shorter than 300-ms, Weber‘s law appears to break down for IOIs below this cutoff (Drake & Botte, 1993).

Admittedly, this timescale division is somewhat arbitrary and has been chosen to cover the perceptual and motor timing of sequences of time intervals. It should be noted that other timescale divisions have been proposed, dividing durations into sub-second (< 1000-ms) time intervals and supra-second (> 1000-ms) intervals (Hellstrom & Rammsayer, 2004; Lewis &

Miall, 2003b; Penney & Vaitilingam, 2008; Petrusic, 1984; Rammsayer, 2008). The sub- and supra-second timescale division is often used in research on animal timing and for human studies in which participants are instructed to subdivide, or segment, longer durations into smaller intervals.

General measures of time perception and production. One difficulty faced when reviewing the research on timing is that a variety of different methodological and analysis techniques have been used to answer questions regarding the experience of time. However,

39 broadly speaking, studies of perceptual and motor timing are concerned with two general measures (Grondin, 2001b, 2010). The first measure concerns the accuracy of the mental representation of time. In other words, these measures focus on the match, or disparity, between subjective and objective duration. Accuracy measures afford investigation of factors that might distort the sense of time. The second measure concerns the variability of the timing system.

While the average of an accuracy measure might suggest that the subjective experience of a duration matches objective duration, variability measures assess the accuracy of the mental representation of duration from trial-to-trial. While this conceptualization does not capture all forms of analyses used in timing research, it provides an overview of the main dependent variables that are discussed in the following review.

Time perception and production tasks. All tasks used to investigate perceptual and motor timing can be broadly divided into two categories. For retrospective tasks, participants are unaware that they will be asked to make an estimate of elapsed duration. Participants experience an event and are asked afterward to estimate how much time has passed. Often, responses are given in a unit of time, such as minutes. One limitation to retrospective tasks is that, often, only one estimate of duration can be obtained from participants, as they will be alerted to the fact that they must attend to time in the future (although see Boltz, 1994). Alternatively, for prospective tasks, participants are explicitly informed before the study that they will be asked to produce or make judgments about the duration of time intervals. The duration-discrimination with rhythmic context task discussed in the Chapter 1 is an example of one prospective timing task.

It is widely agreed that retrospective and prospective tasks tap into very different timing processes. First, while prospective tasks require participants to attend to the duration of a time interval, retrospective judgments are a surprise (Block & Zakay, 1997; Zakay & Block, 1997,

40

2004). Second, these tasks can differ in the timescale they consider. Prospective tasks are often used to investigate durations lasting up to a few seconds, while retrospective timing tasks are usually used to investigate durations longer than 10 seconds.

The main finding of studies that compare these two tasks is that estimates of the duration of a time interval are more accurate for prospective tasks, compared to the retrospective tasks

(Brown & Stubbs, 1988; Hicks, Miller, & Kinsbourne, 1976). However, these differences might be duration dependent, as some studies using retrospective tasks show that estimates about the duration of a time interval are relatively accurate for up to approximately 20 seconds, but longer durations tend to be underestimated (Hicks, 1992; Predebon, 1996; but, see Grondin & Plourde,

2007). Based on the relatively poor accuracy of duration estimates and methodological problems associated with retrospective tasks, most timing research has utilized prospective tasks.

Studies utilizing prospective tasks have relied on different methodologies to assess the mean and variability of the mental representation for the duration of time intervals. This is, in part, due to the timescale under investigation. Certain methods lend themselves to the study of short durations, while others are more appropriate for longer durations. Furthermore, certain tasks are appropriate for the study of the perception and production of isolated time intervals, while other tasks are more appropriate for the study of both isolated time intervals and sequences of time intervals. In this dissertation, I will the divide the different methodologies into two different types of tasks: production or perception. Production tasks, typically, require some form of a motor response to produce either a single time interval or a sequence of time intervals. In most production tasks, the time to initiate the motor response is factored in to the produced time interval. If initiation of the motor response is slowed, then a longer time interval is produced.

Alternatively, perception tasks often require a comparative judgment between two or more time

41 intervals. Although these tasks often require a motor response, such as a button press, the time to initiate the motor response does not factor into task performance. The distinction between these tasks is important when considering timing in a population with a movement disorder, such as individuals with PD. Studies that utilize production tasks in this population can be difficult to interpret, as it becomes unclear whether potential differences between individuals with PD and controls are due to disordered movement or impairments in timing. While I have divided these tasks into separate categories, for obvious reasons, most researchers agree that both types of tasks rely on a common timing mechanism (Ivry & Hazeltine, 1995).

Two basic production tasks used in timing research are duration production and reproduction tasks. For this dissertation, I have also included the discussion of verbal estimation tasks along with duration production tasks. While verbal estimation tasks do not require a motor response that factors into duration production, the task is conceptually similar and provides complementary findings to production tasks that will become evident in Chapter 4.

Studies using verbal estimation and duration production tasks have focused on participant‘s mental representation of isolated time intervals lasting longer than 1000-ms. Longer durations are, often, investigated with these tasks, due to the nature of the response required.

Verbal estimation tasks present participants with a duration and have them provide a verbal estimate of the duration, usually in a specific time unit, such as seconds. Duration production tasks are a variant of this task. Participants are told a duration, such as 3 seconds, which they must produce by holding down a response key for the given duration. One general concern with the use of either task is that they rely on an internal (memory-based) representation of a clock unit duration, such as 5 seconds, which has been experienced outside of the laboratory. As we will see later, this issue can lead to problems interpreting data obtained from these tasks.

42

Duration reproduction tasks do not rely on a memory-based representation for clock unit durations. Instead, the to-be-timed duration(s) is presented in the experimental session and does not have a verbal label attached. For example, participants may hear a tone that last for a certain duration, usually termed the standard, and are asked to reproduce the duration of the tone. A benefit derived from the use of this task is that no verbal label of time is attached to the duration.

Furthermore, this task can easily be extended to the production of sequences of time intervals, allowing researchers to investigate the reproduction of sequences of time intervals in synchrony with a metronome.

The perceptual tasks used in timing research, often, present participants with two, or more, time intervals and have them to compare the relative duration of the intervals. The majority of these tasks use methodologies and analyses similar to those used in psychophysics.

As discussed previously, one benefit of perceptual timing tasks is that they do not require a time- dependent motor response. Two additional benefits of perceptual timing tasks are that they can be used to investigate a wide range of durations, spanning milliseconds – seconds, and can easily be extended to the study of both isolated time intervals and sequences of time intervals.

The experiments described in Chapters 5 and 6 primarily investigate performance on a perceptual timing task. However, one production task was investigated in the experiment described in Chapter 6. A perceptual task was chosen for this dissertation to assess timing in individuals with PD without concern of how motor impairments might affect timing performance.

Production of Time Intervals

The following section will briefly review the behavioral data for the production of both isolated time intervals and sequences of time intervals. First, the production of isolated time

43 intervals are discussed, focusing on the relevant data for verbal estimation, duration production, and duration reproduction tasks. Second, the production of sequences of time intervals are discussed, focusing on spontaneous motor tempo and the synchronization-continuation task. This review is meant for the sake of comparison with the data for older adults and individuals with PD presented in Chapter 4.

Generally speaking, young adult performance on verbal estimation tasks is poor. While verbal estimates of duration tend to increase linearly with objective duration, individuals tend to estimate the duration as shorter than the objective duration, or underestimate duration

(PentonVoak, Edwards, Percival, & Wearden, 1996; Wearden & Culpin, 1998). Moreover, variability for verbal estimation tasks often violates the scalar property, offering little support for

SET (Wearden, 1999; Wearden, Edwards, Fakhri, & Percival, 1998). The poor performance observed in verbal estimation tasks may be due to participants providing estimates of duration that conform to ‗rounded‘ time units, such as one second. Or, poor performance might be due to timescale, as many studies use suprasecond durations so that participants can provide a verbal label for the duration (Wearden & Lejeune, 2008).

In contrast to verbal estimation, duration productions by young adults tend to be longer than the objective duration, or overproduced (Labelle, Graf, Grondin, & Gagne-Roy, 2009;

Wearden & McShane, 1988). Moreover, the variability associated with duration productions has been found to conform to the scalar property, but this effect might be feedback dependent, making it unclear whether the data from this task supports SET (Wearden, 2005; Wearden &

McShane, 1988). The use of feedback in many studies utilizing this task may explain why better performance is observed for this task, compared to verbal estimation tasks.

44

Studies utilizing tasks that require the reproduction of isolated time intervals often report that young adults can reproduce veridical durations up to approximately 3000-ms. However, durations longer than 3000-ms tend to be underproduced (Kagerer, Wittmann, Szelag, & Von

Steinbuchel, 2002). Concerning the variability of duration reproductions, mixed findings are reported regarding whether variability for this task conforms to the scalar property. Some studies provide support for SET, reporting constant CV values across a range of durations, especially in the range 200 – 2000-ms (Rakitin, et al., 1998; Woodrow, 1930), but other studies find that CV decreases with durations up to 1000-ms (Fortin & Rousseau, 1998; Macar, Grondin, & Casini,

1994; Wearden, 2003; Zakay, 1993a, 1993b). Unfortunately, most of the studies investigating the variability of duration reproductions have varied a number of task parameters, such as the timescale investigated and the presence of feedback, which make it difficult to clearly interpret these findings.

Several studies have used duration reproduction tasks to investigate the effects of divided attention on the reproduction of isolated time intervals. Data from these studies suggest that both memory and attention play a role in the timing of isolated time intervals. When a nontemporal task is performed concurrently with the encoding of a to-be-reproduced duration, temporal reproductions tend to be underproduced. Yet, when a nontemporal task is performed concurrently with the reproduction of a duration, reproductions tend to be overproduced (Fortin

& Rousseau, 1998; Macar, et al., 1994; Zakay, 1993b). These findings are consistent with SET, which predicts that divided attention will affect the closing of the switch, resulting in a lower pulse count for the duration recorded in the accumulator. When attention is divided during encoding, either delayed closing of the switch, or ‗flickering‘ of the switch, is predicted to lower the pulse count in the accumulator. Fewer pulse counts will result in underproduction of the

45 duration. However, assuming an accurate pulse count is stored when participants attend only to duration during encoding, divided attention during the reproduction task should result in overproductions of duration. Overproductions of duration are hypothesized to occur due to fewer pulses entering the accumulator during reproduction, requiring more time for the encoded pulse count to be reached.

Alternatively, this pattern of overproduction/underproduction can be explained by memory processes in SET. Some authors argue that this pattern of results is due to an overloading of working memory, which affects the stored mental representation of duration.

Some support for this hypothesis is provided by studies that show relatively accurate duration reproductions when the concurrent task involves only attentional processes and a pattern of underproductions/overproductions of duration when the amount of working memory required for the concurrent task is increased (Brown, 1997; Fortin, 1999; Fortin, Rousseau, Bourque, &

Kirouac, 1993).

In sum, the few studies that have investigated the production of isolated time intervals produce mixed support for SET. The strongest support for SET comes from the effects of divided attention or memory overload on the production of isolated time intervals. However, many studies investigating the production of isolated time intervals report that the variability of productions is not scalar, which is inconsistent with the predictions of SET. Comparatively, no support for entrainment theories of timing is provided by studies investigating the production, or reproduction, of isolated time intervals. In part, this may due to the long duration, non-periodic stimuli used for these tasks.

46

Other research has investigated the production of sequences of time intervals. This research has, typically, used either the spontaneous motor tempo (SMT) or synchronization- continuation tasks.

The SMT task has participants produce a series of taps (e.g., hand) at a rate they deem most comfortable (i.e., not too fast or too slow). Both interval and entrainment models of timing interpret SMT to reflect either the rate of the pacemaker or the endogenous period of the oscillator, respectively, yet there is some debate about what SMT measures (Boltz, 1994;

McAuley, Jones, Holub, Johnston, & Miller, 2006; Vanneste, Pouthas, & Wearden, 2001). Some support for SMT reflecting the rate of the internal clock is that SMT is fairly consistent within an individual over multiple recordings, SMT strongly correlates with an individual‘s perceptual ratings of preferred rate, and that SMT can be ‗sped up‘ or ‗slowed down‘ by inducing stress or relaxation, respectively (Boltz, 1994; McAuley, et al., 2006; Steinberg & Raith, 1985).

The most representative SMT of young adults is approximately 600-ms, but a range between 300 – 800-ms has been reported (Fraisse, 1982; McAuley, et al., 2006; Mishima, 1956;

Smoll & Schutz, 1978). One study reports that SMT is bimodally distributed, with peaks at approximately 300- and 500-ms (Collyer, Broadbent, & Church, 1994). Studies that have recorded multiple SMT measurements find a strong relationship between SMT measurements within an individual, usually between .75 – .95, suggesting that SMT is reliable intraindividually

(McAuley, et al., 2006; Rimoldi, 1951).

The synchronization-continuation task is also commonly used to investigate the production of sequences of time intervals. For this task, participants tap in synchrony with a metronome, set to a target time interval, and continue tapping at the target time interval after the tones stop. The main variables of interest for this task are the mean and standard deviation of

47 taps, or the inter-tap interval (ITI), during either the synchronization or continuation portion of the task.

The majority of research utilizing this task has focused on 1:1 synchronization for sequences of equal time intervals, or one tap for every tone (Repp, 2005a, 2005b). One variable of interest in this research has been the range of target time intervals to which participants can accurately synchronize. Naturally, the lower limit of this range is constrained by the effector used for the tapping task. For hand tapping, several studies report that the lower limit of synchronization is approximately 150 – 200-ms (Fraisse, 1982; McAuley, et al., 2006; Repp,

2003). The upper limit of synchronization is determined by the point at which participants begin to respond to the tones, rather than synchronizing. The upper limit ITI is approximately 2000-ms, but this rate limit varies between studies (Mates, et al., 1994; Miyake, Onishi, & Poppel, 2004).

While upper and lower limits exist for the synchronization-continuation task, one prediction of the interval theory-based W & K model is that within this range of target time intervals the produced ITI during continuation tapping should closely reflect the target time interval during synchronization. This prediction is based on the assumption that the clock can accurately time the target time interval, regardless of duration, and that no underestimations or overestimations of the target time interval occur during continuation tapping (Wing, 1980; Wing

& Kristofferson, 1973). Findings consistent with this hypothesis are often reported for young adults (Vorberg & Wing, 1996; Wing, 1980). Another prediction of the W & K model is that ITI variance should increase linearly with the target time interval. However, support for this prediction is not often found in young adults. Overall, variability in the production of the target time interval better adheres to a generalized version of Weber‘s law, which supports both the interval theory-based component analysis and entrainment models of timing. CV is

48 approximately equal for target time intervals between 250 – 2000-ms, but increases for target times outside of this range (Madison, 2001; Peters, 1989).

A number of other factors, beside the target time interval, have been shown to impair the accuracy of productions for sequences of time intervals. The regularity of IOIs within the sequence matters. Synchronization performance is more accurate for sequences of equal time intervals, compared to sequences of irregularly timed intervals (Repp, 2005a, 2005b). Similarly,

ITIs are less accurate, relative to the target time interval, when participants are instructed to tap in antiphase, or the midpoint between tones (Repp, 2005b). Finally, less accurate ITIs are found when participants must synchronize with visual sequences, as opposed to auditory sequences

(Kolers & Brewster, 1985; Repp & Penel, 2002, 2004).

In sum, studies that have investigated the production of sequences of time intervals report a few main findings. First, the most representative measure of young adult SMT is 600-ms.

Second, young adults can accurately perform the synchronization-continuation task, showing accurate reproductions of the target time interval during continuation, which is consistent with predictions of the W & K model. However, the generalized variability observed in young adults offers little support for the W & K model.

Perception of Time Intervals

The following section will briefly review the behavioral data for the perception of both isolated time intervals and sequences of time intervals. First, the perception of isolated time intervals are discussed, followed by a discussion of the perception of sequences of time intervals.

Figure 8a shows an example of a commonly used task to investigate the perception of isolated time intervals. For the isolated-interval discrimination task, participants are presented with two brief tones in which the interonset interval (IOI) of the tones mark a duration, termed

49 the standard interval. After a brief pause, another pair of tones mark the comparison interval.

Listeners judge the duration of the comparison interval, relative to the standard interval, responding ‗shorter‘ or ‗longer.‘ One dependent variable of interest for this task is the just- noticeable differences (JNDs) or the smallest change in duration necessary to detect a difference between the pair of intervals. Often, relative JNDs are reported to reflect JNDs as a percentage of the standard interval. This conversion makes for easier interpretation of JNDs and also provides a form of the Weber ratio. Relative JNDs for the perception of isolated time intervals are usually reported to be between 6 – 10 % of the standard interval duration (Abel, 1972; Allan, 1979;

Small & Campbell, 1962; Woodrow, 1951). Moreover, relative JNDs adhere to a generalized- version of Weber‘s law. Weber ratios are approximately equal for isolated time intervals between

200 – 2000-ms, but increase outside of this range (Abel, 1972; Getty, 1975; Treisman, 1963).

Another commonly reported finding in studies investigating the perception of isolated time intervals is that participants are sensitive to the temporal characteristics of their environment. The range of durations experienced within an experimental session can affect the perception of isolated time intervals, resulting in a systematic distortion in the time perception.

Specifically, the average duration experienced over the experimental session is perceived veridically, but shorter and longer durations are overestimated and underestimated, respectively

(Jones & McAuley, 2005; McAuley & Jones, 2003; McAuley & Miller, 2007). This finding is consistent with the broader psychophysics literature, which reports that participants develop a long-term memory representation for the ‗average‘ stimulus and base their judgments about a stimulus on this representation (see Helson, 1964 for a review of Adaptation-Level Theory).

Similarly, this finding is consistent with the central tendency effect (Hollingworth, 1910), in

50 which participants over-estimate stimuli smaller in magnitude than the average stimulus and under-estimate stimuli larger in magnitude than the average stimulus.

a) Standard Interval Comparison Interval

Pause b)

Standard Interval Comparison Interval

Standard Sequence Pause Comparison Sequence

Figure 8: Schematic of a a) isolated-interval discrimination task and a b) tempo discrimination task. Both tasks are the same, except that multiple repetitions of the standard and comparison interval are presented in the latter task, resulting in a standard and comparison sequence. For both tasks, participants judge whether the comparison interval is ‗shorter‘ or ‗longer‘ than the standard interval.

Studies investigating the perception of sequences of time intervals have, typically, used two main tasks: tempo discrimination or duration discrimination with rhythmic context. For a tempo discrimination task, listeners are presented with a sequence of IOIs in which either the duration of a single IOI, or a series of IOIs, is shortened or lengthened. Participants are asked to detect whether a change in the sequence to time intervals occurred. Figure 8b shows an example of a tempo discrimination task. The term ‗tempo-discrimination‘ is used for these tasks because

51 participants must make a judgment regarding the tempo (or rate) of the sequence(s) (Friberg &

Sundberg, 1995; Grondin, 2001a). Consistent with the findings reported in studies investigating the production of sequences of time intervals, studies investigating the perception of sequences of time intervals have generally found that sequences of equally timed intervals yield more accurate time judgments than sequences of irregularly timed intervals (Barnes & Jones, 2000;

Drake & Botte, 1993; Ivry & Hazeltine, 1995; McAuley & Jones, 2003; McAuley & Kidd, 1998;

Schulze, 1989). Furthermore, auditory sequences yield more accurate judgments of time than visual sequences (Grondin, 2001a; Grondin & McAuley, 2009).

Consistent with the multiple-look model, multiple repetitions of a duration have been shown to lower discrimination thresholds (Ivry & Hazeltine, 1995; McAuley & Jones, 2003;

McAuley & Kidd, 1998; McAuley & Miller, 2007; Miller & McAuley, 2005). This multiple- interval benefit can result in relative JNDs for sequences of time intervals being as low as 2%

(Drake & Botte, 1993; Friberg & Sundberg, 1995; Michon, 1964). Similar multiple-look benefits have also been reported for visual sequences (Grondin, 2001a).

However, there is a limit to the multiple-interval benefit. Drake and Botte (1993) tested this limit by presenting participants with standard-comparison sequences that contained either n

= 1, 2, 4, or 6 intervals (similar to Figure 8b). Lower discrimination thresholds were found when up to n = 4 intervals were presented in both sequences, but no additional benefit was observed with n = 6 intervals.

While the perception of isolated time intervals adheres to a generalized version of

Weber‘s law, the perception of sequences of time intervals does not hold to any form of Weber‘s law. Instead, relative JNDs are relatively constant for sequences of time intervals with IOIs of

300 – 900-ms, but relative JNDs increase outside of this range (Drake & Botte, 1993; Halpern &

52

Darwin, 1982; Hirsh, Monahan, Grant, & Singh, 1990; Michon, 1964; Schulze, 1989; ten

Hoopen, Boelaarts, Gruisen, Apon, & et al., 1994). This U-shaped function of relative JNDs across sequence IOIs suggests that there is an optimal range of sensitivity for the perception of sequences of time intervals.

One prediction of the multiple-look model is that only the number of standard intervals are responsible for the decrease in discrimination thresholds observed for the tempo discrimination task. However, studies have often covaried both the number of intervals in the standard and comparison sequences. To test the multiple-look model prediction, Miller &

McAuley (2005) independently varied the number of standard and comparison intervals to determine whether the multiple-interval benefit was due to the number of intervals in the standard sequence, comparison sequence, or both. Surprisingly, when the duration of the standard interval was fixed (did not vary) on every trial, the multiple-look advantage was due to multiple intervals in the comparison sequence, as opposed to the standard sequence. However, when the standard interval roved (varied from trial-to-trial), the multiple interval advantage was due to multiple intervals in both the standard and comparison sequences. Consistent with

Adaptation-level theory, the authors interpreted these findings to suggest that participants developed a stable referent memory for the standard interval when it was fixed from trial-to-trial, eliminating any multiple-interval benefit with a fixed standard interval. This finding is relevant for the experiments discussed in Chapters 5 and 6, as it is the basis for predictions regarding one of the conditions.

Consistent with studies investigating the perception of isolated time intervals, studies investigating the perception of sequences of time intervals also show that participants develop a general sense of the ‗pace‘ of their environment (Jones & McAuley, 2005; Miller & McAuley,

53

2005). In agreement with the central tendency effect, when a range of sequence IOIs (rates) are presented randomly over the experimental session, participants tend to gravitate to the mean rate of the session, showing veridical perception for the mean sequence IOI of the session, but overestimation and underestimation for shorter and longer sequence IOIs, respectively (Jones &

McAuley, 2005; Miller & McAuley, 2005). Moreover, participants are sensitive to the statistical properties of their temporal environment. The perception of sequences of time intervals is more accurate when a narrow range of sequence IOIs are presented within an experimental session, compared to a wide range of IOIs (Jones & McAuley, 2005; Miller & McAuley, 2005).

Other studies investigating the perception of sequences of time intervals have used the duration-discrimination with rhythmic context task (see Figure 1). While this task, and the main findings of studies utilizing this task, are discussed in Chapter 1, one additional finding from this task is relevant to understanding the perception of sequences of time intervals. Both interval and entrainment theories of timing predict the ∩-shaped expectancy profile reported for this task.

According to interval models, participants rely on a multiple-look mechanism that averages the

IOIs of the context rhythm and standard interval, but exclude the duration separating the context rhythm and the standard interval from their running average. An accurate representation of the standard interval will obtain for the on-time standard ending, but a less accurate representation will obtain for early- and late-standard endings. Alternatively, entrainment models predict that the oscillator entrains to the context rhythm. Entrainment to the context rhythm will result in synchrony between the oscillator and the on-time standard ending; however, early- and late- standard endings will require period correction for the oscillator to stay in synchrony with the unexpected standard ending. This period correction results in a distortion of the perceived duration of the standard interval.

54

In order to compare predictions by both interval and entrainment models, McAuley and

Jones (2003) halved the context IOI, but maintained the standard-comparison interval duration.

This manipulation resulted in a context rhythm that was twice as fast as the context rhythm previously discussed, but still shared a periodic relationship with the on-time standard ending.

According to interval theories, this manipulation should result in an inaccurate (i.e., very short) representation of the standard interval that will lead to a flat expectancy profile for the early, on- time, and late standard endings. However, entrainment models of timing predict the opposite effect on the expectancy profile. Again, a ∩-shaped profile is predicted, as the period of the oscillator will still be in temporal synchrony for the on-time standard ending, but not for the early- or late-standard endings. The effects of rhythmic context on perceived duration were still found when the faster context rhythm preceded the standard-comparison interval pair, as evident by the ∩-shaped expectancy profile, supporting the predictions from entrainment models of timing. This finding is difficult to explain from a simple averaging model proposed by interval theories (see McAuley & Jones, 2003 for a discussion).

The experiments reported in Chapters 5 and 6 will use the duration discrimination with rhythmic context task to test both the period-correction and DA-mediated distortion hypotheses.

Specifically, the period-correction hypothesis will examine whether expectancy profiles can be eliminated with multiple repetitions of a standard interval. This hypothesis derives from entrainment models of timing, which predict that multiple standard intervals should afford period correction to the standard interval, resulting in a flat (or at least weakened) expectancy profile.

In summary, several general findings emerge from the research investigating the perception and production of both isolated time intervals and sequences of time intervals. First, durations in the range of 300 – 2000-ms have, relatively, constant thresholds; however, durations

55 outside this range show increased thresholds. This pattern of findings best adheres to a generalized form of Weber‘s law, but the perception of sequences of timing intervals has been shown to have a zone of optimal sensitivity. Interestingly, this optimal zone of sensitivity corresponds to the average rates of SMT. Second, some evidence suggests that listeners are sensitive to the pace of their environment, as evident by them ‗picking up‘ on the average experienced duration during an experimental session. Third, the behavioral data reviewed offer mixed support for the main theories of timing. Generally speaking, the data obtained from studies investigating isolated time intervals provide support for interval theories; however, data obtained from studies investigating sequences of time intervals provide support for entrainment theories of timing.

Neural Bases of Duration Perception and Production

The research on timing has seen a boom of interest in locating the neural mechanisms involved in the perceptual and motor timing of both isolated time intervals and sequences of time intervals over the last twenty years. Two fields of research have provided insight into this issue.

Neuropsychological studies have provided investigated the effects of lesioned brain areas on timing. A, relatively, clearer approach has been the use of neuroimaging technology, such as magnetic resonance imaging (fMRI), Positron Emission tomography (PET), and electrophysiological measures to investigate timing. This review will focus on what is currently known about the neural mechanisms of timing through neuropsychological, fMRI, and PET studies. The review is limited to these approaches, as they have provided the most insight into both the cortical and subcortical areas involved in timing; however, see Macar and Vidal (2009) for a review of electrophysiological research on duration perception and production.

56

The following review will begin with a brief overview of the neural bases of perceptual and motor timing for both isolated time intervals and sequences of time intervals. Then, critical issues faced in this research will be discussed. Finally, a review of research on the production of isolated time intervals and sequences of time intervals and the perception of isolated time intervals and sequences of time intervals is given.

One early assumption of research investigating the neural bases of timing, which was also a concern for behavioral studies of timing, was whether the same brain areas were involved in the perception and production of duration. Several studies have provided consistent evidence that the same brain areas are involved in both the perception and production of time intervals, further supporting the common-timing mechanism hypothesis (Coull, Vidal, Nazarian, & Macar, 2004;

Lewis, Wing, Pope, Praamstra, & Miall, 2004; Rao, Mayer, & Harrington, 2001). Moreover, several neuroimaging studies of timing consistently report several cortical and subcortical regions identified as playing a role in timing processes. Among these areas are the basal ganglia

(BG; Larsson, Gulyas, & Roland, 1996; Rao, et al., 2001; Schubotz, Friederici, & von Cramon,

2000), cerebellum (Jueptner, Flerich, Weiller, Mueller, & Diener, 1996; Jueptner, et al., 1995), supplementary motor area (SMA; Jancke, Loose, Lutz, Specht, & Shah; Penhune, Zatorre, &

Evans, 1998; Rao, et al., 2001), premotor cortex (PMC; Brunia, de Jong, van den Berg-Lenssen,

& Paans, 2000; Schubotz, et al., 2000), and prefrontal regions (PFC; Lewis & Miall, 2002; Rao, et al., 2001; Rubia, et al., 1998). Converging evidence is reported by neuropsychological studies, with lesions to the BG (Grahn & Brett, 2009; Malapani, Deweer, & Gibbon, 2002; Malapani, et al., 1998; O'Boyle, et al., 1996), cerebellum (Ivry & Keele, 1989; Spencer, Ivry, & Zelaznik,

2005), SMA (Halsband, Ito, Tanji, & Freund, 1993), PMC (Halsband, et al., 1993), and PFC

(Mangels, Ivry, & Shimizu, 1998) resulting in timing impairments. Taken together, these

57 findings strongly suggest that while a common timing mechanism exists, one brain area is not uniquely responsible for timing. Instead, the perception and production of duration requires interactions between different brain areas, with each area potentially providing a different role in timing processes.

Additional evidence suggests that timing is lateralized in the brain. Both neuroimaging and neuropsychological studies report that the right hemisphere is involved in the perception and production of both isolated time intervals and sequences of time intervals (Drane, Lee, Loring, &

Meador, 1999; Funnell, Corballis, & Gazzaniga, 2003; Handy, Gazzaniga, & Ivry, 2003;

Harrington, Haaland, & Knight, 1998; Vidalaki, Ho, Bradshaw, & Szabadi, 1999). Interestingly, right hemispheric lateralization holds for both auditory and visual stimuli, suggesting that this lateralization is not specific to audition (Alpherts, et al., 2002; Grondin & Girard, 2005).

New imaging technologies have led to significant advances in the identification of the brain areas involved in the perception and production of both isolated time intervals and sequences of time intervals. Yet, there is still considerable debate over the specific brain areas involved these processes and the role each area might play in the perception and production of time intervals. Many of the same critical issues discussed for the behavioral study of timing underlie this debate. Specifically, interpretation of this data must involve careful consideration of both the duration(s) and the required processes for the timing task under investigation. These issues are discussed in the next section.

Critical Issues in Determining the Neural Bases of Timing

Timescale. Like behavioral studies, research examining the neural bases of timing has investigated a range of durations, spanning milliseconds – seconds. However, unlike the timescale proposed in this dissertation, which considers the range of durations over which the

58 perception and production of sequences of time intervals is stable, much of the research examining the neural bases of timing has focused on subsecond durations or suprasecond durations (Harrington, Lee, Boyd, Rapcsak, & Knight, 2004; Ivry, 1996; Lewis & Miall, 2003a,

2003b). This boundary is often used because 1) differences in neural activity during timing tasks are often observed around 1000-ms (see Grondin, 2010 for a review; Harrington, Lee, et al.,

2004; Ivry, 1996; Lewis & Miall, 2003a, 2003b) and 2) most of the research examining the neural bases of timing has considered only the perception and production of isolated time intervals.

Although controversial, two areas that have been reported to show duration-dependent involvement in timing processes, through both neuropsychological and neuroimaging techniques, are the cerebellum (Ivry, 1996; Lee, et al., 2007; Lejeune, 1998; Lewis & Miall, 2003b; Nichelli,

Alway, & Grafman, 1996) and the BG (Harrington, Boyd, et al., 2004; Jahanshahi, Jones,

Dirnberger, & Frith, 2006; Jueptner, et al., 1995; Pouthas, et al., 2005; Rao, et al., 2001). No duration dependent involvement of the frontal cortex (Koch, Oliveri, Carlesimo, & Caltagirone,

2002; Koch, Oliveri, Torriero, & Caltagirone, 2003; Lewis & Miall, 2006), SMA (Ferrandez, et al., 2003; Jahanshahi, et al., 2006; Kudo, et al., 2004; Tregellas, Davalos, & Rojas, 2006), and parietal cortex (Bueti, Bahrami, & Walsh, 2008) has been reported for the perception and production of duration. One proposal for the time-dependent nature of the cerebellum and BG was offered by Ivry and colleagues, who proposed that the cerebellum is involved in timing durations in the subsecond range, while the BG is involved in timing durations in the suprasecond range. (Diedrichsen, Ivry, & Pressing, 2003; Ivry, 1996). However, later studies found that both the BG and cerebellum are involved for timing millisecond – seconds durations, suggesting that these areas are not duration dependent (Belin, et al., 2002; Harrington, Boyd, et

59 al., 2004; Harrington & Haaland, 1999; Jahanshahi, et al., 2006; Jueptner, et al., 1995; Koch, et al., 2007; Lewis & Miall, 2003a; Pouthas, et al., 2005; Tracy, Faro, Mohamed, Pinsk, & Pinus,

2000).

Another timescale issue is that timing durations longer than 1000-ms might recruit additional cognitive processes, such as memory and attention. Consistent with this hypothesis, activation of brain areas associated with memory, attention, and other cognitive processes have been found for durations longer than 1000-ms (Coull, et al., 2004; Harrington & Haaland, 1999;

Rao, et al., 2001). We will return to this issue in the following section.

Time perception and production tasks. In addition to the issue of timescale, a variety of different timing tasks are employed by neuropsychological and neuroimaging studies of timing. Related to the timescale issue just discussed, the primary concern with some timing tasks is that additional, non-timing specific, cognitive processes might be required to perform the task.

Many authors have questioned whether certain timing impairments or areas of activation better reflect attention or working memory processes, rather than timing per se (Buhusi & Meck, 2005;

Harrington, Haaland, & Knight, 1998; Rao, et al., 2001). For example, from an interval perspective, perceptual tasks investigating long durations may require sustained attention during encoding of duration and working memory to temporarily store the duration for later comparison.

Consistent with this hypothesis are reports of activation of some brain areas thought to be involved in attention and working memory during timing tasks. Both neuroimaging and neuropsychological studies show involvement of the PMC, parietal, and frontal cortices for the perception of isolated time intervals (Coull, 2004; Harrington & Haaland, 1999; Rao, et al.,

2001).

60

Strong support for the role of attention in the perception of isolated time intervals comes from research showing that attention to duration can modulate levels of brain activation. In an elegantly designed fMRI study, Coull, Vidal, Nazarian, and Macar (2004) varied the level of participant‘s attention to either the color attributes of a square or the duration the square was presented and had participants make a perceptual judgment about the attended dimension. Task difficulty for the visual stimulus and duration was equated. Increasing allocation of attention to either stimulus showed an increase in brain activity; activity in the visual area (V4) systematically increased with more allocation of attention to the visual stimulus and similar increases in the activation of the cortico-striatal loop (i.e., pre-SMA, PMC, PFC, temporal cortex, intraparietal sulcus, and putamen) were observed when attention was directed to duration.

Moreover, the pattern of neural activation was supported by behavioral measures. Systematic decreases in reaction time were found as the level of neural activation for the to-be-attended stimulus increased.

As already seen in this brief review, several different brain areas are involved in the perception and production of isolated time intervals and sequences of time intervals. However,

Lewis and Miall‘s (2003b) proposed distinction between automatic and controlled timing systems clarifies the potential involvement of different brain areas, based on the requirements of the time production or perception task. Tasks that recruit the automatic timing systems tend to show activation of motor areas, such as the SMA, sensorimotor cortex, PMC, and the BG, but not brain areas believed to be involved in attention and memory, such as the PFC and parietal cortex (Buhusi & Meck, 2005; Lewis & Miall, 2003b). Comparatively, tasks that recruit the cognitively-controlled system often find activation of the PFC, parietal cortex, cerebellum, BG, and SMA, which supports the proposal that attention and memory are necessary to perform these

61 tasks (Lewis & Miall, 2003b). It should be noted that certain brain areas show activation in tasks that recruit both timing systems, such as the SMA, BG, and cerebellum. Two potential explanations for the redundancy of activation in these areas are that 1) this activation might reflect general processes required for the perception and production of time intervals or 2) the tasks used to study the perception and production of time intervals are unable to differentiate the roles of these brain areas in timing (Buhusi & Meck, 2005).

Based on the data reviewed and the work of Coull et al. (2004), there is substantial support for the hypothesis that the perception of isolated time intervals relies on the cognitively- controlled timing system. However, it is less clear whether tasks that require the perception of sequences of time intervals rely on the automatic or cognitively-controlled timing system. On one hand, the perception and production of sequences of time intervals appear to rely solely on the automatic timing system. In both the perception and production of sequences of time intervals, future durations are generally predictable. Furthermore, both the perception and production of sequences of time intervals lend themselves to motor movement, either directly or through ‗tapping along to the rhythm.‘ Yet, perceptual and motor timing for sequences of time intervals is often stable for durations lasting longer than 1000-ms (i.e. 300 – 2000-ms) and, more than likely, requires some form of attentional processes (Grahn & Brett, 2007).

Neural Mechanisms Recruited for the Production of Time Intervals

The following section will briefly review the neural data for the production of both isolated time intervals and sequences of time intervals. First, the production of isolated time intervals is discussed. Second, the production of sequences of time intervals is discussed.

Early studies investigating the neural bases of the production of isolated time intervals relied heavily on neuropsychological patient groups. Timing impairments were often reported for

62 individuals with PD and cerebellar lesions, leading many to suggest that a single timing mechanism might be located in either brain area (Ivry & Keele, 1989; Pastor, Artieda, et al.,

1992). However, inconsistent reports of timing impairments in either individuals with PD or cerebellar lesions made it difficult to tease apart which brain area was responsible for duration production, or whether these two brain areas played separate roles in duration production.

Subsequent research has shown that lesions to the frontal cortex can also lead to impairments in the production of duration (Nichelli, Clark, Hollnagel, & Grafman, 1995). Taken together, these findings suggest that several brain areas may be recruited during the production of isolated time intervals (Harrington & Haaland, 1999).

Compared to the neuropsychological research on the production of isolated time intervals, few studies have examined the production of isolated time intervals using neuroimaging techniques. However, of the few available studies, activation of the SMA, PFC, temporal gyrus, and cerebellum have been reported (Brunia, et al., 2000; Lewis & Miall, 2003b;

Tracy, et al., 2000).

A commonly used task to investigate the neural bases of the production of sequences of time intervals in individuals with neuropsychological deficits has been the synchronization- continuation task. This task has also been used by neuroimaging studies to assess the neural mechanisms involved in the production of sequences of time intervals. Findings across studies employing both techniques are rather consistent.

Compared to controls, performance impairments on the synchronization-continuation task have been observed in individuals with lesions of the cerebellum (Ivry & Keele, 1989;

Spencer & Ivry, 2005; Spencer, Zelaznik, Diedrichsen, & Ivry, 2003), SMA (Halsband, et al.,

1993), PMC (Halsband, et al., 1993), and the superior temporal gyri (Penhune, Zatorre, &

63

Feindel, 1999). Additionally, individuals with PD tend to perform worse on this task, compared to controls, suggesting the involvement of the BG in the production of sequences of time intervals (Duchek, Balota, & Ferraro, 1994; Elsinger, et al., 2003; Harrington, Haaland, &

Hermanowicz, 1998; O'Boyle, et al., 1996; Yahalom, Simon, Thorne, Peretz, & Giladi, 2004).

Unfortunately, no consistent pattern of errors is found for a specific lesion. Lesions of the SMA,

PMC, and superior temporal gyri do not result in a pattern of impairments that separates one lesion group from the others. While some studies report that individuals with PD show higher clock and motor variability on the task when the W & K model is applied to the tapping data, these findings are mixed (Ivry & Keele, 1989; O'Boyle, et al., 1996; Pastor, Artieda, et al., 1992).

Neuroimaging studies utilizing tasks that require the production of sequences of time intervals report converging evidence with the findings of neuropsychological studies. Commonly reported areas of activation during the performance of this task are the BG, cerebellum, SMA,

PFC, PMC, and superior temporal gyri (Lewis & Miall, 2003b; Lewis, et al., 2004; Penhune, et al., 1998; Schubotz, et al., 2000). The areas activated during this task do not appear to be rate

(duration) dependent for sequences of IOIs between 300 – 600-ms, but these are the only durations that have been directly compared (Rao, et al., 1997). Furthermore, studies comparing activation during both the synchronization and continuation portions of the task provide conflicting results. One study reported no differences in activation between either portion of the task (Jancke, et al., 2000), but another study reported additional activation of the premotor loop,

SMA, putamen, and thalamus during the continuation portion of the task (Rao, et al., 1997).

64

Neural Mechanisms Recruited for the Perception of Time Intervals

The following section will briefly review the neural data for the perception of both isolated time intervals and sequences of time intervals. First, the perception of isolated time intervals is discussed. Second, the perception of sequences of time intervals is discussed.

Neuropsychological studies investigating the perception of isolated time intervals have focused primarily on individuals with PD and cerebellar lesions. Although the data are mixed, impairments have been reported for both neuropsychological patient groups, suggesting that the

BG and cerebellum are involved in the perception of isolated time intervals (Harrington,

Haaland, & Hermanowicz, 1998; Ivry, Keele, & Diener, 1988; Jueptner, et al., 1995).

Additionally, right hemispheric damage, specifically to the frontal and parietal cortices, has been shown to cause impairments in the perception of isolated time intervals (Harrington, Haaland, &

Knight, 1998).

Neuroimaging studies have provided a somewhat clearer answer to the role of specific brain areas in the perception of isolated time intervals. Studies show rather consistently that the following brain areas are activated during these tasks: SMA, PMC, frontal operculum, PFC, parietal cortex, BG, and cerebellum (Lewis & Miall, 2003b). While the exact role of these brain areas in the perception of isolated time intervals is not fully understood, researchers have made hypotheses regarding the role each area might play in timing. The BG are hypothesized to be involved in encoding duration or the location of the clock (Meck, Hinton, & Matell, 1998; Meck

& Malapani, 2004; Rao, et al., 2001). The frontal operculum has been hypothesized to be involved in the process of time estimation (Coull, et al., 2004; Lewis & Miall, 2003b; Schubotz

& von Cramon, 2001). Some have argued that activation in the SMA and PMC reflects the initiation of a response, but studies that control for a motor response still report activation of this

65 area, making the role of these areas unclear (Coull & Nobre, 1998; Harrington, Haaland, &

Knight, 1998; Lewis & Miall, 2003a, 2003b). Finally, the PFC and parietal cortex are hypothesized to be involved in working memory and/or decision processes and temporal attention, respectively (Coull, et al., 2004; Harrington & Haaland, 1999; Harrington, Haaland, &

Knight, 1998; Rao, et al., 2001).

In a seminal study, Rao, Mayer, & Harrington (2001) utilized an event-related fMRI to separate the involvement of different brain areas in the duration encoding process, comparison process, and response generation for an isolated-interval duration discrimination task. Based on predictions of SET, scans were time-locked to what the authors believed were the separate stages of isolated-interval discrimination. Comparisons of neural activation were made between a isolated-interval duration discrimination task (standard IOI = 1200-ms), a pitch-discrimination task, and a sensorimotor control task that required a button press after passively listening to equally timed tone pairs of identical pitch. Through an elaborate series of subtractions for brain activity during the three tasks, higher levels of activation were observed in the BG and the inferior parietal cortex at the beginning of the trial, leading the authors to suggest that these areas are involved in encoding duration. In accord with SET, they propose that the BG is the ‗clock‘ and that the inferior parietal cortex is involved in attention processes and pulse collection.

Activation of the cerebellum only occurred at the end of the trials, leading the authors to suggest that this area is involved in time perception, but not the encoding of duration. Additional activity in the PMC and PFC was also observed at the end of trials, leading the authors to suggest that these areas are used to carry out the memory and decision stages described by SET, respectively.

Compared to the research on neural bases of the perception of isolated time intervals, only a few studies have examined the neural bases of the perception of sequences of time

66 intervals. Grahn & Brett (2009) compared individuals with PD and older adult controls on a perceptual discrimination task in which participants were presented with both metrically simple and complex rhythmic sequences. The IOIs that made up both rhythmic tone sequences shared simple integer-ratios (1:2:3:4). The key manipulation was the grouping of the IOIs. For the metrically simple rhythms, the IOIs were grouped in a manner that gave rise to an easily perceivable beat, or pulse. In other words, the IOIs of metrically simple rhythms were grouped in a manner similar to the metrical structure of Western music. The IOIs that made up the metrically complex rhythms shared the same interval ratios; however, these IOIs were grouped so that it was difficult to abstract a simple beat from the sequences. It should be stressed that both the sequences shared the same IOI ratios; however, the main manipulation was the ease with which a regular beat could be perceived. Metrically complex rhythms had no regular reoccurrence of a beat. Participants heard a standard rhythm and a comparison rhythm and judged whether they were the ‗same‘ or ‗different.‘ For the metrically simple sequences, lower discrimination thresholds were observed in individuals with PD, compared to older adults.

However, no differences in discrimination thresholds were found between the groups for the metrically complex rhythms. These findings suggest that the BG is involved in the perception of sequences of time intervals. However, the null difference between individuals with PD and older adults suggests that the damage to the BG, caused by PD, does not completely interfere with the perception of sequences of time intervals.

Converging evidence for the involvement of the BG in the perception of sequences of time intervals also comes from fMRI studies. Grahn and colleagues have consistently shown activation of both the BG and motor areas for quite different perceptual timing tasks (Grahn &

Brett, 2007; Grahn & McAuley, 2009). An fMRI study utilizing the task described in the

67 previous paragraph found activation in the BG, anterior superior temporal gyri, inferior frontal gyrus and pre-SMA/SMA for metrically simple rhythms. Moreover, Grahn & McAuley (2009) conducted an fMRI study in which participants were presented with both a typical isolated- interval discrimination task and a rhythmically ambiguous tone sequence that could be perceived either as an isolated time interval or as a sequence of time intervals. Further supporting the involvement of the motor areas in the perception of sequences of time intervals, greater activation of the SMA, PMC, and insula were observed for individuals who were more likely to pick up on the rhythm of the sequence, compared to those who heard the sequence as an isolated- time interval.

Taken together, Grahn and colleagues show that similar brain areas found to be active during the production of sequences of time intervals are activated during the perception of sequences of time intervals. Additionally, their findings also offer supporting evidence for the involvement of the BG and motor areas in the perception of sequences of time intervals. This latter finding is not surprising, as the BG and SMA have strong connections through striato- thalamocortical loops (Alexander, Delong, & Crutcher, 1992). Finally, it should be noted, that the involvement of the motor areas during the perception of sequences of time intervals suggest that the activity observed in motor areas during the production of sequences of time intervals tasks is not solely due to the motor requirements of the production task. Motor responses only occurred at the end of the trial for perceptual tasks, and response-related activity could be subtracted out with the appropriate control task.

Support for the role of the BG in the perception of sequences of time intervals is relevant for this dissertation. The experiment described in Chapter 6 tests one hypothesis regarding the role of the BG and dopamine (DA) in the perception of sequences of time intervals. Based on the

68

PD-related impairments in the perception of sequences of time intervals reported by Grahn and colleagues, individuals with PD were predicted to show larger effects of rhythmic context on perceived duration in the experiment described in Chapter 6.

This brief review of the neural bases of perceptual and motor timing for both isolated time intervals and sequences of time intervals highlights two main findings. First, many of the same brain areas are involved in the perceptual and motor timing of both isolated time intervals and sequences of time intervals. These findings suggest that a basic timing mechanism might exist, which involves the cerebellum, BG, SMA, PMC, and PFC. However, second, other brain areas may play independent roles in timing, based upon which timing system is recruited for the timing task. This finding is consistent with the proposal of the cognitively-controlled and automatic timing systems. Additional support for this proposal is provided by the involvement of brain areas thought to be involved in attention, memory, and other cognitive processes for the perceptual and motor timing of isolated time intervals, which is consistent with the proposed cognitively-controlled timing system. Similarly, the involvement of motor areas for the perceptual and motor timing of sequences of time intervals is, potentially, consistent with the proposed automatically-controlled timing system. Finally, neural support for the role of the BG in timing is quite strong. Not only is the involvement of the BG implicated across studies that have varied the durations and tasks under investigation, but clear anatomical connections with motor areas make it likely that the BG plays an important role in the timing of sequences of intervals.

69

CHAPTER 4: EFFECTS OF AGING AND PD ON THE PERCEPTION AND PRODUCTION OF DURATION

Aging and Timing

In youth the days are short and the years are long; in old age the years are short and the days

long. ~Nikita Ivanovich Panin

Age-related changes in the perceptual and motor timing of isolated time intervals and sequences of time intervals have generated a considerable amount of interest within the field of timing research.5 In part, this interest has been fueled by older adult‘s contradictory statements about their experiences with time. Older adults often report that ―the seasons come around quicker every year,‖ and, yet at other times, ―the minutes pass like hours‖— providing sentiments similar to those expressed in the Panin quote. Surprisingly, these seemingly paradoxical notions are consistent with the research findings regarding age-related changes in timing. As we will see later, this time paradox can be resolved. Separate from the time paradox of aging, age-related changes in timing are frequently reported (Block, Zakay, & Hancock, 1998;

Lustig, 2003; McAuley, et al., 2006; Vanneste, et al., 2001; Wearden, 2005). Many researchers have become interested in understanding these age-related changes, as they might provide insight into how cognitive processes, such as attention and memory, play a role in timing.

The section of age-related changes in timing is organized in the following manner. First, theoretical predictions concerning age-related changes in timing from both interval and entrainment models are discussed. Second, behavioral data of perceptual and motor timing for both isolated time intervals and sequences of time intervals are discussed. The purpose of these subsections is to clarify age-related changes in timing, so that specific timing impairments in PD can be discussed in the latter half of this chapter.

70

Theoretical Perspectives on Aging and Timing

Both interval and entrainment theories of timing generate predictions regarding changes in timing across the lifespan. However, only predictions concerning older adults will be discussed in this dissertation.6

Both interval and entrainment theories attribute the ability to time event duration(s) to an internal ‗clock.‘ Generally speaking, both theoretical perspectives predict that the clock slows with age. This general prediction fits well within traditional theories of cognitive slowing used to explain cognitive decline in older adults; however, it should be noted that cognitive slowing theories refer to a slowing of processing speed, as opposed to the clock (Cerella, 2001;

Salthouse, 1996).

Interval and entrainment theories differ most in their explanations for why age-related changes in timing occur. Interval theories of timing primarily hypothesize that age-related changes in timing result from impairments of attention or memory. In contrast, entrainment theories hypothesize that age-related changes in timing result from systematic changes in the preferred rate of events and the range of rates that are accessible in older adulthood.

Interval approaches, primarily scalar expectancy theory (SET), attribute age-related changes in timing to impairments of either attention or memory. Since attention and memory impairments are commonly reported in the cognitive aging literature, many researchers have used this model to investigate whether aging affects these cognitive processes for timing (Craik

& Jennings, 1992; Jennings & Jacoby, 1993). While SET does not argue that these are the only stages of the model susceptible to aging effects (e.g., decision processes), few detailed predictions have been made regarding these other processes.

71

One prediction of SET is that aging affects the clock stage of the model. An age-related slowing of the clock is predicted to occur, resulting in the collection of fewer pulses for a to-be- timed event. Or, attentional impairments can affect the closing of the switch, or result in a

‗flickering‘ of the switch, which leads to the accumulation of fewer, or missed, pulses for a to- be-timed event. Unfortunately, these two hypotheses are not clearly disambiguated in the literature, as they make similar predictions.

Another prediction of SET is that aging affects the memory stage of the model.

Specifically, distortions in the memory for a time interval are predicted. Memory distortions can result from 1) a loss of pulses, in either reference or long-term memory, or 2) through mixing of memories for two, or more, durations stored in reference memory.

Entrainment theories attribute age-related changes in timing to both a slowing of the clock and a narrowing of the range of accessible rates in older adulthood (Drake, Jones, &

Baruch, 2000; McAuley, et al., 2006). Remember that entrainment approaches are concerned with how internal periodic activity, such as an oscillator(s), synchronizes with a driving rhythm.

The oscillator(s), in the latent state, is proposed to maintain a regular periodicity, termed P0.

When a driving rhythm is present, the periodicity of P0 is driven to synchronize with the driving rhythm, or entrain (period match).

Entrainment theories generate two hypotheses concerning age-related impairments in timing. First, the preferred period hypothesis predicts that the latent oscillator will slow in older

adulthood. Specifically, P0 is predicted to slow with age, thereby slowing the clock. The second prediction from this approach concerns the range of rates accessible to older adults. The entrainment region hypothesis predicts that the range of accessible rates is rather narrow in early childhood, widens during late childhood through middle adulthood, and, again, narrows in older

72

adulthood. In more formal terms, narrower period matching limits between P0 and the driving rhythm are predicted for older adults. Target times outside the entrainment region are predicted to show systematic distortions in the perception and production of duration.

Data of Aging and Timing

The review of the behavioral data for aging and timing is organized in the following manner. First, I review the findings from studies investigating the production of both isolated time intervals and sequences of time intervals in older adults. Second, I review the behavioral data from studies investigating the perception of both isolated time intervals and sequences of time intervals in older adults.

Production of time intervals. Research findings from studies investigating age-related changes in the production of isolated time intervals has often been interpreted within the context of interval models of timing, specifically SET. In part, this is due to the clock and memory stages of SET providing simple predictions regarding the production of isolated time intervals by older adults. As the quote at the beginning of the chapter suggests, the experience of time in older adults appears to be paradoxical. Early studies investigating age-related changes in the production of isolated time intervals provided results consistent with this paradox. Some research suggested that the clock speeds up with age, while other research suggested that the clock slows down with age (Cohen, 1967; James, 1890). However, application of SET to this data provides a potential explanation for the mixed results of earlier research.

The appearance of mixed results in early research investigating age-related changes in the production of duration for isolated intervals can be explained by the fact that these studies did not clearly distinguish between the use of retrospective and prospective tasks (Block, et al.,

1998). Further complicating the issue was that many early studies had older adults perform one

73 of three tasks: verbal estimation, production, or reproduction. Distinguishing these types of tasks is very important, as the slow-clock hypothesis provided by SET provides very different predictions for these tasks. SET predicts that a slowed clock in older adults will lead to underestimation of the duration of isolated time intervals, but overproduction of the duration of a time interval. Underestimation is predicted to occur because the slower clock rate will lead to fewer pulses being collected in the accumulator during a to-be-timed event. On the other hand, overproduction is predicted because it will take longer to reach the appropriate pulse count for the to-be-timed event. This pattern of underestimation and overproduction, especially for interval theories of timing, is the diagnostic marker of a slower clock. However, it should be mentioned that these predictions assume no correction for changes in clock rate occur (e.g., that the system does not readjust itself for the new clock rate), nor that older adults compensate for their slowed clock (e.g., adjust their responses so that more or less pulses, respectively, accumulate).

Block, Zakay, & Hancock (1998) conducted a meta-analysis of early studies that compared young- and older-adults on duration production tasks. The main variable they considered in this meta-analysis was the ratio of subjective-to-objective time between the two groups. This ratio allowed for comparisons across duration estimation and production tasks—for duration production tasks, the objective time was the participant‘s production and the subjective time was the duration they were instructed to produce (or reproduce). The main finding from this meta-analysis was that older adults had a higher ratio, compared to young adults. This finding is somewhat surprising in that it suggests the opposite pattern of findings predicted by the slowed clock hypothesis of SET—older adults overestimated and underproduced durations. Moreover, they found that age-related effects in the production of duration were limited to certain duration production tasks. While no differences were found between young- and older-adults for

74 reproduction tasks, differences were found for verbal estimation and production tasks. The latter finding is even more surprising, as later studies show age effects for reproduction tasks.

However, as the authors note, one should be skeptical of this finding, as there were only three studies in this meta-analysis that used a duration reproduction task and small effect sizes were observed in all three studies.

The findings of Block, Zakay, & Hancock (1998), at first, seem to conflict with slowed clock hypothesis. However, interval models, specifically SET, provide a reasonable account for this disparity. SET predicts that attentional processes affect the closing of the switch between the pacemaker and the accumulator. With normal attention, the switch will close at the beginning of a to-be-timed event and remain closed until the end of the to-be-time event. However, impairments of attention are predicted to result in either a delay of the switch closing, or a

‗flickering‘ of the switch, both of which result in some pulses being missed between the pacemaker and accumulator.

When interpreted through SET, the findings of Block, Zakay, & Hancock (1998) provide some support for the attention impairment hypothesis. Many of the studies reviewed in the meta- analysis used tasks that required participants to either verbally estimate or produce a duration tied to a specific clock unit. Since these tasks required participants to estimate or produce durations based on a verbal label of time, they may have relied on their memories for these durations, which were formed outside the lab. Attentional distractions in the outside world may have resulted in missed pulses and, hence, inaccurate memories were formed. However, in the lab, no such distractions were present. Participants might have been relying on a ‗faster‘ clock when performing these tasks. According to this hypothesis, differences in attention to time may explain why time appears to go faster for older adults in the Block, Zakay, and Hancock study.

75

Stronger support for attentional impairments affecting productions of isolated intervals by older adults comes from studies investigating the role of divided attention on duration production. Craik and Hay (1999) compared young- and older-adult performance on a verbal estimation and production task. Participants performed both tasks while making judgments about visual stimuli. When attention was divided between timing and non-timing tasks, the pattern of performance reported by Block, Zackay, & Hancock (1998) switched; both younger- and older- adults underestimated and overproduced durations during the experiment. Although both groups showed the general slow clock pattern, the effect was more pronounced for older adults. These findings suggest that when attention is divided between timing and non-timing tasks, older adults show a pattern of results consistent with the slow clock hypothesis. Moreover, these findings support the attentional impairment hypothesis in older adults. When older adults are paying full attention to time, ―time flies;‖ however, when attention to time is limited, time moves rather slowly.

Interval theories have been useful for explaining the patterns of data reported for the production isolated time intervals. However, both interval and entrainment theories make predictions regarding the production of sequences of time intervals.

A shared prediction from both theoretical perspectives of timing is that the clock slows with age. The primary task used to assess the rate of the ‗internal clock‘ is spontaneous motor tempo (SMT). For this task, individuals tap their hand at a rate they deem most comfortable (i.e., not too fast or too slow). A consistent finding across studies examining SMT in both young- and older-adults is that SMT slows with age. While studies report the SMT of young adults to be approximately 600-ms, SMT for older adults is between 700 – 1100-ms, which further supports

76 the slowed clock hypothesis (Baudouin, Vanneste, & Isingrini, 2004; McAuley, et al., 2006;

Vanneste, et al., 2001).

McAuley et al. (2006) provide additional support for the slowed-clock hypothesis. In addition to SMT, the measure of preferred perceptual tempo (PPT) was obtained from participants ranging in age from 4 – 95 years of age. The PPT task is similar to SMT, except that participants made a scaled perceptual rating about whether the rate of a tone sequence was ‗too fast,‘ ‗just right,‘ or ‗too slow.‘ McAuley et al. (2006) had participants in all age groups judge the rate of equally timed tone sequences presented over a wide range of IOIs. Consistent with previous research, PPT for young adults was approximately 600-ms (McAuley, et al., 2006;

Mishima, 1956); however, older adults were found to have a slower PPT of approximately 700- ms.

The majority of research on age-related changes in the production of sequences of time intervals has come from studies that employ the synchronization-continuation task. One advantage of this task is that both the W & K and the component analysis model can be applied to the data to assess the effects of aging on clock estimates.

Two studies have compared young- and older-adult performance on the synchronization- continuation task for a target time interval of 550-ms (Greene & Wiliams, 1993; Ivry & Keele,

1989). Ivry and Keele (1989) found that, overall, older adults produced significantly longer inter- tap intervals (ITIs ; 550.1-ms) compared to younger adults (535. 4-ms). Additionally, older adults were significantly more variable in their continuation ITIs than young adults. Application of the W & K model to their data revealed that the increase in variability observed in older adults was attributable to increased variability in the clock; no motor variability differences were found between the groups.

77

A similar pattern of results were reported by Greene and Williams (1993). No age-related differences were found in mean ITI during the continuation portion of the task between young- and older-adults. Similarly, no statistical differences in variability were found between groups; however, group differences in variability were quite large (20 – 50 % greater in older adults).

The W & K model revealed no statistical differences between young- and older-adults, but clock variance was markedly higher for the older adults. Taken together, these two studies suggest that older adults do show age-related changes in their ability to continue tapping the target time interval in the absence of pacing tones. Additionally, the high clock, but not motor, estimates of variability found for older adults support the general hypothesis that age-related changes occur for the production of sequences of time intervals.

However, conflicting results are reported by Vanneste, Pouthas, and Wearden (2001).

Young- and older-adults performed the synchronization-continuation task at target time intervals between 300 – 700-ms, in 100-ms increments. No differences in mean ITI or variability were found between the groups. Additionally, no differences were found in the clock or motor estimates of variability from either the W & K or component analysis models, suggesting no differences between young- and older-adults in the production of sequences of time intervals. It should be noted that some older adults yielded data that gave estimates of negative motor variability, which makes these estimates questionable. 7

McAuley et al. (2006) conducted the most extensive study of age-related changes in the production of sequences of time intervals. Among other tasks, participants performed the synchronization-continuation task for a wide range of target time intervals: 150-, 225-, 337-,

506-, 759-, 1139-, and 1709-ms. The W & K and component analysis models were also applied to their tapping data, but yielded uninterpretable negative estimates of motor variability. In

78 addition to these models, participants ITIs during continuation tapping were analyzed for drift, or a systematic change toward faster or slower tapping. Interestingly, older adults showed greater drift toward their SMT for the slowest target time intervals, compared to young adults. The greater drift toward SMT found for older adults supports entrainment theories, suggesting that the slower rates were outside the entrainment region for older adults. Since older adults were unable to entrain to the slow rates, their continuation tapping drifted back toward their SMT.

In sum, the available data on the production of duration for isolated time intervals and sequences of time intervals in older adults is mixed. One relatively consistent finding is that older adult performance on time production tasks favors the slow clock hypothesis. Another consistent finding is that age-related impairments of attention are, at least, partially responsible for age-related changes in the production of isolated time intervals. This finding is consistent with the predictions of SET. Studies of duration production for sequences of time intervals report mixed results. Some studies provide evidence for an age-related change in the ‗clock,‘ but other studies report no differences between younger- and older-adults for the production of sequences of time intervals. Finally, when participants must produce sequences of time intervals over a wide range of rates, older adults have difficulty maintaining slow rates compared to young adults, which offers some support for entrainment theories.

Perception of time intervals. In addition to the research on the production of duration, a few studies have investigated the perception of both isolated time intervals and sequences of time intervals in older adults. One benefit to the study of the perception of duration is that these tasks do not require a verbal label of time to be assigned to a duration. Instead, a duration(s) is learned and tested in an experimental session, with no mention of clock-based time units. Thus,

79 participant‘s previous experience of a given duration should not affect performance on these tasks.

Some research has compared young- and older-adult performance on tasks that require a comparative perceptual judgment about the duration of two stimuli. These studies find that older adults underestimate the duration of isolated time intervals, compared to young adults, which further support the slowed-clock hypothesis (Craik & Hay, 1999; McCormack, Brown, Maylor,

Darby, & Green, 1999; Vanneste, Perbal, & Pouthas, 1999; Wearden, Wearden, & Rabbitt,

1997b).

Other research has investigated the perception of isolated time intervals while manipulating attentional demands during the task, providing some support for the impaired attention in older adulthood hypothesis. When older adults can fully attend to time, no differences are observed between young- and older-adults; however, when attention is divided older adults underestimate duration (Vanneste, et al., 1999). Moreover, modality effects suggest that impairments in attention may affect older adults‘ ability to perceive duration accurately. A classic finding in the research on timing is that sounds are judged longer in duration than lights

(Goldstone & Goldfarb, 1964a, 1964b; Walker & Scott, 1981). One explanation for this result is that tones capture and hold attention more so than lights (Posner, 1976; Spence & Driver, 1997).

According to SET, this finding might suggest that the greater attention paid to tones results in a quicker closing of the switch, resulting in more pulses being collected for tones, compared to lights (Penney, Gibbon, & Meck, 2000). Consistent with this hypothesis, older adults show a greater modality effect than young adults, with a greater effect observed for dual attention tasks that require dividing attention between tones and lights (Lustig & Meck, 2001, 2002). Taken

80 together, these findings suggest that older adults have an impairment of attentional control that affects their perception of isolated time intervals.

Other research has investigated the potential for age-related memory impairment effects on the perception of duration for isolated time intervals. Some data support the hypothesis that memory impairments might also explain differences observed in the perception of isolated time intervals between young- and older-adults. Unfortunately, memory effects are more difficult to test within the context of SET, as there are is not a diagnostic memory impairment pattern.

Memory impairments are marked only by timing performance that is unresponsive to feedback

(Lustig, 2003). To add to this problem, there is some debate as to whether some memory effects can also be explained by attention (Lustig, 2003).

The strongest support for memory impairments playing a role in the perception of isolated time intervals is provided by a study that found a relationship between working memory scores and age-related impairments in timing (Perbal, Droit-Volet, Isingrini, & Pouthas, 2002).

Additional support for the potential role of age-related memory impairments in time perception is provided by McCormack et al. (1999) and Wearden et al. (1997a), who both used two timing tasks to compare time perception in young- and older-adults. In one task, listeners were trained on a short duration, termed the standard. In test trials, participants were presented with a range of durations, centered on the standard, and judged whether the new durations were the same duration as the standard, responding ‗yes‘ or ‗no.‘ In another version of this task, participants were trained on a short- and a long-duration. During test trials, participants responded whether the test duration was more like the learned short- or long-duration. The main finding from these studies was that older adults tended to underestimate duration, specifically for long durations.

One explanation for these results is that older adults‘ memory for the standard is more variable

81 than young adults (Wearden, et al., 1997b). However, based on this explanation, similar results should have also been observed for short durations. An alternative explanation is that older adults might remember the standard interval as being longer than it objectively was, resulting in underestimation of the long durations (McCormack, et al., 1999). This interpretation is at least consistent with the available data and the tendency for young children to show the opposite pattern of performance in the same study. However, an attention-based explanation for these findings also exist. Participants may have become bored or distracted during the task, resulting in the missing of pulses for only the long durations (Lustig, 2003). Unfortunately, further assessment of these three explanations is impossible based on the current data.

Of the studies that have investigated the perception of isolated time intervals in older adults, relatively few have examined variability on these tasks. Based on studies that report variability measures, most consistently show increased variability in older adults, compared to young adults (Vanneste, et al., 1999; Wearden, et al., 1997b). However, one study suggests no variability differences between young- and older-adults (Rammsayer, 2001). These mixed results might be due to either the time scale of durations investigated or the specific task used. On the other hand, larger age-related effects on variability are consistently observed for divided attention tasks (Lustig & Meck, 2001; Perbal, et al., 2002; Vanneste & Pouthas, 1999). Both young- and older-adults show increased variability on divided attention tasks, but variability in older adults is much greater.

Only one known study has investigated the effect of aging on the perception of sequences of time intervals. Fitzgibbons and Gordon-Salant (2001) used a tempo-discrimination task (see

Figure 8b) to compare discrimination thresholds between young- and older-adults. Participants heard equally timed standard and comparison sequences. Four different rates were presented:

82 standard IOIs were100-, 200-, 400-, and 600-ms. Participants were presented with a standard sequence and two comparison sequences; their task was to identify which of the two comparison sequences was different from the standard sequence. Three main findings emerged from this study. First, the qualitative pattern of tempo discrimination thresholds in older adults is consistent with previous studies of tempo discrimination in young adults. For both age groups, thresholds were lowest for the 200 – 600-ms standard IOIs, but higher for the 100-ms standard

IOI. This finding is consistent with Drake and Botte (1993), in that optimal tempo discrimination was found within the range of 300 – 800-ms, but relative JNDs increased outside this range.

Second, while a similar qualitative pattern of thresholds were found, older adults had significantly higher JNDs compared to young adults. Third, JNDs for older adults were more variable compared to young adults.

The work of Fitzgibbons and Gordon-Salant (2001) is relevant to this dissertation, as

Chapter 6 describes an experiment in which the perception of sequences of time intervals was investigated in older adults. Although the tasks slightly differ, the findings of Fitzgibbon and

Gordon-Salant suggest that older adults should show age-related changes (i.e., lower proportion of correct responses) in the discrimination of sequences of time intervals, compared to young adults.

In sum, research examining the perception of both isolated time intervals and sequences of time intervals suggests that age-related changes in the perception of duration occur. Most of the data on the perception of isolated time intervals support the attention and memory hypotheses of SET. The one known study that has examined the perception of sequences of time intervals in older adults suggests that age-related changes in the perception of sequences of time intervals also occur. While age-related changes in the perception of sequences of time intervals were

83 found, data still supported the optimal range of tempo discrimination proposed by the multiple- look hypothesis.

PD and Timing

The neuropsychological approach to the study of the perceptual and motor timing of isolated time intervals and sequences of time intervals has focused, primarily, on individuals with PD. This group has been of particular interest due to the effect of PD on the basal ganglia

(BG) and dopaminergic cells, which have both been suggested to play a role in perceptual and motor timing.

The review of studies investigating perceptual and motor timing in individuals with PD is organized in a manner similar to the preceding section, except additional factors relevant to the study of individuals with PD are discussed. First, theoretical predictions regarding perceptual and motor timing impairments in PD are reviewed. Second, the behavioral data of the perception and production of both isolated time intervals and sequences of time intervals by individuals with PD are discussed. Finally, the section closes with a brief discussion of the effects of disease severity on timing, the heterogeneity of performance on timing tasks in individuals with PD, the role of dopamine (DA) in timing, and the effects of deep brain stimulation (DBS) on timing.

Theoretical Perspectives on Timing in PD

Compared to entrainment models of timing, interval theories make the strongest predictions regarding perceptual and motor timing impairments in PD. In most cases, these predictions are an extension of the aging predictions generated by SET. In part, this is due to PD generally occurring during older adulthood. However, other predictions concern the effect of PD on the BG and DA. Interval theories propose that the ‗clock,‘ in whole or part, is located within the BG, specifically in the substantia nigra pars compactus and striatum (Meck, 1986).

84

Moreover, SET proposes that DA controls the clock stage of the model (Buhusi, 2003; Meck,

1996; Meck & Benson, 2002). Based on these two proposals, SET predicts that individuals with

PD have a slower clock rate, compared to young- and older-adults, due to the degeneration of dopaminergic cells within the BG. However, DA-enhancing medications are predicted to normalize clock rate (Perbal, et al., 2005).

SET also predicts that attention and the memory impairments affect timing in individuals with PD. These predictions are consistent with the age-related cognitive changes observed in these processes and are biologically plausible, given the finding that PD-related damage to the

BG effects the frontostriatal circuitry (see Meck & Benson, 2002 for a review). However, SET makes no specific predictions regarding changes in these processes associated with PD that differ from the age-related predictions (see Malapani & Rakitin, 2003 for a discussion).

No PD specific predictions have been generated by entrainment theories of timing.

However, one can easily extend the age-related predictions of this model to PD. Assuming that the BG is involved in entrainment processes, individuals with PD are predicted to show a greater

slowing of the clock, compared to older adults. In other words, P0 should be slower, compared to

older adults. Period matching limits between P0 and the driving rhythm should also be narrower for individuals with PD, compared to older adults. In other words, a narrower entrainment region is predicted for individuals with PD, compared to older adults.

Data of Timing in PD

Production of time intervals. Interestingly, no known research has investigated retrospective timing in individuals with PD. Similarly, only a handful of studies have examined the production of isolated time intervals in individuals with PD—although several duration production tasks have been compared within these few studies. The limited number of studies

85 investigating the production of isolated time intervals is interesting, as SET makes very strong predictions regarding duration productions by individuals with PD. According to SET, the slower clock rate proposed for individuals with PD should result in individuals with PD showing greater underestimations and overproductions of durations, compared to older adults, for verbal estimation and duration production tasks, respectively. However, based on the prediction that

DA-enhancing medications normalize clock rate, overestimations/underproductions should be ameliorated when individuals with PD are tested on medication.

Studies have utilized verbal estimation tasks to investigate whether individuals with PD show impairments in the production of isolated time intervals. Only one of these studies has investigated verbal estimation in the subsecond range, reporting no impairments in PD

(Wearden, et al., 2008). A few studies have examined verbal estimation in the more common suprasecond range (3 – 60 seconds).8 These studies, generally, find that individuals with PD reliably underestimate durations, compared to older adult controls, which is consistent with the predictions of the slowed-clock hypothesis in PD (Lange, et al., 1995; Riesen & Schnider, 2001).

Other studies that have used duration production tasks also support the slowed-clock hypothesis in PD. For durations in the 10 – 60 seconds range, individuals with PD overproduced durations when they were off medication. However, overproductions were ameliorated when individuals with PD were on medication, relative to older adult controls (Lange, et al., 1995).

Similar findings have been reported in another study for durations ranging between 30 – 120 seconds. Again, individuals with PD overproduced these durations, compared to older adult controls, suggesting that overproductions are found over a wide range of durations (Jones, et al.,

2008). However, it should be noted that Jones et al. (2008) did not find a reliable difference between controls and individuals with PD. The reason behind these null results is unclear, but the

86 findings do not appear to be due to the use of longer durations used in their study. Interestingly,

Jones et al. also found that duration productions were less accurate when individuals with PD were on medication, compared to off. In addition to accuracy measures, Jones et al. compared the variability of duration productions for young-, older-adults, and individuals with PD. Variability of duration productions were equivalent between individuals with PD and controls, but less variability was found when individuals with PD were on medication, compared to off medication. Taken together, these findings are consistent with the slowed-clock hypothesis and suggest that DA-enhancing medication may reduce both overproductions and variability in duration production tasks.

Similar to the work discussed in the aging section, some research has investigated the effect of divided attention on timing in individuals with PD. Perbal et al. (2005) conducted a study in which they had both individuals with PD (on medication) and older adult controls produce 5, 14, and 38 second durations in both a divided-attention and control task. The divided- attention task required participants to read random numbers, presented on a computer screen, aloud while performing the duration production task. The control task required participants to count aloud on their own. For the control conditions, both individuals with PD and older adults accurately produced the three durations. However, in the divided-attention condition, older adults overproduced all three durations, while individuals with PD were found to underproduce all three durations. Moreover, no variability differences were found between groups, but duration productions in both groups were more variable for the divided-attention task, compared to the control task.

The results of this study are somewhat difficult to interpret. On one hand, the finding that individuals with PD (on medication) performed the same as older adults on the control task is

87 consistent with the general finding that DA-enhancing medications correct overproductions in individuals with PD. Similarly, the finding that older adults show overproductions and increased variability in duration productions for the divided-attention task is consistent with the literature on age-related changes in timing. On the other hand, it is surprising that individuals with PD underproduced durations in the divided-attention task. No theoretical explanations exist for this result. The authors argue that DA-enhancing drugs might speed up the rate of the internal clock in individuals with PD or that temporal memory might be impaired in PD, but neither explanation completely captures the pattern of results found in this study.

Studies using duration reproduction tasks have found that for relatively short durations, between 250 – 2000-ms, individuals with PD (both on- and off-medication) reproduce durations as accurately as controls (Jones, et al., 2008). The PD group (on medication) was found to be less variable than controls, suggesting improved performance when on DA-enhancing medications.

For longer durations (e.g., 6 and 9 seconds), individuals with PD showed greater variability in their reproductions, compared to older adults; however, medication state did not affect performance (Pastor, Artieda, et al., 1992). The results of these studies suggest that only the variability of duration reproductions is affected by PD. While DA-enhancing medications appear to correct the increased variability observed in individuals with PD for short durations, no medication-state effect was observed for the longer durations.

A variant of the duration reproduction task was used by Pastor et al. (1992) to investigate duration reproduction in individuals with PD. For the task, participants were presented with a standard duration while visual time markers flashed on a computer screen. Three flashing rates were presented (200-, 303-, and 625-ms). Participants were instructed to count along with the flashes and, during reproduction, respond when they had reached the total count of the time

88 markers presented during the standard. Counting while timing should have improved timing performance on the task, as it has been shown to improve timing in controls (Grondin, 1992;

Hinton & Rao, 2004; Petrusic, 1984). However, surprisingly, when the rate of the time markers was fast (i.e., 200- or 303-ms), individuals with PD showed greater variability in their reproductions, compared to controls. In contrast, no variability differences were observed for the slower time marker rate (i.e., 625-ms). Medication state modulated the latter findings. Variability was significantly reduced when participants were tested on medication for the fast time marker rates; however, no medication state effect was observed for the slower time marker. The rate specific findings reported in this study suggest that individuals with PD could accurately reproduce durations when slow markers were provided, but that their reproductions did not benefit from the presentation of the fast markers. This finding is consistent with predictions from entrainment models of timing, which suggest a narrowed entrainment region. From this viewpoint, the fast markers may have been outside the entrainment region for individuals with

PD, reducing any benefit of marker presentation during duration reproduction. However, individuals with PD (on medication) were able to benefit from the faster marker rates, suggesting that DA may play a role in modulating the width of the entrainment region.

Other research has investigated the effect of divided-attention on duration reproductions, comparing individuals with PD (on medication) to older adult controls (Perbal, et al., 2005). For the task, participants were presented with a visual square that marked the standard duration (5,

14, or 38 seconds). During the divided-attention condition, participants read aloud a series of random numbers while learning and reproducing the standard duration. For the control condition, participants counted aloud while learning and reproducing the standard duration. No group differences were found for the accuracy of reproductions. Both groups accurately reproduced

89 durations in the control condition and underproduced durations for the divided-attention condition. Similarly, no group differences in the variability of reproductions were found for the control condition. However, greater variability in reproductions were observed for the divided- attention task in individuals with PD, compared to controls. These findings are consistent with the general finding that individuals with PD show greater variability on duration productions of isolated time intervals. Furthermore, these findings suggest that timing impairments observed in

PD may result, in part, from attention-related impairments.

In addition to the research on the production of isolated time intervals, several studies have also investigated the production of sequences of time intervals. These studies have used the

SMT, maximal tapping rate, and synchronization-continuation tasks.

As discussed previously, one measure used to assess the rate of the internal clock is SMT.

Only one known study has investigated SMT in individuals with PD (Yahalom, et al., 2004).

Individuals with PD were found to have slower SMT (684-ms), compared to older adult controls

(581-ms). Instead of SMT, most studies investigating the production of sequences of time intervals in individuals with PD have investigated the fastest (maximum) producible tapping rate

(Harrington, Haaland, & Hermanowicz, 1998; Konczak, Ackermann, Hertrich, Spieker, &

Dichgans, 1997; Pastor, Jahanshahi, et al., 1992; Yahalom, et al., 2004). Primarily, this task serves as a control measure to ensure that participants can synchronize with the target time interval for the synchronization-continuation task. Slower maximal tapping rates (≈ 200-ms) are reported for individuals with PD, compared to older adults (≈ 180-ms; however, see Spencer &

Ivry, 2005). Furthermore, maximal tapping rate is affected by medication state. Individuals on medication have a faster maximal tapping rate, compared to when they are off medication

(Pastor, Artieda, et al., 1992). While slower SMT supports the slowed-clock hypothesis, the

90 slowed maximal tapping rate for individuals with PD is difficult to interpret. Slowing of maximal tapping rate could reflect changes in duration production, which would be consistent with the slowed-clock hypothesis, or simply motor limiting effects of the disease.

Several studies have also examined synchronization-continuation in individuals with PD.

The main questions of interest for these studies are whether individuals with PD 1) show greater variability in their reproductions of the target time interval, compared to older adult controls, 2) can accurately reproduce the target time interval during the synchronization portion of the task, and 3) can accurately reproduce the target time during continuation portion of the task.

In regard to the first question, several studies have applied the W & K (1973) model to separate total ITI variability into clock and motor components. Initially, it was thought that this model could determine whether timing impairments found in PD were due to a disordered clock or the result of motor variability due to the motor symptoms of disease. However, this question remains unanswered, as violations to the W & K model are often observed when applied to synchronization-continuation tapping data obtained from individuals with PD. Specifically, negative motor variability is often reported (Duchek, et al., 1994; Harrington, Haaland, &

Hermanowicz, 1998; Ivry & Keele, 1989; O'Boyle, et al., 1996; Pastor, Jahanshahi, et al., 1992).

Therefore, discussions of clock and motor estimates for individuals with PD should be interpreted skeptically. In the remainder of this section, cases in which estimates of clock and motor variability were calculated are designated appropriately. Undesignated references of variability refer to total variability only.

One consistent finding across studies examining the production of sequences of time intervals in individuals with PD is that the target time interval is underproduced during the continuation portion of the task. Some studies report statistically significant underproductions

91

(Harrington, Haaland, & Hermanowicz, 1998; Ivry & Keele, 1989; O'Boyle, et al., 1996), while others do not (Duchek, et al., 1994; Elsinger, et al., 2003; Yahalom, et al., 2004). Studies consistently show that medication state does not affect the underproduction of the target time during continuation (Elsinger, et al., 2003; O'Boyle, et al., 1996). Mixed findings, however, are reported for ITI variability during the continuation portion of the task. More studies report that individuals with PD show greater ITI variability during continuation tapping, compared to controls (Elsinger, et al., 2003; Harrington, Haaland, & Hermanowicz, 1998; O'Boyle, et al.,

1996; Pastor, Jahanshahi, et al., 1992), but a few suggest no difference in ITI variability

(Duchek, et al., 1994; Ivry & Keele, 1989; Spencer & Ivry, 2005).

Studies that have compared synchronization-continuation ITIs in individuals with PD

(both on- and off-medication) and controls have reported mixed findings (Elsinger, et al., 2003;

Ivry & Keele, 1989; O'Boyle, et al., 1996; Pastor, Jahanshahi, et al., 1992; Spencer & Ivry,

2005). A few studies report that medication state affects variability of continuation ITIs

(O'Boyle, et al., 1996; Pastor, Artieda, et al., 1992). Increased total variability of continuation taps, along with increased clock and motor variability, were found when participants were tested off medication, compared to on medication. Off medication, individuals with PD also had higher total, clock, and motor variability, compared to older adult controls. No differences were observed between individuals with PD (on medication) and older adults (O'Boyle, et al., 1996;

Pastor, Artieda, et al., 1992). Another study reports that medication state did not affect accuracy or variability of duration reproductions in individuals with PD, but that PD does affect the accuracy and variability of duration productions on the synchronization-continuation task.

Elsinger et al. (2003) found no differences in either mean ITI or total variability between on- and off-medication states, but significant differences in both ITI and total variability were found

92 between individuals with PD (on- and off-medication) compared to controls. Unfortunately,

Elsinger et al. (2003) did not calculate W & K estimates for clock or motor variability, so it is unclear whether medication state effected these estimates. Finally, Ivry and colleagues (Ivry &

Keele, 1989; Spencer & Ivry, 2005) report no differences between individuals with PD (on- and off-medication) compared to themselves or older adult controls.

Two factors may be responsible for the discrepant results reported between medication state and PD-related impairments on the synchronization-continuation task. First, Ivry & Keele

(1989) do not report medication information, nor disease stage, for their sample. Based on their report that participants were unable to walk and had severe bradykinesia, it is likely that the sample consisted of individuals with severe PD who may not have been responsive to their PD- related medications. Without disease stage/severity or medication information, it is difficult to interpret their null findings. Furthermore, only seven individuals were analyzed for this analysis, which calls into question the power of their statistical tests. Second, both O‘Boyle, Freeman, &

Cody (1996) and Pastor et al. (1992) do not account for drift in their continuation tapping data.

Drift, or participant‘s ITIs systematically speeding up or slowing down, relative to the target time, is often observed during continuation tapping (McAuley, et al., 2006; Vanneste, et al.,

2001). The presence of drift in continuation ITIs will increase total variability, in turn affecting estimates of clock and motor variability. In light of this, the medication state effects reported by

O‘Boyle, Freeman, & Cody (1996) and Pastor et al. (1992) might be explained by their failure to correct for drift. One possible reinterpretation of their findings is that individuals tested off medication might have a greater tendency to drift during the continuation portion of the task.

However, it is impossible to assess this possibility based on the available data.

93

Most studies that utilize the synchronization-continuation task to investigate the production of duration for sequences of time intervals in individuals with PD have only examined one target time interval (≈ 600-ms). Few studies have investigated synchronization- continuation across multiple rates. This factor is of potential interest since individuals with PD might show greater drift during continuation tapping, as predicted by entrainment models of timing.

Studies investigating synchronization-continuation performance across a range of rates report mixed results. Harrington, Haaland, and Hermanowicz (1998) found no differences between individuals with PD and older adult controls for target time intervals of either 300- or

600-ms. Yet, when asked to synchronize with a wider range of target time intervals, individuals with PD show a similar pattern of overproduction and underproduction for fast and slow rates, respectively. Freeman, Cody, and Schady (1993) investigated performance on the synchronization-continuation task, using a wider range of target time intervals than other studies:

200-, 250-, 333-, 500-, and 1000-ms. Interestingly, the findings for older adults were consistent with other studies investigating performance on this task across a similar range of target rates

(McAuley, et al., 2006). Older adults showed a pattern of overproduction of the fastest rates and underproduction of the slowest rates. Individuals with PD showed greater overproductions and underproductions, respectively, compared to older adult controls.

Pastor et al. (1992) report similar results for the following target times: 400-, 500-, 666-,

1000-, and 2000-ms. They found that for continuation tapping individuals with PD, both on- and off-medication, underproduced the target time for the 1000- and 2000-ms durations, compared to controls. However, for the other target times, medication state differentially effected continuation

ITIs. Individuals with PD, off medication, overproduced the target time intervals relative to

94 controls, although this difference did not reach statistical significance. However, individuals with

PD, on medication, underproduced the target time intervals, which is a consistent finding among most other studies that have used the synchronize-continue task. Medication state also affected the variability of ITIs in individuals with PD. Less variable ITIs were observed when participants were on medication, but this was only statistically reliable for the 400-, 500-, and 666-ms target times. Application of the W & K model to their data revealed that individuals with PD showed greater total, clock, and motor variability, relative to controls for all target time intervals.

Violations of the W & K model precluded statistical analyses of the effects of disease severity and medication state on total, clock, and motor variability. Although the general pattern of the results suggests that total, clock, and motor variability increased with worse disease severity.

Similarly, lower total, clock, and motor variability were observed when participants were on medication, compared to off.

A consistent finding reported by studies that have investigated performance on the synchronization-continuation task over a range of target time intervals is that individuals with

PD underproduce and overproduce the slowest and fastest rates, respectively, presented within an experimental session. These findings provide some support for the narrowed entrainment region hypothesis proposed by entrainment models.

Other research has used the synchronization-continuation task to investigate whether impairments in the production of duration for sequences of time intervals observed in PD might be explained by the motor symptoms of the disease. Specifically, these studies were concerned with the possibility that PD differences might be due to participants tapping with the PD-affected hand. The two studies have investigated synchronization-continuation performance intraindividually in individuals with bilaterally asymmetrical signs of PD suggest a difference in

95 performance between the ‗most affected‘ and ‗least affected‘ hands (Ivry & Keele, 1989;

O'Boyle, et al., 1996). O‘Boyle, Freeman, and Cody (1996) found that total, clock, and motor variability were larger in the ‗most affected‘ hand, but these estimates were not statistically different compared to the ‗least affected‘ hand. Ivry & Keele (1989) report similar results, except that motor variability was roughly equivalent between both hands. Unfortunately, these findings do not differentiate whether potential differences in the performance of individuals with PD on the production of sequences of time intervals are due to motor symptoms of the disease or impairments of the clock.

Elsinger et al. (2003) investigated potential changes in brain activation when individuals with PD performed the synchronization-continuation task. Individuals with PD (on- and off- medication) and older adult controls performed the synchronization-continuation task during fMRI scanning. In this study, participants synchronized hand taps to a 600-ms target time interval. During the synchronization portion of the task, all groups showed activation of the left sensorimotor cortex (SMC), bilateral superior temporal gyrus (STG), and right cerebellum, but individuals with PD (on- and off-medication) showed reduced spatial extent of activation.

Behaviorally, no difference between the groups was found for synchronization. However, for the continuation portion, group differences were observed in both the neural and behavioral data. For controls, the SMC, SMA, and cerebellum were activated. However, for individuals with PD, activation differences were found for medication state. On medication, activation in the SMA, although decreased compared to controls, putamen, thalamus, and SMC were observed. Off medication, only activation in the SMC was observed. Behaviorally, individuals with PD showed significant underproduction of the target time, compared to controls, but this did not vary for medication state. Similarly, higher mean ITI variability was found for individuals with PD, but

96 no medication-state differences were found. Taken together, these findings suggest that levodopa/carbidopa partially normalizes brain activation patterns for the synchronization- continuation task in individuals with PD, but behavioral tasks might not be sensitive enough to pick up on these differences.

In sum, the findings from studies investigating the production of duration for isolated time intervals and sequences of time intervals are mixed. While it is still unclear whether an impairment in the production of time intervals occurs in PD, several studies provide some evidence for such an impairment. The most consistent finding is that this impairment affects the variability of duration productions. Similarly, reports on the effect of DA-enhancing medications on duration productions are mixed. However, several studies report findings that suggest DA- enhancing medications normalize duration production, or at least improve productions relative to an off medication state. Finally, there is mixed support for the PD-related predictions of models of timing. Again, data from the production of duration for isolated time intervals best supports intervals theories, while data from tasks that require the production of sequences of time intervals best support entrainment models of timing.

Perception of time intervals. One potential problem with duration production and reproduction tasks is that they require a motor response, such as a key press, to end the to-be- timed event. Hence, the time to initiate movement is also included in the produced duration.

Since PD is characterized by slowed initiation of movement, some have argued that differences in duration productions and reproductions observed in individuals with PD may reflect this motor impairment, as opposed to a timing impairment. This has led some researchers to investigate the perception of both isolated time intervals and sequences of time intervals, as duration perception tasks do not require a timed motor response.

97

All studies that have examined the perception of isolated time intervals in the 300 – 600- ms range have reported a general pattern of increased discrimination thresholds (worse performance) for individuals with PD, compared to older adult controls (Harrington, Haaland, &

Hermanowicz, 1998; Ivry & Keele, 1989; Wearden, et al., 2008).9 Yet, threshold values were only found to be significantly worse in one study, or approaching significance in another

(Harrington, Haaland, & Hermanowicz, 1998; Wearden, et al., 2008). While Ivry and Keele

(1989) found higher discrimination thresholds in individuals with PD, this difference was rather small. Based on the available evidence, there is strong support to suggest that the perception of isolated time intervals is impaired in individuals with PD. However, it is unclear why mixed results have been found. One potential explanation for the null findings reported by some studies is a lack of statistical power, as discrimination thresholds tend to be highly variable in individuals with PD.

Other research has examined the perception of isolated time intervals for longer durations to determine whether the PD-related timing impairments are duration dependent. Smith, Harper,

Gittings, and Abernathy (2007) compared discrimination thresholds for both subsecond and suprasecond durations in individuals with PD and older adult controls. For subsecond durations, no reliable differences in thresholds were observed between individuals with PD and controls, but thresholds were higher for individuals with PD. However, discrimination thresholds were significantly higher for suprasecond durations in individuals with PD compared to controls.

These findings provide some evidence that the perception of suprasecond durations might be impaired in individuals with PD, while the perception of subsecond durations is, relatively, spared. However, this finding conflicts with other studies that report timing impairments in PD for subsecond durations.

98

Finally, Riesen & Schnider (2001) investigated the effect of divided-attention on the perception of isolated time intervals. Participants concurrently timed two temporally overlapping durations, marked by squares on a computer screen, and judged which of the two squares was presented for a longer duration. Both a short duration (200-ms) and a long duration (1000-ms) were investigated. Overall, individuals with PD showed impaired discrimination, relative to older adult controls, for both the short and long durations when attention was divided. Unfortunately, no data were presented for duration discrimination in the control condition, so it is unclear whether these impairments are due to divided attention or, more simply, impairment in the discrimination of isolated time intervals. The results of this study provided additional support for the general finding that PD-related impairments in the perception of isolated time intervals are found for both subsecond durations, and possibly, suprasecond durations. Furthermore, these findings are consistent with SET. SET predicts that greater impairments in attention for individuals with PD, compared to older adults, are partially responsible for the timing impairments in this population. The duration discrimination difference found between these groups in the divided-attention task supports this hypothesis.

Only one known, published, study has investigated the perception of sequences of time intervals in individuals with PD (Grahn & Brett, 2009). This study is described in greater detail in Chapter 3. Both individuals with PD (on medication) and older adults were presented with a rhythmic tone sequence that varied in metrical complexity. For metrically simple sequences, sequence IOIs were grouped in a manner that gave rise to an easily perceivable beat, or rhythmic pulse. Metrically complex sequences contained the same IOIs, but were grouped so that it was difficult to abstract a simple beat from the sequences. Participants heard a standard and comparison rhythm and judged whether the comparison rhythm was the ‗same‘ or ‗different,‘

99 relative to the standard rhythm. Individuals with PD showed worse rhythm discrimination for metrically simple sequences, compared to controls; however, no differences in discrimination were found between individuals with PD and older adults for metrically complex sequences.

Taken together, these findings suggest that individuals with PD are impaired in the perception of sequences of time intervals. However, based on the available data, it is unclear why this impairment might occur.

This dissertation extends research on the perception of sequences of time intervals in individuals with PD. Specifically, it investigates the effects of rhythmic context on the perceived duration of the standard interval. Based on the reports of impaired perception of sequences of time intervals, individuals with PD were predicted to also show greater effects of rhythmic context on the perceived duration of the standard interval, compared to young- and older-adult controls.

Disease Severity. One factor often considered by research on perceptual and motor timing for isolated time intervals and sequences of time intervals in individuals with PD is whether a relationship exists between disease severity and measures of duration production or perception. Very few studies report a relationship between measures of duration production or perception and the Unified Parkinson Disease Rating Scale-motor examination or the Hoehn and

Yahr scale (Harrington, Haaland, & Hermanowicz, 1998; Pastor, Jahanshahi, et al., 1992; Perbal, et al., 2005; Wearden, et al., 2008). The studies that find relationships between timing and disease severity do not share any common tasks or measures that would suggest PD severity affects any one particular aspect of timing.

Heterogeneity of performance in PD. As evident throughout this review, the findings of studies investigating the perceptual and motor timing of both isolated intervals and sequences of

100 intervals are quite mixed. Several potential explanations are offered throughout the review that may explain some of the mixed results. However, other conflicting findings do not have clear resolution. In part, the mixed findings might simply reflect testing a highly variable neurological population. Merchant, Luciana, Hooper, Majestic, and Tuite (2008) investigated whether the heterogeneity observed in PD might explain the mixed results reported in the timing literature.

Participants completed a battery of perceptual and motor timing tasks (synchronization- continuation, duration reproduction, and an isolated-interval discrimination task), neuropsychological tests, and PD-related measures. Cluster and linear discriminant analyses separated participants into groups. The main finding of this study was that individuals with PD could be separated into low- and high-variability timers. Low-variability timers were found to not differ from controls on timing tasks, while high-variability timers differed significantly from controls. The mixed results reported in studies of timing in PD may, in part, be explained by sampling differences between these two groups of timers with PD.

Role of DA in timing. Early animal timing studies suggested that DA modulates clock speed (Maricq, Roberts, & Church, 1981; Meck, 1983). Administration of DA agonists, such as amphetamine, sped up the clock in rats while DA agonists, such as haloperidol, slowed the clock

(see Meck, 1996 for a review). Later pharmacological studies conducted in neurologically- normal humans found slightly different effects of DA in human timing. The administration of levodopa to neurologically normal young- and older-adults resulted in an overproduction of duration on a reproduction task, which suggests that increased levels of DA slow the clock

(Rakitin, Scarmeas, Li, Malapani, & Stern, 2006). However, administration of DA agonists to neurologically normal participants improved duration discrimination thresholds, while DA

101 antagonists worsened thresholds (Rammsayer, 1989; Rammsayer, 1997; Rammsayer & Vogel,

1992).

Pharmacological studies of the effects of DA on human and animal timing suggest a fairly robust role of DA in the perception and production of duration for isolated time intervals.

Hence, one prediction is that individuals with PD (on- and off-medication) will show medication- state differences on perceptual and motor timing tasks. However, mixed findings have been reported. The majority of studies suggest that DA-enhancing medications improve perceptual and motor timing impairments in individuals with PD (Elsinger, et al., 2003; Jones, et al., 2008;

Lange, et al., 1995; Malapani, et al., 1998; Pastor, Artieda, et al., 1992). One study even reports that individuals with PD perform better than controls, when on medication (Jones, et al., 2008).

However, other studies suggest no medication-state differences in perceptual and motor timing

(Elsinger, et al., 2003; Ivry & Keele, 1989; O'Boyle, et al., 1996; Wearden, et al., 2008).

Other research has identified a DA-dependent distortion in the remembered duration of a time interval in individuals with PD (Jones, et al., 2008; Koch, et al., 2004; Koch, Brusa, Oliveri,

Stanzione, & Caltagirone, 2005; Perbal, et al., 2005; Rakitin, Scarmeas, Malapani, & Stern,

2002). The clearest example of this distortion is reported by Malapani and colleagues (Malapani, et al., 2002; Malapani, et al., 1998), who used the peak-interval procedure to compare productions of isolated time intervals between older adults and individuals with PD (tested both on and off DA-enhancing medication). Participants were trained on both a visually presented short and long duration (e.g., 8 and 21 seconds, respectively) and asked to reproduce the short or long duration during testing. Both training and testing occurred in separate blocks. Older adults and individuals with PD (on medication) reproduced the durations with similar accuracy.

However, individuals with PD, trained and tested off medication, showed a systematic pattern of

102 overproduction and underproduction of the short and long durations, respectively. This pattern of distortions in the production of isolated time intervals has been termed the ‗migration effect,‘ as it appears that the reproductions of both the short and long durations migrated toward each other.

Interestingly, this pattern of distortions only occurs when two time intervals must be produced.

When a single duration is trained and tested, individuals with PD (off medication) showed the opposite pattern of overproduction for the longer duration, but accurate productions of the 21- second duration were found when on medication. The finding that individuals with PD gravitated toward the mean of a set of experienced time intervals only when off medication suggests that the distortion is DA-dependent.

Based on the data that suggests DA-enhancing medication corrects impairments in the production of duration, this dissertation investigated whether DA-enhancing medications modulated the effect of rhythmic context on perceived duration. Specifically, the DA-mediated distortion hypothesis predicted that larger effects of rhythmic context would be found when individuals with PD were tested off medication, compared to on medication.

Effects of deep brain stimulation on timing. Only two known studies have investigated the effect of DBS on timing, at least at the timescale considered in this review (Koch, et al.,

2004; Malapani & Rakitin, 2003). Both studies used duration reproduction tasks. Koch et al.,

(2004) tested individuals with PD, who were receiving both subthalamic nucleus (STN)-targeted

DBS and were taking levodopa/carbidopa, in three separate therapeutic conditions: on-DBS/off- medication, off-DBS/on-medication, off-DBS/off-medication. Participants reproduced both 5- and 15-second durations. Compared to older adult controls, individuals with PD, who were tested off-DBS/off-medication, significantly overproduced the 5 second duration and underproduced the 15 second duration, which is consistent with the non-DBS migration effect discussed earlier.

103

Furthermore, individuals tested off-DBS/off-medication had significantly greater over- and underproductions, respectively, compared to both on-DBS/off-medication or off-DBS/on- medication. However, no differences in duration reproductions were observed when participants were tested on-DBS/off-medication or off-DBS/on-medication.

Malapani and Rakitin (2003) report similar findings in individuals receiving STN- targeted DBS and globus pallidus (GP)-targeted DBS. Specifically, they investigated whether

DBS would eliminate the migration effect in the same manner as DA-enhancing medications.

STN-targeted DBS was shown to eliminate the migration effect; however, GP-targeted DBS did not eliminate migration. These findings suggest that the STN, but not the GP, are involved in reproducing durations. Moreover, these findings suggest that both STN-targeted DBS and levodopa/carbidopa therapies have equivalent effects on correcting timing impairments in individuals with PD.

In summary, research investigating impairments in perceptual and motor timing in individuals with PD is quite mixed. Further compounding this issue is that a number of potentially important factors relevant to timing impairments in PD (e.g., medication state, reports of disease severity, analysis conducted) have varied between studies. However, a few clear findings emerge from this research. First, the majority of studies suggest a timing impairment in individuals with PD. Second, DA-enhancing medications are often reported to improve timing impairments in PD. Third, individuals with PD (off treatment) show distortions in their reproductions of two remembered durations. Several studies report results consistent with the migration effect when individuals with PD are off treatment. Taken together, these findings suggest that individuals with PD should show greater effects of rhythmic context on the perceived duration of the standard interval for the experiment reported in Chapter 6. Moreover,

104 the general finding that timing impairments in PD are ameliorated with DA-enhancing medications suggests that the effect of rhythmic context on the perceived duration of the standard interval should be weakened when individuals with PD are tested on medication, compared to off medication.

105

CHAPTER 5: EXPERIMENT 1—EFFECT OF RHYTHMIC CONTEXT ON DURATION DISCRIMINATION IN YOUNG ADULTS

Experiment Overview

Experiment 1 tested young adults to accomplish two aims. The first aim was to obtain baseline measures of duration discrimination performance for one- and five-standard interval sequences that were not preceded by a context rhythm. Participants judged whether the duration of a variable comparison time interval was ‗shorter,‘ ‗longer,‘ or the ‗same duration‘ as a fixed

600-ms standard time interval. Second, this experiment provided an initial evaluation of the period-correction hypothesis by preceding the one- and five-standard interval sequences with a to-be-ignored context rhythm. The context rhythm consisted of a sequence of 6 tones with inter- onset intervals (IOIs) that took on values of 500-, 600-, or 700-ms; this resulted in three conditions where the first tone that marked the 600-ms standard interval ended late, on time or early, respectively, relative to the context rhythm. Participants were instructed to ignore the context rhythm and judge the duration of the comparison interval relative to the standard interval. Standard-comparison sequences with a to-be-ignored context rhythm provided expectancy profiles for young adults for the one- and five-standard interval sequences. Based on previous research, a ∩-shaped expectancy profile was predicted for the one-standard interval sequence. According to the period-correction hypothesis, increasing the number of repetitions of the standard interval should flatten (or at least weaken) the expectancy profile for the five- standard interval sequence.

Method

Design

The experiment utilized a 2 (number of standard intervals: one, five) x 2 (context condition: no-context, with-rhythmic-context) within-subjects design. For the with-rhythmic-

106 context condition, there was an additional within-subject factor of standard ending with three levels: early, on time, late.

Participants

Participants were 15 undergraduate students (10 females) with normal hearing from

Michigan State University. The group ranged in age from 18 – 24 years (M = 19.6, SD = 1.9) and varied in years of formal musical training (M = 2.9, SD = 2.7). Participants were recruited through the online Human Participation in Research recruitment system and course credit was awarded for participation.

Equipment

Stimuli were generated offline using Audacity Version1.2.6 for Microsoft Windows.

Stimulus presentation and response collection occurred using E-Prime software (Psychology

Software Tools, Inc., Pittsburgh, PA) running on a Dell PC-compatible computer. Stimuli were presented binaurally through Sennheiser HD 280 Pro headphones.

Stimuli

Figure 9 shows a schematic of the no-context condition tone sequences. The sequences contained either one- or five-standard time interval(s) of equal duration, followed by an IOI and a comparison interval. The standard interval was marked by either two- or six-tones with a 600- ms IOI on every trial. There was a 1200-ms IOI between the onset of the final tone of the standard sequence and the onset of the first tone of the comparison interval. The comparison IOI randomly varied from trial-to-trial and took on one of three values yoked to the 600-ms standard

IOI (540-ms, 600-ms, and 660-ms).

Figure 10 shows a schematic of the with-rhythmic-context condition tone sequences. For these sequences, a to-be-ignored context rhythm preceded standard-comparison sequences that

107 were timed identically to the no-context sequences. The to-be-ignored context rhythm consisted of a sequence of six tones that marked out five equal intervals (a context IOI), which on a given trial was equal to 500-ms, 600-ms, or 700-ms. The IOI between the onset of the final tone of the context rhythm and the onset of the first tone marking the beginning of the standard interval was always equal to twice the duration of the context IOI (i.e., 1000-ms, 1200-ms, or 1400-ms, respectively). Consequently, the onset of the tone marking the end of the first standard interval was either on time, or unexpectedly early or late, relative to an extrapolation of the rhythm of the context sequence. For context IOIs of 500-ms, 600-ms, or 700-ms, the initial standard time interval ended ‗late,‘ ‗on time,‘ and ‗early,‘ respectively.

All stimulus tones were 50-ms in duration and had a fundamental frequency of 440 Hz.

Procedure

Participants were tested in two sessions separated by at least one week. The context condition factor was blocked within a session and the order was counterbalanced. Task instructions were read aloud by the experimenter while participants studied a diagram of the tone sequences (similar to Figures 9 or 10). Each trial began with a 200-ms (high-pitched) warning tone followed by a 2000-ms interval before the onset of the first tone of the sequence.

No-context condition. Participants first completed a single familiarization block that consisted of 12 trials. The number of standard intervals randomly varied. Changes between the standard and comparison were large; comparison IOIs were 420-ms, 600-ms, or 780-ms.

Feedback was given after each trial to ensure participants understood the task.

No-Context Condition: One-Standard Interval Sequence Standard Interval Comparison Interval IOI = 600-ms IOI = 540-ms, 600-ms, and 660-ms

IOI = 1200-ms

No-Context Condition: Five-Standard Interval Sequence

Standard Interval Comparison Interval IOI = 600-ms IOI = 540-ms, 600-ms, and 660-ms

IOI = 1200-ms

Figure 9. Schematic of the no-context condition sequences. Circles represent tones. Participants were presented with either one- or five-standard intervals and a single comparison interval; standard IOI(s) were 600-ms on every trial. Their task was to judge the duration of the comparison interval relative to the standard interval and respond ‗shorter‘, ‗same duration,‘ or ‗longer.‘ 108

With-Rhythmic Context Condition: One-Standard Interval Sequence Context Rhythm Standard Interval Comparison Interval IOI = 500-ms, 600-ms, or 700-ms IOI = 600-ms IOI = 540-ms, 600-ms, and 660-ms

Context IOI 2x context IOI IOI = 1200-ms

With-Rhythmic Context Condition: Five-Standard Interval Sequence Context Rhythm Standard Sequence Comparison Interval IOI = 500-ms, 600-ms, or 700-ms IOI = 600 ms IOI = 540, 600, and 660 ms

Context IOI 2x context IOI IOI = 1200-ms

Figure 10. Schematic of the with-rhythmic-context condition sequences. Circles represent tones. A to-be-ignored context rhythm preceded standard-comparison sequences that were timed identical to the no-context sequences. The context rhythm consisted of six tones that marked out five equal intervals (context IOI) equal to 500-ms, 600-ms, or 700-ms, resulting in the initial standard interval ending either late, on time or early, respectively, relative to the context rhythm. Participants were explicitly instructed to ignore the context rhythm and judge the duration of the comparison interval relative to the standard interval, responding ‗shorter‘, ‗same duration,‘ or ‗longer.‘ 109

110

Two test blocks were presented, each of which contained 30 trials. For test blocks, the number of standard intervals was blocked to eliminate any uncertainty about how many standard intervals participants would hear in a block. The order of blocks was counterbalanced. Equal numbers of 420-ms, 600-ms, or 780-ms comparison IOI trials were presented within a block. Ten observations were obtained for each level of number of standard intervals and comparison IOI, resulting in 60 total trials. The order of trials within each block was randomized, with the order of randomized blocks counterbalanced. Participants were instructed to rest the index finger of their dominant hand on a marker centered on the response box. They responded after each trial by pressing a button labeled ‗shorter,‘ ‗same,‘ or ‗longer‘ on the response box. An unlimited amount of time was given for responses, followed by a 2000-ms pause before the warning tone signaling the start of the next trial. No feedback was provided during test blocks. The experimental session lasted approximately one-half hour, with a five-minute break after the first test block.

With-rhythmic-context condition. Participants first completed two familiarization blocks. The first familiarization block was identical to the familiarization block for the no- context condition. The second familiarization block consisted of 18 trials. For each trial, a context rhythm preceded one- or five-standard interval(s) and a comparison interval. The number of standard intervals and context IOI varied randomly. Participants were explicitly instructed to ignore the context rhythm and judge the duration of the comparison interval relative to the standard interval, responding ‗shorter,‘ ‗same duration,‘ or ‗longer.‘ All other aspects of the second familiarization block were identical to the first.

Six test blocks were then presented that each contained 30 trials. For test blocks, the number of standard intervals and standard ending (relative to the context IOI) were blocked to

111 eliminate any uncertainty about the context rhythm and number of standard intervals participants would hear in a block. An equal number of 540-ms, 600-ms, or 660-ms comparison IOI trials were presented within a block. In total, ten observations were obtained for each level of number of standard intervals, standard ending, and comparison IOI, resulting in 180 trials. The order of trials within each block was randomized and the order of randomized blocks was counterbalanced. Responses were collected in the same manner as the no-context condition. No feedback was given during test blocks. The experimental session lasted approximately one hour; participants were given a five-minute break after every two blocks.

Predictions

There were two predictions for Experiment 1. First, no differences were predicted in the proportion of correct responses (PC) between the one- and five-standard interval sequences for the no-context condition. At first glance, this prediction may appear counterintuitive due to the multiple-interval benefit discussed earlier (Grondin, 2001a; Ivry & Hazeltine, 1995; McAuley &

Kidd, 1998). However, previous research has shown that the multiple-interval benefit is eliminated when the standard duration is fixed from trial-to-trial (Miller & McAuley, 2005). The use of a fixed 600-ms standard interval in this experiment is predicted to eliminate any multiple- interval benefit for the no-context condition. However, second, the period-correction hypothesis predicts differences in PC values for the one- and five-standard interval sequences in the with- rhythmic-context condition. Specifically, repetitions of a standard interval are predicted to afford period correction to the standard sequence, eliminating the effect of rhythmic context. Therefore, a ∩-shaped expectancy profile is predicted for the one-standard interval sequence and a flat (or at least weakened) expectancy profile is predicted for the five-standard interval sequence. In other

112 words, an interaction between the number of standard intervals and standard ending is predicted for the with-rhythmic-context condition.

Data Analysis

The main dependent variable of interest was PC values. All statistical analyses were performed using SPSS version 17.0 for Windows (SPSS, 2008). An alpha level of 0.05 was used for all analyses.

Results

Data analysis focused on two ANOVAs, due to the additional factor of standard ending for the with-rhythmic-context condition. The first ANOVA was conducted to assess potential differences between the two context conditions. A 2 (number of standard intervals) x 2 (context condition) repeated-measures ANOVA on PC revealed a main effect of context condition, F(1,

2 14) = 12.60, MSE = 0.009, p = .47, p < .01, but not a main effect of the number of standard

2 intervals, F(1, 14) = 0.40, MSE = 0.01, p = .03, p = .54, or an interaction between the two

2 factors, F(1, 14) = 3.50, MSE = 0.008, p = .20, p = .08. Figure 11 shows that higher PC values were found for the with-rhythmic-context condition (M = .72, SEM = .03) compared to the no- context condition (M = .64, SEM = .03). We will return to a discussion of this finding after presenting the other analyses.

Due to the additional factor of standard ending for the with-rhythmic-context condition, a

2 (number of standard intervals) x 3 (standard ending) repeated-measures ANOVA was conducted on PC values for the with-rhythmic-context condition (Figure 12). Consistent with other studies examining the effect of rhythmic context on perceived duration, a main effect of

2 standard ending obtained, F(2, 28) = 27.56, MSE = 0.009, p = .66, p <.001. Higher PC were found for the on-time standard ending (M = .72; SEM = .03), compared to the early- and late-

113 standard endings (M = .59; SEM = .02 and M = .55; SEM = .02, respectively). Tukey HSD comparisons of the three standard endings indicated that PC for the on-time standard ending was significantly greater than the early- and late-standard endings (ps < .01), but the early- and late- standard endings were not reliably different from each other (ps > .05). No main effect of the

2 number of standard intervals was found, F(1, 14) = 0.32, MSE = 0.008, p = .02, p =.58.

Finally, as predicted by the period-correction hypothesis, a significant interaction between the

2 number of standard intervals and standard ending obtained, F(2, 28) = 4.50, MSE = 0.006, p =

.24, p =.02. A ∩-shaped expectancy profile was found for the one-standard interval sequence, but a weakened expectancy profile was found for the five-standard interval sequence.

1 One-Standard Interval Sequence 0.9 Five-Standard Interval Sequence 0.8 0.7

0.6

0.5 PC 0.4 0.3 0.2 0.1 0 No-Context With-Rhythmic-Context Condition Condition

Figure 11: PC as a function of context condition for the one- and five-standard interval sequences. Error bars represent SEM and the dashed line denotes chance responding.

Trend analyses revealed a significant quadratic trend was found for both the one-standard

2 interval sequence, F(1, 14) = 32.55, MSE = 0.006, p = .70, p < .001, and the five-standard

2 interval sequence, F(1, 14) = 15.75, MSE = 0.005, p = .53, p = .01. However, a lower effect

114 size was observed for the five-standard interval sequence, compared to the one-standard interval sequence. Consistent with the period-correction hypothesis, the lower effect size found for the five-standard interval sequence suggests that repetitions of the standard interval weakened the expectancy profile.

1 One-Standard Interval Sequence 0.9 Five-Standard Interval Sequence 0.8 0.7

0.6

0.5 PC 0.4 0.3 0.2 0.1 0 Early On-Time Late Standard Ending

Figure 12: PC as a function of standard ending for the one- and five-standard interval sequences in the with-rhythmic-context condition. Error bars represent SEM and the dashed line denotes chance responding.

In summary, three main findings were obtained in Experiment 1. First, consistent with previous research, the use of a fixed standard interval eliminated any multiple-interval benefit for the no-context condition; similar PC values were observed for the one-and five-standard interval sequences in the no-context condition. Second, this experiment replicated previous research on the effects of rhythmic context on perceived duration. Participants were unable to ignore the context rhythm when making duration-discrimination judgments, resulting in the ∩-shaped expectancy profile for one-standard interval sequence. However, in support of the period- correction hypothesis, the five-standard interval sequence weakened, but did not completely eliminate, the effects of rhythmic context on perceived duration. I will return to the issue of the

115 profile not being eliminated in Chapter 7. Third, higher PC values were obtained for the with- rhythmic-context condition when compared to the no-context condition. This last finding was somewhat surprising, but is consistent with the findings reported by other studies using this task

(Barnes & Jones, 2000; McAuley & Jones, 2003). At least one potential explanation for the higher PC values found for the with-context rhythm condition is that the rhythm provided by the context reinforced the standard duration in the on-time standard ending condition, enhancing performance in this condition. Consistent with this explanation, higher PC were found for both the one- and five-standard interval sequences in the with-context condition (Ms = .75 and .69, respectively), compared to the one- and five-standard interval sequences in the no-context condition (Ms = .62 and .65, respectively).

116

CHAPTER 6: EXPERIMENT 2—EFFECT OF RHYTHMIC CONTEXT ON DURATION DISCRIMINATION IN OLDER ADULTS AND INDIVIDUALS WITH PD

Experiment Overview

Experiment 2 was similar to Experiment 1, except that older adults and individuals with

PD were tested. As in Experiment 1, participants heard the one- and five-standard interval sequences with both no-context and with-rhythmic-context. The aims of Experiment 2 were threefold. First, the experiment tested whether general aging affected the effects of rhythmic context on perceived duration. Stronger ∩-shaped expectancy profiles were predicted for the one-standard interval sequences with-rhythmic-context for both individuals with PD and older adult controls, compared to young adults. Similarly, stronger expectancy profiles were predicted for both individuals with PD and older adult controls for the five-standard interval sequences with-rhythmic-context, compared to young adults. Second, the experiment tested the DA- mediated distortion hypothesis by comparing the strength of the expectancy profiles generated by the three groups. DA loss in individuals with PD was predicted to result in stronger ∩-shaped expectancy profiles for the one-standard interval sequences and stronger expectancy profiles for the five-standard interval sequences, compared to older adult controls and young adults. Third, the experiment extended the test of the DA-mediated distortion hypothesis by assessing whether

DA-enhancing medications modulated the effects of rhythmic context on perceived duration in individuals with PD by testing the same individuals on- and off-medication. Stronger expectancy profiles were predicted for the with-rhythmic-context sequences when individuals with PD were tested off medication, compared to when they were tested on medication.

117

Method

Design

The general design of the experiment was a 2 (group: PD, control) x 2 (number of standard intervals: one, five) x 2 (context condition: no-context, with-rhythmic-context) mixed- measures design. Individuals with PD and older adult controls made relative duration judgments about the time intervals comprising standard-comparison sequences with one- and five-standard intervals in the no-context and with-rhythmic-context conditions. For the with-rhythmic-context condition there were three additional levels of standard ending; the tone marking the end of the initial standard interval could be early, on-time, or late relative to the preceding context rhythm.

Additionally, individuals with PD were tested in all conditions both on- and off-medication.

Participants

Individuals with PD were recruited through PD support groups in Northwest Ohio.

Thirteen participants (3 females) with mild-to-moderate idiopathic PD, as diagnosed by a neurologist, participated in the experiment. Individuals with PD ranged in age from 40 – 77 years

(M = 63.9, SD = 9.2) and varied in years of formal musical training (M = 4.5, SD = 13.7).

Participants had normal hearing, no other neurological diagnoses or surgeries, and were receiving DA-enhancing medication(s), such as levodopa/carbidopa. Average years post- diagnosis were 7.0 (SD = 5.3). Table 1 provides participant characteristics for these individuals.

Participants were excluded from the study if they scored below a 123 (maximum raw score =

144) on the Dementia Rating Scale (DRS-2).13 Monetary compensation was offered for participation.

Table 1 Participant characteristics for individuals with PD Participant Agea Sex Years post DRS-2 MEQ Mean Education UPDRS-motor UPDRS-motor DA-enhancing diagnosisa raw total CES-Db (years) epoch 2 Epoch 3 medicationsc OFF ON OFF ON State State State State 1 75 Female 10 138 66 9.5 18 16 8 22 11 1 2 61 Male 10 137 67 12 12 28 14 35 38 1, 3, 5, 6 3 60 Male 9 140 48 10.5 16 25 18 23 17 1, 3, 5, 6, 9 4 60 Male 20 142 68 8.5 19 44 35 30 20 1, 2, 3, 6 5 62 Male 1.5 138 60 2.5 16 7 5 19 15 3, 4, 6, 7 6 67 Male 3 140 43 34 12 16 13 52 28 1, 3, 7 7d 64 Male 4 137 66 14 17 20 8 43 28 4, 6, 12 8 60 Male 3 143 61 4.5 16 12 10 26 17 5, 8, 11 9d 64 Male 3 139 52 21 12 38 26 42 34 1, 3, 7 10 77 Male 13 126 61 13 16 45 35 52 54 1, 3, 6 11 40 Female 7 142 49 2.5 13 12 10 11 9 1, 3, 6 12 72 Female 5 142 76 14.5 16 22 14 20 13 1, 4, 5, 6 13 68 Male 3 143 48 4 14 33 27 21 19 1, 8, 9, 10 Mean 63.9 7.0 139 58.9 11.6 15.2 24.5 17.2 30.5 23.3

Note. DRS-2 = Dementia Rating Scale (2nd edition); MEQ = Morningness-Eveningness Questionnaire; CES-D = Center for Epidemiologic Studies Depression Scale; UPDRS-Motor = Unified Parkinson‘s disease Rating Scale-motor examination. aAge and Years Post Diagnosis are based on epoch 1.bMean CES-D is reported, since participants completed the CES-D questionnaire during both epochs 2 and 3. c1 = levodopa/carbidopa (regular); 2 = levodopa/carbidopa and entacapone (Stalevo); 3 = levodopa/carbidopa controlled release (CR); 5 = ropinirole (Requip); 6 = ropinirole-extended release (Requip XL); 6 = rasagiline (Azilect); 7 = clonazepam; 8 = amantadine; 9 = entacapone (Comtan); 10 = pramipexole dihydrochloride (Mirapex); 11 = selegiline; 12 = profenamine hydrochloride (Parsitan). dParticipants were removed from analyses due to failure to complete all testing blocks or below chance performance in all blocks (both on- and off-medication), respectively.

118

119

Older adult controls were recruited from the Northwest-Ohio area through flyers posted at senior centers and in the community. The group consisted of ten older adults (9 females) with normal hearing and no history of neurological diagnoses or surgeries. The group ranged in age from 55 – 87 years (M = 62.7, SD = 10.8) and varied in years of formal musical training (M =

8.6, SD = 15.2). Table 2 provides participant characteristics for the older adult controls.

Monetary compensation was given for participation.

Table 2 Participant characteristics for older adults Participant Agea Sex DRS-2 MEQ Mean Education raw total CES-Db (years) 1 87 Female 138 57 14 14 2 61 Female 143 59 4.5 17 3 78 Female 142 60 3 15.5 4 58 Male 142 61 7 16 5 58 Female 141 69 4.5 19 6 56 Female 141 33 6.5 13 7 57 Female 142 58 5.5 12 8 57 Female 142 42 6.5 17 9 60 Female 143 47 21.5 20 10 55 Female 144 63 4 16 Mean 62.7 141.8 54.9 7.7 16.0

Note. DRS-2 = Dementia Rating Scale (2nd edition); MEQ = Morningness-Eveningness Questionnaire; CES-D = Center for Epidemiologic Studies Depression Scale. aAge reported for epoch 1. bMean CES-D is reported, since the CES-D questionnaire was administered during both epochs 2 and 3.

Equipment

The equipment was the same as Experiment 1. Additionally, spontaneous motor tempo

(SMT) was recorded using a response board consisting of two copper plates, located in front of the participants. Press and release times of participant taps were recorded to the nearest millisecond by a PC running customized software.

Stimuli

The stimuli were the same as Experiment 1.

120

Procedure

Figure 13 provides a diagram of the testing epochs for older adults and individuals with

PD. Older adults completed a total of three testing sessions. Since individuals with PD were tested both on- and off-medication, the PD group completed a total of five testing sessions. All testing sessions occurred on separate days.

Testing sessions for both groups occurred over three epochs. Epoch 1 was the same for both older adults and individuals with PD. Both the DRS-2 and the Morningness-Eveningness

Questionnaire (MEQ) were administered (a description of the screening measures is provided in

Appendix C).

Epoch 2 occurred the following day. For this epoch, participants were presented with one of the context conditions (either the no-context condition or with-rhythmic-context condition).

Context condition order was counterbalanced. For older adult controls, epoch 2 consisted of one testing session. For individuals with PD, epoch 2 consisted of one testing session while on medication and another testing session while off medication. On- and off-medication testing occurred on consecutive days and the order of on- and off-medication sessions were counterbalanced. All testing sessions lasted between 1.5 and 2.5 hours, depending on the context condition presented.

Epoch 3 occurred at least one month after epoch 2. During epoch 3, participants were presented with the context condition that was not presented during epoch 2. Again, older adult controls participated in one testing session and individuals with PD participated in two testing sessions, once on- and once off-medication. For individuals with PD, the order of on- and off- medication testing was counterbalanced. All testing sessions lasted between 1.5 and 2.5 hours, depending on the context condition presented.

121

In addition to perceptual measures of timing, spontaneous motor tempo (SMT) and scores on a depression screening were obtained from both groups during epochs 2 and 3. SMT was collected from participants at the beginning of testing sessions during both epochs 2 and 3. This measure was collected to assess whether the IOIs presented during the experiment were within the range of accessible rates (entrainment region) for participants. For individuals with PD, SMT was collected while they were on medication. Participants were instructed to tap their dominant hand in an even and regular manner at their ‗most comfortable and natural rate of tapping.‘

Thirty-one taps (30 time intervals) were collected from each measurement. A total of four SMT measurements were taken from each participant. Two measures were obtained during both epochs 2 and 3, with the time between measurements approximating ten minutes.

The Center for Epidemiological Studies-Depression scale (CES-D) was also administered to both groups at the beginning of the testing sessions during both epochs 2 and 3 to assess any potential effects of depression on perceptual timing. For individuals with PD, the scale was administered while they were on medication.

For the on medication sessions, individuals with PD maintained their usual medication schedule and were tested approximately one hour after morning administration of their normal

DA-enhancing medications. For the off medication sessions, participants were tested in the morning after they had abstained from taking DA-enhancing medications for at least 8 hours prior to testing (mean withdrawal time = 12.62 h, range = 9 – 15 h). These sessions represented a practically-defined off medication state (Defer, Widner, Marie, Remy, & Levivier, 1999). On- and off-medication testing sessions occurred on back-to-back days.

Epoch 1 Epoch 2 Epoch 3

Med State: On OR Off Med State: On OR Off Med State: On OR Off Med State: On OR Off Context: No OR With No OR With Context: No OR With Context: No OR With SMT (On) SMT (On) SMT (On) SMT (On) PD CES-D (On) CES-D (On) CES-D (On) CES-D (On) Group UPDRS UPDRS UPDRS UPDRS Symptom/Medication Symptom/Medication Symptom/Medication Symptom/Medication

DRS-2 MEQ

Older Adult Context: No OR With Context: No OR With SMT SMT Controls CES-D CES-D

1 2 3 > 30 > 31 Testing Day Figure 13: Diagram of the three testing epochs for older adults and individuals with PD. Dementia Rating Scale-2nd edition (DRS-2); Morningness-Eveningness Questionnaire (MEQ); on- or off-medication (on OR off); no-context condition or with-rhythmic context condition (No OR With); spontaneous motor tempo (SMT); Center for Epidemiologic Studies Depression Scale (CES-D); Unified Parkinson‘s Disease Rating Scale-motor examination (UPDRS); self-reports of symptom severity and medication effectiveness

(Symptom/Medication). 122

123

Two general measures were obtained from individuals with PD to verify on- and off- medication states. The UPDRS-motor section was administered at the start of each testing session during epochs 2 and 3. A total of four measurements were obtained. Participants also self-reported disease severity and medication effectiveness throughout the four testing sessions.

In total, individuals with PD self-rated disease severity and medication effectiveness ten times.

Due to differences in the duration of testing sessions, two self-ratings were obtained during the no-context condition testing sessions and three self-ratings were obtained during the rhythmic context condition testing sessions.

All participants were tested in a quiet room in their homes. Testing sessions for older adult controls were scheduled approximately one-hour after the participants‘ normal waking times.

Predictions

Predictions for Experiment 2 are organized into two sections. The first section discusses the predictions for the no-context condition. The second section discusses the predictions for the with-rhythmic-context condition.

No-context condition. The no-context condition served as a baseline measure of duration discrimination in individuals with PD and older adult controls. These sequences did not directly test either the period-correction or DA-mediated distortion hypotheses. However, based on age- related changes in timing and the reported impairments in timing for individuals with PD, there were four predictions for the no-context condition.

First, based on the changes in perceptual and motor timing reported in older adults, both with and without PD, overall lower PC was predicted for these two groups, compared to young adults. Second, based on the perceptual and motor timing impairments reported for individuals

124 with PD (both on- and off-medication), overall lower PC was predicted for individuals with PD, compared to older adult controls. Third, based on previous research that reports improvements in timing performance when individuals with PD are tested on medication, overall higher PC values were predicted for individuals with PD when tested on medication, compared to off medication.

Fourth, consistent with the results of Experiment 1, no differences in PC values were expected between the one- or five-standard interval sequences for the no-context condition. The use of a fixed 600-ms standard interval in the no-context condition was predicted to eliminate any multiple-interval benefit in duration discrimination.

With-rhythmic-context condition. There were three main predictions for the with- rhythmic-context condition. First, based on the age-related changes in perceptual and motor timing reported for older adults, both older adult controls and individuals with PD were predicted to have stronger expectancy profiles for both the one- and five-standard interval sequences, compared to young adults. A stronger ∩-shaped expectancy profile was predicted for both groups, compared to young adults, for the one-standard interval sequences. Five-standard interval sequences were still predicted to weaken the expectancy profiles, but a less weakened expectancy profile was predicted for older adult controls and individuals with PD, compared to young adults. In other words, a three-way interaction was predicted between group, the number of standard intervals, and standard ending.

Second, based on the DA-mediated distortion hypothesis, DA loss in individuals with PD was predicted to result in stronger expectancy profiles for both one- and five-standard interval sequences, compared to older adult controls and young adults. Again, a stronger ∩-shaped expectancy profile was predicted for the one-standard interval sequences for individuals with

PD, compared to older adult controls and young adults. Similarly, a less weakened expectancy

125 profile was predicted for the five-standard interval sequences for individuals with PD, compared to older adult controls and young adults. In other words, a three-way interaction was again predicted between group, the number of standard intervals, and standard ending.

The third prediction was also based on the DA-mediated distortion hypothesis. Due to the reported improvements in perceptual and motor timing reported when individuals with PD are on

DA-enhancing medications, differences in the strength of expectancy profiles were predicted when individuals with PD were tested on- and off-medication. Specifically, stronger expectancy profiles were predicted when individuals with PD were tested off medication, compared to on medication. In other words, a three-way interaction was predicted between medication state, the number of standard intervals, and standard ending.

Data Analysis

The main dependent variable of interest was PC values. All statistical analyses were performed using SPSS version 17.0 for Windows (SPSS, 2008). An alpha level of .05 was used for all analyses. The data reported in Experiment 1 were used for comparisons between young adults, older adult controls, and individuals with PD.

Eleven participants with PD are included in the final PC analyses because data from two participants with PD were removed. Participant 7 was removed due to a failure to complete all testing blocks. Participant 9 was removed due to below-chance performance for all testing sessions (both on- and off-medication). Exclusion of these participants did not change the overall pattern of the final results. Data from participants 7 and 9 are included in the analysis of spontaneous motor tempo (SMT) discussed later; however, participant 10 was unable to perform the SMT task due to severe tremor.

126

Results

Two sets of analyses were conducted for Experiment 2. The first set of analyses compared individuals with PD (on medication), older adult controls, and young adults. The second set of analyses considered the DA-mediated distortion hypothesis in more detail by comparing individuals with PD both on- and off-medication.

Comparison of Individuals with PD to Older-Adult Controls and Young Adults

Both individuals with PD and older adult controls were compared on a number of measures that could potentially impact interpretation of their data (see Tables 1 and 2). No significant differences were found between groups for age (MPD = 63.9; MOA = 62.7), years of education (MPD = 15.2; MOA = 16), years of formal music training (MPD = 5.6; MOA = 8.6), and scores on the DRS-2 (MPD = 139.0; MOA = 141.8), CES-D (MPD = 11.6; MOA = 7.7), MEQ (MPD

= 58.9; MOA = 54.9)(ps > .09). Furthermore, no reliable differences were found between young adults (M = 2.9), individuals with PD, and older adult controls for years of formal musical training (p > .1).

The first set of analyses compared individuals with PD (on medication), older adult controls, and young adults to assess age-related and PD-specific predictions regarding changes in duration perception without rhythmic context. Individuals with PD (on medication) were chosen for these analyses to be consistent with previous studies of perceptual and motor timing in PD.

A 2 (number of standard intervals) x 2 (context condition) x 3 (group) mixed-measures

ANOVA on overall PC was conducted to test predictions of whether older adult controls, individuals with PD, or both showed changes in duration discrimination, compared to young adults. No main effect of the number of standard intervals, context condition, or group was found

(ps > .10). However, a significant interaction obtained between the context condition and group,

127

2 F(2, 33) = 3.84, MSE = 0.011, p = .19, p =.03. The interaction appears to be driven by the higher PC found for young adults in the with-rhythmic-context condition discussed in

Experiment 1 (Figure 14). We will return to a discussion of this finding in Chapter 7. No other interactions between factors were found (ps > .13).

1 Individuals with PD (On-Medication) 0.9 Older Adult Controls 0.8 Young Adults 0.7

0.6

0.5 PC 0.4 0.3 0.2 0.1 0 No-Context With-Rhythmic-Context Condition Condition

Figure 14: PC as a function of context condition and group. Error bars represent SEM and the dashed line denotes chance responding.

Next, a 2 (number of standard intervals) x 3 (standard ending) x 3 (group) mixed- measures ANOVA was conducted on PC values for the with-rhythmic-context condition to assess the additional factor of standard ending for the three groups. Specifically, this analysis tested the predictions regarding the effects of rhythmic context on perceived duration in both older adult controls and individuals with PD. Critically, a significant interaction between the

2 number of standard intervals and standard ending obtained, F(2, 66) = 6.34, MSE = 0.006, p =

.07, p <.01. This interaction was due to a ∩-shaped expectancy profile for the one-standard interval sequence, but a much weaker expectancy profile for the five-standard interval sequence, which was predicted by the period-correction hypothesis (Figure 15).

128

2 Additionally, a main effect of group obtained, F(2, 33) = 3.47, MSE = 0.037, p = .17, p

=.04. Young adults had higher PC values (M = .62, SEM = .02) compared to older adults and individuals with PD (M = .54, SEM = .03 and M = .57, SEM = .02, respectively). Tukey HSD comparisons of the three groups indicated that young adults had significantly higher PC than older adults (p < .05), but no other comparisons were significant (ps > .05). This effect was likely driven by the higher PC values for the on-time and late standard endings for one-standard interval sequences observed in young adults.

2 A main effect of standard ending was also found, F(2, 66) = 31.164, MSE = 0.01, p =

.49, p <.001. PC values were higher for the on-time standard ending (M = .65, SEM = .02), compared to the early- or late-standard endings (M = .57, SEM = .01 and M = .51, SEM = .01, respectively). Additionally, a main effect of the number of standard intervals was found, F(1, 33)

2 = 5.31, MSE = 0.011, p = .14, p = .03. PC values were higher for the five-standard interval sequence (M = .59, SEM = .02), compared to the single-standard interval sequence (M = .56,

SEM = .01). No other main effects or interactions between factors were found (ps > .06).

Quadratic trend analyses suggested that the strength of the expectancy profiles varied by group. Six separate trend analyses were conducted on PC values for each of the three groups for the one- and five-standard interval sequences. For individuals with PD (on medication), a significant quadratic trend was found for the one-standard interval sequence, but not the five-

2 standard interval sequence, F(1, 10) = 5.62, MSE = 0.012, p = .36, p = .04 and F(1, 10) = 0.26,

2 MSE = 0.013, p = .03, p = .62, respectively. A significant quadratic trend was found for older

2 adults for both the one- and five-standard interval sequences, F(1, 9) = 16.96, MSE = 0.005, p

2 = .65, p = .003 and F(1, 9) = 5.68, MSE = 0.007, p = .39, p = .04 , respectively. Similarly, a significant quadratic trend was found for young adults for both the one- and five-standard

129

2 interval sequences, F(1, 14) = 32.55, MSE = 0.013, p = .70, p < .001 and F(1, 14) = 15.75,

2 MSE = 0.006, p = .53, p = .001 , respectively. Contrary to the predictions that older adults and individuals with PD would show greater effects of rhythmic context on perceived duration, the effect size measurements of the trend analyses suggest that younger adults had the strongest expectancy profiles, compared to older adults, and then individuals with PD, for both one- and five-standard interval sequences. Furthermore, stronger expectancy profiles were observed in older adult controls, compared to individuals with PD. We will return to a discussion of these findings in Chapter 7.

In summary, the analyses comparing individuals with PD (on medication), older-adult controls, and young adults highlight a few important findings. First, contrary to some previous studies that have reported duration discrimination impairments in individuals with PD, no differences were observed between individuals with PD, older adult controls, or young adults in the no-context condition. This finding suggests no age-related changes or PD impairment in the perception of duration when rhythmic context is not present. However, for the with-rhythmic context condition, lower PC was found for older adults, compared to young adults. Lower overall

PC was found for individuals with PD also, but PC values did not significantly differ from young adults. Second, contrary to the general aging and PD-specific timing impairment predictions, young adults showed the strongest expectancy profiles, and individuals with PD the weakest expectancy profiles, for the one-standard interval sequence. The same pattern of results was observed for the five-standard interval sequence, with young adults showing the strongest expectancy profiles. Taken together, the results of these analyses offer support for the period- correction hypothesis, consistently showing that five standard intervals weakens the expectancy

130 profile. However, the current analyses offered little support for the DA-mediated distortion hypothesis.

a)

1 Individuals with PD (On Medication) 0.9 Older Adult Controls 0.8 Young Adults 0.7

0.6

0.5 PC 0.4 0.3 0.2 0.1 0 Early On-Time Late Standard Ending

b) 1 Individuals with PD (On Medication) 0.9 Older Adult Controls 0.8 Young Adults 0.7

0.6

0.5 PC 0.4 0.3 0.2 0.1 0 Early On-Time Late Standard Ending

Figure 15: PC as a function of standard ending and group for the a) one-standard interval sequence and the b) five-standard interval sequence in the with-context-rhythm condition. Error bars represent SEM and the dashed line denotes chance responding.

131

Comparison of Individuals with PD On- and Off-medication

In order to further test the DA-mediated distortion hypothesis, a second set of analyses compared individuals with PD both on- and off-medication.

A 2 (medication state) x 2 (number of standard intervals) x 2 (context condition) repeated-measures ANOVA was conducted on overall PC to test the prediction that individuals with PD, off medication, would show lower PC for the no-context condition when compared to on medication. No main effects or interactions were found (ps > .08). Figure 16 shows mean PC values for this analysis.

1 On Medication 0.9 Off Medication 0.8 0.7

0.6

0.5 PC 0.4 0.3 0.2 0.1 0 One-Standard Five-Standard One-Standard Five-Standard Interval Interval Interval Interval Sequence Sequence Sequence Sequence No-Context Condition With-Rhythmic-Context Condition

Figure 16: PC as a function of context condition, number of standard intervals, and medication state for individuals with PD. Error bars represent SEM and the dashed line denotes chance responding.

Due to the with-rhythmic-context condition having three levels of the standard ending factor, a 2 (medication state) x 2 (number of standard intervals) x 3 (standard ending) repeated- measures ANOVA was conducted on PC values for the with-rhythmic-context condition. Of most interest, a significant interaction was found between medication state and standard ending,

2 F(2, 20) = 3.83, MSE = 0.004, p = .28, p = .04. As seen in Figure 17, this interaction was

132 driven by stronger expectancy profiles for both the one- and five-standard interval sequences when individuals with PD were tested off medication, compared to on medication. This interaction offers some support for the DA-mediated distortion hypothesis. Additionally, main effects were found for both the standard ending and number of standard intervals, F(2, 20) =

2 2 8.88, MSE = 0.02, p = .47, p = .002 and F(1, 10) = 15.86, MSE = 0.008, p = .61, p = .003, respectively. Higher PC values were found for the on-time standard ending (M = .64, SEM =

.04), compared to the early- or late- standard endings (M = .56, SEM = .03 and M = .53, SEM =

.03, respectively). Higher PC values were also found for the five-standard interval sequences (M

= .61, SEM = .03), compared to one-standard interval sequences (M = .55, SEM = .03). No interaction was found between the number of standard intervals and standard endings (p = .40 ).

No other main effects or interactions were significant (ps > .39).

1 Early Standard Ending 0.9 On-Time Standard Ending 0.8 Late Standard Ending 0.7

0.6 0.5 PC 0.4 0.3 0.2 0.1 0 One-Standard Five-Standard One-Standard Five-Standard Interval Interval Interval Interval Sequence Sequence Sequence Sequence

On Medication Off Medication

Figure 17: PC as a function of standard ending, number of standard intervals, and medication state for individuals with PD in the with-context rhythm condition. Error bars represent SEM and the dashed line denotes chance responding.

133

Trend analyses clarified the medication state and standard ending interaction. Although the three-way interaction between medication state, standard ending, and number of standard intervals that was predicted by the DA-mediated distortion hypothesis, did not obtain, a significant quadratic trend was found for individuals with PD, tested on- and off-medication, for

2 the single-standard interval sequence, F(1, 10) = 5.62, MSE = 0.006, p = .13, p = .04 and F(1,

2 10) = 9.95, MSE = 0.016, p = .50, p = .01, respectively. A significant quadratic trend was also found for the five-standard interval sequence when individuals with PD were tested off

2 medication, F(1, 10) = 11.25, MSE = 0.007, p = .53, p <.01. However, the quadratic trend was not significant for the five-standard interval sequence when the PD group was on medication,

2 F(1, 10) = 0.26, MSE = 0.01, p = .03, p = .62. Taken together, the trend analyses offer some support for the DA-mediated hypothesis. Effect size measures of the trend analyses suggest that a stronger expectancy profile was found for the one-standard interval sequence when individuals with PD were tested off medication, compared to on medication. Furthermore, multiple repetitions of the standard interval weakened the expectancy profile when participants were on medication, but not when the same individuals were off medication. This finding suggests that

DA might play a potential role in period-correction processes.

PD severity ratings. The UPDRS-Motor examination provided a measure of PD severity at the time of testing. UPDRS scores range between 0 (no motor symptoms) – 108 (extreme motor symptoms). As seen in Table 1, some participants showed drastic changes in their UPDRS scores between epochs 2 and 3. A 2 (medication state) x 2 (epoch) ANOVA for repeated- measures was conducted on UPDRS scores to assess whether UPDRS scores significantly differed between epochs 2 and 3. A main effect of medication state was found, F(1, 10) = 35.19,

2 MSE = 13.02, p = .78, p < .001. Mean UPDRS scores were lower when participants were on

134 medication (M = 19.41, SEM = 3.15) compared to off medication (M = 25.86, SEM = 3.34).

However, there was no effect of epoch, nor an interaction between factors (ps > .23 ). These findings support that PD symptoms were improved in the on medication state, compared to the off medication state. Furthermore, mean UPDRS scores did not significantly differ between epochs 2 and 3.

Participants‘ self-reports of symptom severity and medication effectiveness further validated the medication state factor. Both symptom severity and medication effectiveness self- ratings were significantly correlated with UPDRS scores, r(42) = -.53, p < .001 and r(42) = -.38, p =.012, respectively. Participants self-reported worsened symptoms or less medication effectiveness with higher UPDRS scores.

PD characteristics and duration discrimination. Consistent with most previous studies examining perceptual or motor timing in individuals with PD, no reliable relationships were found between disease duration, disease severity, or self-ratings of symptom severity/medication effectiveness and PC values for the on-time standard endings for either of the context conditions.

The lack of a relationship between any of these variables suggests that neither duration discrimination, or the effects of rhythmic context, changed across PD duration or severity.

In summary, the analysis comparing individuals with PD, both on- and off-medication highlights two points. First, duration discrimination for the no-context condition revealed no differences between the individuals with PD, tested both on- and off-medication. This finding suggests that DA-enhancing medication does not improve the perception of duration in individuals with PD when no context rhythm is present. Second, some support for the DA- mediated distortion hypothesis was found. Trend analyses suggested that medication state did affect the strength of the expectancy profiles for the one-standard interval sequences. A stronger

135 expectancy profile was found when individuals with PD were tested off medication, compared to on medication. Furthermore, for the five-standard interval sequences, medication state differentially affected the strength of the expectancy profile. A flat expectancy profile was found when individuals with PD were tested on medication, but only a weakened expectancy profile was found when participants were tested off medication.

Comparison of Spontaneous Motor Tempo in Individuals with PD to Older Adult Controls

Spontaneous motor tempo (SMT) measures were obtained from both individuals with PD and older adult controls to investigate potential differences in preferred tapping rate. Two measurements of SMT were collected at the beginning of the session during epochs 2 and 3.

SMT measurements obtained from individuals with PD were taken while on medication.

A 2 (epoch) x 2 (group) mixed-measures ANOVA on median SMT revealed a main

2 effect of group, F(1, 18) = 9.02, MSE = 63518.39, p = .33, p < .01. Individuals with PD had a faster SMT (M = 378.8, SEM = 56.4) than older adult controls (M = 618.2, SEM = 56.4).

Interestingly, the SMT of individuals with PD found in this study is also faster than the ≈ 600-ms

SMT reported for young adults (Fraisse, 1982; McAuley, et al., 2006; Mishima, 1956; Smoll &

Schutz, 1978). Neither a main effect of epoch, nor an interaction between factors was found (ps

> .66). Analysis of mean SMT revealed the same pattern of results.

To assess whether SMT was reliable between the two testing epochs, correlations were conducted on average and median SMT values for Epochs 2 and 3. Due to the group differences in SMT, all values were converted to z-scores centered on the mean SMT for the group. A significant relationship was found for both average and median SMT between the two testing sessions, which were separated by at least a month (all rs (20) > .49, p = .03). This finding supports previous studies that report SMT is consistent within an individual over time.

136

In sum, Experiment 2 revealed three important findings. First, little support was found for the predicted differences between individuals with PD, older adult controls, and young adults. In general, individuals with PD did not show signs of impairment in perceptual timing, compared to older adult controls and young adults for either the no-context or with-rhythmic context conditions. Second, the largest PD-related differences in timing performance were found when individuals with PD were compared on- and off-medication. Specifically, medication state mediated the effect of rhythmic context on perceived duration. Stronger expectancy profiles were observed for the one-standard interval sequence when off medication, compared to on medication. Furthermore, a flattened expectancy profile was observed for the five-standard interval sequence when on medication, compared to off medication. Third, SMT was found to be very fast in individuals with PD, compared to older adult controls and young adults. All three of these findings are discussed in greater detail in the following chapter.

137

CHAPTER 7: GENERAL DISCUSSION

The aim of this dissertation was to test two hypotheses regarding the effects of rhythmic context on perceived duration. Specifically, two experiments tested the period-correction and the dopamine (DA)-mediated distortion hypotheses. The period-correction hypothesis posited that the effects of rhythmic context on perceived duration should be eliminated (or at least weakened) with multiple repetitions of a standard interval. The DA-mediated distortion hypothesis posited that the effects of rhythmic context on perceived duration are mediated by DA. This hypothesis predicted stronger expectancy profiles for individuals with PD when tested off DA-enhancing medication, compared to on medication. Moreover, due to the loss of dopaminergic cells in PD, stronger expectancy profiles were predicted for individuals with PD, compared to older adult controls and young adults. Four main findings contribute to the current understanding of the effects of rhythmic context on perceived duration in young adults, older adults, and individuals with PD.

The first main finding of this dissertation was that increasing the number of standard intervals weakened the effects of rhythmic context on perceived duration. This was found in both

Experiments 1 and 2 for young adults, older adults, and individuals with PD. A ∩-shaped expectancy profile was found for all groups for the one-standard interval sequence in the with- rhythmic-context condition. However, weakened expectancy profiles were found for all groups for the five-standard interval sequence in the with-rhythmic-context condition. This pattern of results offers support for the period-correction hypothesis and entrainment models of timing.

Specifically, the pattern of results suggests that even though participants were explicitly instructed to ignore the context rhythm, they were unable to do so. Instead, participants entrained to the context rhythm, resulting in the lower PC values for unexpected standard endings (i.e., late

138 and early) for the one-standard interval sequences. However, multiple standard intervals afforded period correction to the 600-ms standard interval, thereby weakening the expectancy profile. We will return to a discussion of why the expectancy profile was not completely eliminated with five-standard intervals later in the chapter.

The second main finding of this dissertation was that one of the basic predictions of the

DA-mediated distortion hypothesis was not met. Specifically, this hypothesis predicted a stronger ∩-shaped expectancy profile for individuals with PD for the one-standard interval sequence in the with-rhythmic-context condition and a less weakened expectancy profile for the five-standard interval sequence in the with-rhythmic-context condition for individuals with PD, compared to young adults. However, the opposite pattern of results were found; stronger expectancy profiles were found for young adults for both one- and five-standard interval sequences, compared to individuals with PD. We will return to a discussion of at least one possible explanation for this pattern of results later in the chapter.

However, the third main finding of this dissertation was that repetitions of the standard interval differentially affected individuals with PD when they were on- and off-medication. For the five-standard interval sequence, a flat expectancy profile was found for PD participants when they were on medication, but not when they were off medication. Moreover, a flattened expectancy profile was only found for individuals with PD when on medication; a flattened expectancy profile was not found for young adults or older adult controls with the five-standard interval sequence in the with-rhythmic-context condition. This finding provided evidence for the potential involvement of DA in the period-correction process and, additionally, provided some support for the DA-mediated distortion hypothesis.

139

Finally, the fourth main finding of this dissertation was that trend analyses suggested individuals with PD had stronger expectancy profiles for the one-standard interval sequences when tested off medication, compared to on medication. While the basic prediction of the DA- mediated distortion hypothesis was not met, this finding provided some support for the hypothesis, suggesting that DA might mediate the effects of rhythmic context on perceived duration.

In addition to the main findings of this dissertation, several other findings contribute to current issues in the research on timing by individuals with PD and older adults. First, no differences in PC values were found between individuals with PD when they were tested on- or off-medication. Moreover, no differences in PC values were found between older adult controls and individuals with PD for the one-standard interval sequence in the no-context condition.

These two findings were somewhat surprising, as the single-standard interval sequence in the no- context condition is comparable to the stimuli used in previous studies that report either impairments in the perception of isolated time intervals for individuals with PD or that DA- enhancing medications ameliorate timing impairments in PD. The findings of this dissertation are consistent with previous studies that have used this type of sequence and found no impairments in the perception of isolated time intervals in individuals with PD (Ivry & Keele,

1989; Smith, et al., 2007; Wearden, et al., 2008) and other studies that have shown no effect of

DA-enhancing medications on timing in individuals with PD (Elsinger, et al., 2003; Ivry &

Keele, 1989; O'Boyle, et al., 1996; Wearden, et al., 2008). However, these findings conflict with several studies that report an impairment in the perception of isolated time intervals for individuals with PD (Harrington, Haaland, & Hermanowicz, 1998; Ivry & Keele, 1989; Smith, et

140 al., 2007; Wearden, et al., 2008) and other studies that show DA-enhancing medications ameliorate this impairment (Jones, et al., 2008; Malapani, et al., 1998; Perbal, et al., 2005).

There are at least two possible explanations for why group or medication differences were not found for the single-standard interval sequences in the no-context condition. First, the performance of both older adults and individuals with PD was quite variable. Since the number of participants in these two groups was rather low, the null effect of the group and medication state may be explained by low statistical power. Second, the null effect of group and medication state may be due to the sensitivity of the PC measure used in this dissertation. Some support for the second explanation is provided by a previous study that investigated the perception of sequences of time intervals in individuals with PD (Grahn & Brett, 2009). Grahn and Brett obtained both the measures of d-prime, a measure of perceptual sensitivity, and PC for their task.

Clear statistical differences obtained between older adult controls and individuals with PD for the d-prime measure, but several parallel statistical tests failed to find significant group differences with the PC measure. Taken together, these explanations suggest that larger sample sizes and use of a more robust measure of sensitivity than PC may be necessary to find reliable differences between individuals with PD and older adult controls for the perception of isolated time intervals.

Another current issue in the research on aging and timing is whether age-related changes in the perception of duration occur. Although not clear cut, this dissertation found some evidence of age-related differences in the perception of duration in both the no-context and with-rhythmic- context conditions. The significant interaction between context condition and group (Figure 14) may, in part, be driven by lower PC values found for older adult controls. Furthermore, a difference in overall PC values between older adults and young adults was found for the with-

141 rhythmic-context condition. While strong interpretation of these group differences is difficult within the current study, the differences may suggest that age-related changes in duration perception occur. Future research is necessary to clarify which aspects of duration perception may be affected by aging.

This dissertation also extended research on SMT in individuals with PD, which has been investigated by only one other known study (Yahalom, et al., 2004). This dissertation found that individuals with PD (on medication) had a faster SMT (≈ 380-ms) compared to older adult controls (≈ 620-ms) and those reported by other studies for younger adults (≈ 600-ms). Visual inspection of the data confirmed that the average SMT value was not driven by outliers. Of the

40 SMT measures obtained in this dissertation, only four SMT measurements were greater than

600-ms and these values varied across individuals. This finding conflicts with Yahalom et al., who reported slower SMT in individuals with PD (684-ms), compared to older adults. In contrast to the faster SMT observed in individuals with PD in this dissertation, the SMT value obtained for older adults was similar to those reported by other studies (Baudouin, et al., 2004; McAuley, et al., 2006; Vanneste, et al., 2001).

The reason for the conflicting SMT values reported by this dissertation and Yahalom et al. (2004) for individuals with PD is unclear. The SMT value reported in this dissertation was not within the range of tremor frequencies (4 – 6 Hz) reported for individuals with PD, so it is unlikely that faster SMT is explained by participant‘s tremor (Lang & Lozano, 1998). Moreover, since both this dissertation and Yahalom et al. collected SMT measurements at the start of the experimental session and while participants were on medication, the differences in SMT are not likely due to these factors.

142

At least one possible explanation for the different SMT values obtained by this dissertation and Yahalom et al. is that the two studies varied the hand used to produce SMT. This dissertation had participants produce their SMT with their dominant hand, while Yahalom et al. had participants produce their SMT with the ‗least affected‘ hand. A reanalysis of the SMT data assessed this explanation for the different findings between studies. Participants were separated into groups, based on whether they produced SMT with their ‗least affected‘ or ‗most affected‘ hand. Group assignment was determined by scores on the UPDRS for hand tasks. No differences were found between groups, suggesting that the faster SMT found in this dissertation was not due to participants tapping with their ‗least‘ or ‗most‘ affected hand. Further research is necessary to clarify estimates of SMT in individuals with PD between the two studies.

Although the findings of this dissertation provide support for the period-correction hypothesis, by showing that multiple standard intervals weakened expectancy profiles for the with-rhythmic-context condition, one question that remains unanswered is why the five-standard interval sequences did not flatten expectancy profiles across the groups. This finding was not too surprising, as previous studies have shown period-correction processes are slower than phase- correction processes, due to period correction processes being dependent on phase correction

(Large & Jones, 1999; McAuley & Jones, 2003). Unpublished studies conducted in our lab on young adults have shown that ten-standard interval sequences can flatten expectancy profiles.

However, it is still unclear how many standard intervals are necessary to flattened expectancy profile within the range of five- to ten-standard intervals.

One unanticipated finding of this dissertation was that young adults‘ performance on the duration-discrimination with rhythm context tasks was enhanced, relative to the no-context condition, by the reinforcing context rhythm for the on-time standard ending. Higher PC values

143 were found for both the one- and five-standard interval sequences for the on-time standard ending condition (Ms = .75 and .69, respectively), compared to the one- and five-standard interval sequences in the no-context condition (Ms = .62 and .65, respectively). A similar rhythm enhancing effect has been reported by previous research using this task (Barnes & Jones, 2000;

McAuley & Jones, 2003).

A likely explanation for this finding is provided by entrainment models of timing.

Remember that participants heard a sequence of equally-timed intervals for the on-time standard ending in the with-rhythmic-context condition. For these sequences, no period correction is necessary for the oscillator to entrain and participants hear multiple repetitions of the standard interval, which should result in enhanced performance in this condition. At first glance, this explanation seems to conflict with the data obtained in the no-context condition, as the temporal regularity of the five-standard interval sequences in the no-context rhythm should have resulted in a similar enhancing effect of the reinforcing rhythm. However, the use of a fixed standard interval explains why this effect was not found for the no-context sequences. Since the 600-ms standard interval was fixed during the testing session, period correction was not necessary for the sequences in the no-context condition, thereby eliminating any multiple-interval benefit (see

Miller & McAuley, 2005). However, for the sequences in the with-rhythmic-context condition, the rate of the context rhythm changed (roved) for each test block. The roving context rhythm would require the oscillator to entrain to a range of rates (i.e., 500-ms, 600-ms, and 700-ms) during the course of the testing session, resulting in the enhanced effect of the reinforcing rhythm for the on-time standard ending.

Interestingly, no enhancing effect of the reinforcing rhythm was found for either individuals with PD or older adult controls. This finding offers further suggestion that age-

144 related changes in rhythm perception might occur; specifically, individuals with PD and older adult controls may not have picked up on the ‗beat‘ provided by the context rhythm in this condition, resulting in the lack of an enhancing effect of the reinforcing rhythm for these groups.

Some support for this explanation is provided by a study that reports higher PC values for young adults, compared to older adults and individuals with PD, on a beat perception task (Grahn &

Brett, 2007, 2009). However, this dissertation did not find support for the differences between older adult controls and individuals with PD reported in the same study (Grahn & Brett, 2009).

Further research is necessary clarify these group differences, and possible age-related changes, in the perception of sequences of time intervals.

Another unanticipated finding, based on the trend analyses conducted in Experiment 2, was that young adults had stronger expectancy profiles for both the one-standard interval sequence, followed by older adults, and then individuals with PD. Similarly, less flattened expectancy profiles were found for young adults, followed by older adult controls, and then individuals with PD. This finding was opposite the general aging prediction proposed by the DA- mediated distortion hypothesis. Although very speculative, at least one possible explanation for this finding, which is related to the previous explanation for young adult performance in

Experiment 1, is that older adult controls and individuals with PD may not have been as strongly entrained by the context rhythm as young adults. In accord with entrainment models of timing, weaker entrainment may be realized by either the peak of the oscillator being less coupled with the tone onsets of the context rhythm or the width of the attentional pulse, or attentional focus in time, widening with age (Large & Jones, 1999). If these groups were more weakly entrained by the context rhythm, then weaker effects of rhythmic context on perceived duration would be predicted. Specifically, early- and late-standard endings might not be perceived as occurring as

145 unexpectedly in time, resulting in a weakened expectancy profile similar to that found in

Experiment 2. Similarly, the five-standard interval sequences may have led to a flatter expectancy profile for these groups, as less period correction might be necessary for an oscillator weakly entrained by the context rhythm. Alternatively, if young adults were strongly entrained by the context rhythm, they would likely be more affected by the unexpected onsets of the early- and late-standard endings, resulting in a stronger ∩-shaped profile, similar to the profile found in

Experiment 1. Presumably, an oscillator strongly entrained by the context rhythm would also take longer (e.g., more cycles) for accurate period correction to occur, resulting in stronger expectancy profiles found for young adults for the five-standard interval sequences.

Previous studies investigating motor timing in older adults and individuals with PD provide some indirect support for this explanation. For the synchronization-continuation tapping task, both older adults and individuals with PD have been shown to overproduce or underproduce the fastest and slowest target durations, respectively, presented within the experimental session during continuation tapping (Harrington, Haaland, and Hermanwicz, 1998;

McAuley et al., 2006). One interpretation of this finding is that the fastest and slowest rates are outside of the range of durations to which these groups can entrain (e.g., the entrainment region).

However, it is currently unclear what age-related or disease-related changes occur that might lead to the narrowing of accessible rates later in the lifespan. If this change is due to a generally weaker coupling between the individual and the environment, then similar changes might also occur within the range of accessible rates.

At least one problem with this interpretation is that it would not predict the results obtained for the on- and off-drug comparison in individuals with PD. Since individuals with PD should, presumably, be the group most weakly entrained by the context rhythm, weaker

146 expectancy profiles should be found for both the one- and five-standard interval sequences in the with-rhythmic-context condition for individuals with PD (off medication). However, the opposite was found. Further research is needed to determine the entrainment region of individuals with

PD and to develop a better understanding of the mechanisms responsible for the age-related changes in entrainment.

One limitation of this dissertation is that the testing sessions for controls and individuals with PD were not equally matched; individuals with PD were tested on each of the two context conditions on back-to-back days, while the control groups were not. This design aspect of the study was due to the medication state condition for individuals with PD. Logistical reasons limited similar testing conditions being carried out for young adults and older adult controls. The primary concern raised by this design issue is that a portion of the results for the group comparisons may reflect practice effects in individuals with PD. To investigate this possibility, the performance of individuals with PD (both on- and off-medication), older adult controls, and young adults was compared for only the first testing session of Epoch 2. The same pattern of results reported in Experiment 2 was found, suggesting that the pattern of results reported for individuals with PD are not due, solely, to practice effects. Furthermore, the general pattern of results found for each of the groups was evident through visual inspection of the data during the first testing session of Epoch 2.

While the group comparisons reported in this dissertation are subject to the concern of an underlying practice effect, is should be highlighted that the comparison of individuals with PD, when on- and off-medication, are not subject to this concern. Therefore, this is the cleanest comparison between groups reported in this dissertation.

147

The main findings of this study have implications for models of timing. The discussion of the implications of these findings for models of timing will focus on entrainment models, as predictions from these models most closely match the pattern of findings reported for the paradigm used in this dissertation (Barnes & Jones, 2000; Large & Jones, 1999; McAuley &

Jones, 2003).

One implication of these findings is that entrainment models of timing can be extended to generate predictions regarding potential timing impairments in individuals with PD. No known hypotheses have been generated or tested from the entrainment perspective regarding PD. This dissertation suggests that entrainment models can successfully generate and test hypotheses regarding timing in PD. Furthermore, this dissertation suggests that the neuropsychological approach to the study of timing is a potentially useful approach to test these hypotheses, as neuropsychological groups may highlight the role of different brain areas and neurotransmitter systems in entrainment. Investigation of timing in PD is of interest, as it converges with other evidence that suggest the BG-thalamo-premotor loop might be involved in entrainment processes

(Grahn & McAuley, 2009; Grahn, 2009; Grahn & Brett, 2009).

Extending the previous implication, one main finding of this dissertation is initial evidence that DA may play a role in the period-correction process. This finding has implications for entrainment theories of timing, as no known research has reported a link between a neurotransmitter system and entrainment. While additional research is necessary to better assess the possible role of DA in period correction, this finding does provide an important step toward a better understanding the neurobiological bases of entrainment.

This dissertation contributed to the broader timing literature in three ways. The first contribution of this dissertation was the test of the period-correction hypothesis. While studies

148 that have used the duration-discrimination with rhythmic context task report effects of rhythmic context on perceived duration, no previous studies have investigated whether the effects are eliminated (or weakened) with multiple standard intervals. This dissertation provided support for the period-correction hypothesis by showing that multiple standard intervals weakened the expectancy profile. The second contribution of this dissertation was that it provided some evidence for the role of DA in period correction. The third contribution of this dissertation was the test of the DA-mediated distortion hypothesis. Only one previous study has investigated the perception of sequences of time intervals in individuals with PD, reporting an impairment in the perception of sequences of time intervals in PD (Grahn & Brett, 2009). However, the role of DA- enhancing medications on this impairment was not investigated. While this dissertation found little support for the DA-mediated distortion hypothesis, the findings do suggest that DA- enhancing medications may play a role in correcting potential timing impairments in PD if period correction is necessary for the task.

The goals of this dissertation were twofold. The first goal of this dissertation was to better understand the mechanisms involved in time perception, specifically the role of the BG and DA were investigated. The second goal of this dissertation was to better understand time perception impairments in individuals with PD. Concerning the first goal, two clear findings extend current understanding of the role of the BG and DA in time perception. First, this dissertation found evidence that DA might play a role in period-correction processes. The second finding suggests that damage to the BG does not result in as strong of an effect of rhythmic context as proposed by the DA-mediated distortion hypothesis. However, as discussed previously, the current findings raise several questions to be tested by future research.

149

In regards to the second goal of this dissertation, two main findings inform current understanding of time perception impairments in PD. First, no difference in PC values was found between individuals with PD (both on- and off-medication) and control groups for the one- standard interval sequence in the no-context condition. This finding supports previous research that reports no impairment in the perception of isolated time intervals in individuals with PD.

Second, and related to the first goal of this dissertation, period correction is improved when individuals with PD are on DA-enhancing medications. Based on this finding, one prediction for future research utilizing tasks that require period correction in individuals with PD is that task performance should be improved in individuals with PD, compared to themselves (off medication) and control groups.

In conclusion, the most significant results of this dissertation are that the effects of rhythmic context on perceived duration can be weakened with multiple standard intervals. This finding supported the period-correction hypothesis. However, little support was found for the

DA-mediated distortion hypothesis. No support was found for the prediction that stronger expectancy profiles would be found for individuals with PD, compared to young adults.

However, for individuals with PD, multiple standard intervals were found to flatten the expectancy profile when individuals were on medication, but not when they were off medication.

This finding provided evidence for the potential role of DA in the period-correction hypothesis.

Finally, trend analyses provided some support for the DA-mediated distortion hypothesis. The effects of rhythmic context on perceived duration were found for the single-standard interval sequences in the with-rhythmic-context condition when individuals with PD were tested off medication, compared to on medication. In sum, this dissertation found support for the period- correction hypothesis, but found mixed support for the DA-mediated distortion hypothesis.

150

APPENDIX A: REFERENCES

Abel, S. M. (1972). Discrimination of temporal gaps. Journal of the Acoustical Society of

America, 52(2), 519-524.

Albin, R. L., Young, A. B., & Penney, J. B. (1989). The functional-anatomy of basal ganglia

disorders. Trends in Neurosciences, 12(10), 366-375.

Alexander, G. E., Delong, M. R., & Crutcher, M. D. (1992). Do cortical and basal ganglionic

motor areas use motor programs to control movement. Behavioral and Brain Sciences,

15(4), 656-665.

Allan, L. G. (1979). The perception of time. Perception and Psychophysics, 26(5), 340-354.

Allan, L. G. (1998). The influence of the scalar timing model on human timing research.

Behavioural Processes, 44(2), 101-117.

Allan, L. G., & Church, R. M. (2002). Honoring the career of professor John Gibbons [Special

Issue]. Learning and Motivation, 33(1).

Alpherts, W. C. J., Vermeulen, J., Franken, M. L. O., Hendriks, M. P. H., van Veelen, C. W. M.,

& van Rijen, P. C. (2002). Lateralization of auditory rhythm length in temporal lobe

lesions. Brain and Cognition, 49(1), 114-122. doi: 10.1006/brcg.2001.1490

Antal, A., Bandini, P., Keri, S., & Bodis-Wollner, I. (1998). Visuo-cognitive dysfunctions in

Parkinson's disease. Clinical Neuroscience, 5(2), 147-152.

Artieda, J., Pastor, M. A., Lacruz, F., & Obeso, J. A. (1992). Temporal discrimination is

abnormal in Parkinson's disease. Brain, 115(1), 199-210.

Barnes, R., & Jones, M. R. (2000). Expectancy, attention, and time. Cognitive Psychology, 41(3),

254-311.

151

Baudouin, A., Vanneste, S., & Isingrini, M. (2004). Age-related cognitive slowing: The role of

spontaneous tempo and processing speed. Experimental Aging Research, 30(3), 225-239.

Belin, P., McAdams, S., Thivard, L., Smith, B., Savel, S., Zilbovicius, M., et al. (2002). The

neuroanatomical substrate of sound duration discrimination. Neuropsychologia, 40(12),

1956-1964.

Benabid, A. L. (2003). Deep brain stimulation for Parkinson's disease. Current Opinion in

Neurobiology, 13(6), 696-706. doi: 10.1016/j.conb.2003.11.001

Benabid, A. L., Chabardes, S., Mitrofanis, J., & Pollak, P. (2009). Deep brain stimulation of the

subthalamic nucleus for the treatment of Parkinson's disease. Lancet Neurology, 8(1), 67-

81.

Bergman, H., & Deuschl, G. (2002). Pathophysiology of Parkinson's disease: From clinical

neurology to basic neuroscience and back. Movement Disorders, 17(Supplement 3), S28-

S40.

Block, R. A., & Zakay, D. (1997). Prospective and retrospective duration judgments: A meta-

analytic review. Psychonomic Bulletin and Review, 4(2), 184-197.

Block, R. A., Zakay, D., & Hancock, P. A. (1998). Human aging and duration judgments: A

meta-analytic review. Psychology and Aging, 13(4), 584-596.

Bloem, B. R., Hausdorff, J. A., Visser, J. E., & Giladi, N. (2004). Falls and freezing of gait in

Parkinson's disease: A review of two interconnected, episodic phenomena. Movement

Disorders, 19(8), 871-884.

Bodis-Wollner, I. (2003). Neuropsychological and perceptual defects in Parkinson's disease.

Parkinsonism and Related Disorders, 9(1), S83-S89.

152

Bodis-Wollner, I., & Jo, M. Y. (2006). Getting around and communicating with the environment:

Visual cognition and language in Parkinson's disease. Journal of Neural Transmission,

70(Supplement), 333-338.

Bohnen, N. I., Minoshima, S., Giordani, B., Frey, K. A., & Kuhl, D. E. (1999). Motor correlates

of occipital glucose hypometabolism in Parkinson's disease without dementia. Neurology,

52(3), 541-546.

Boltz, M. G. (1994). Changes in internal tempo and effects on the learning and remembering of

event durations. Journal of Experimental Psychology: Learning, Memory, and Cognition,

20(5), 1154-1171.

Brown, P., & Marsden, C. D. (1998). What do the basal ganglia do? Lancet, 351, 1801-1804.

Brown, S. W. (1997). Attentional resources in timing: Interference effects in concurrent temporal

and nontemporal working memory tasks. Perception and Psychophysics, 59(7), 1118-

1140.

Brown, S. W., & Stubbs, D. A. (1988). The psychophysics of retrospective and prospective

timing. Perception, 17(3), 297-310.

Brunia, C. H. M., de Jong, B. M., van den Berg-Lenssen, M. M. C., & Paans, A. M. J. (2000).

Visual feedback about time estimation is related to a right hemisphere activation

measured by PET. Experimental Brain Research, 130(3), 328-337.

Bueti, D., Bahrami, B., & Walsh, V. (2008). Sensory and association cortex in time perception.

Journal of Cognitive Neuroscience, 20(6), 1054-1062.

Buhusi, C. V. (2003). Dopaminergic mechanisms of interval timing and attention. In W. H.

Meck (Ed.), Functional and neural mechanisms of interval timing. Boca Raton, FL: CRC

Press.

153

Buhusi, C. V., & Meck, W. H. (2005). What makes us tick? Functional and neural mechanisms

of interval timing. Nature Reviews Neuroscience, 6(10), 755-765.

Cerella, J. (2001). Aging and information rate. In J. E. Birren & K. W. Schaie (Eds.), Handbook

of the psychology of aging (3rd ed., pp. 201-221). San Diego, CA: Academic Press.

Church, R. M. (1984). Properties of the internal clock. Annals of the New York Academy of

Sciences, 423, 566-582.

Cohen, J. (1967). Psychological Time in Health and Disease. Springfield, IL: C. C. Thomas.

Collyer, C. E., Broadbent, H. A., & Church, R. M. (1994). Preferred rates of repetitive tapping

and categorical time production. Perception and Psychophysics, 55(4), 443-453.

Coull, J. T. (2004). fMRI studies of temporal attention: allocating attention within, or towards,

time. Cognitive Brain Research, 21(2), 216-226.

Coull, J. T., & Nobre, A. C. (1998). Where and when to pay attention: The neural systems for

directing attention to spatial locations and to time intervals as revealed by both PET and

fMRI. Journal of Neuroscience, 18(18), 7426-7435.

Coull, J. T., Vidal, F., Nazarian, B., & Macar, F. (2004). Functional anatomy of the attentional

modulation of time estimation. Science, 303(5663), 1506-1508.

Craik, F. I. M., & Hay, J. F. (1999). Aging and judgments of duration: Effects of task complexity

and method of estimation. Perception and Psychophysics, 61(3), 549-560.

Craik, F. I. M., & Jennings, J. M. (1992). Human memory. In F. I. M. Craik & T. A. Salthouse

(Eds.), The Handbook of Aging and Cognition. (pp. 51-110). Mahwah, NJ Lawrence

Erlbaum Associates.

Crystal, J. D. (2007). The psychology of time: A tribute to the contributions of Russell M.

Church [Special issue]. Behavioural Processes, 74(2).

154

Defer, G., Widner, H., Marie, R., Remy, P., & Levivier, M. (1999). Core assessment program for

surgical interventional therapies Parkinson's disease (CAPSIT-PD). Movement Disorders,

14, 572-584.

Delong, M. R. (1990). Primate models of movement-disorders of basal ganglia origin. Trends in

Neurosciences, 13(7), 281-285.

Diederich, N. J., Alesch, F., & Goetz, C. G. (2000). Visual hallucinations induced by deep brain

stimulation in Parkinson's disease. Clinical Neuropharmacology, 23(5), 287-289.

Diedrichsen, J., Ivry, R. B., & Pressing, J. (2003). Cerebellar and basal ganglia contributions to

interval timing. In W. H. Meck (Ed.), Functional and Neural Mechanisms of Interval

Timing (pp. 457-484). Boca Raton, FL: CRC Press.

Drake, C., & Botte, M. (1993). Tempo sensitivity in auditory sequences: Evidence for a multiple-

look model. Perception and Psychophysics, 54(3), 277-286.

Drake, C., Jones, M. R., & Baruch, C. (2000). The development of rhythmic attending in

auditory sequences: Attunement, referent period, focal attending. Cognition, 77(3), 251-

288.

Drane, D. L., Lee, G. P., Loring, D. W., & Meador, K. J. (1999). Time perception following

unilateral amobarbital injection in patients with temporal lobe epilepsy. Journal of

Clinical and Experimental Neuropsychology, 21(3), 385-396.

Droit-Volet, S., & Wearden, J. H. (2001). Temporal bisection in children. Journal of

Experimental Child Psychology, 80(2), 142-159.

Dubois, B., & Pillon, B. (1997). Cognitive deficits in Parkinson's disease. Journal of Neurology,

244(1), 2-8.

155

Duchek, J. M., Balota, D. A., & Ferraro, F. R. (1994). Component analysis of a rhythmic finger

tapping task in individuals with senile dementia of the Alzheimer type and in individuals

with Parkinson's disease. Neuropsychology, 8(2), 218-226.

Eisler, H. (1976). Experiments on subjective duration 1868-1975: Collection of power function

exponents. Psychological Bulletin, 83(6), 1154-1171.

Elsinger, C. L., Rao, S. M., Zimbelman, J. L., Reynolds, N. C., Blindauer, K. A., & Hoffmann,

R. G. (2003). Neural basis for impaired time reproduction in Parkinson's disease: An

fMRI study. Journal of the International Neuropsychological Society, 9(7), 1088-1098.

Fahn, S. (2008). The history of dopamine and levodopa in the treatment of Parkinson's disease.

Movement Disorders, 23, S497-S508. doi: 10.1002/mds.22028

Fahn, S., & Elton, R. (1987). The UPDRS Development Committee Recent Developments in

Parkinson’s Disease (Vol. 2, pp. 153-163). Florham Park: MacMillan.

Fearnley, J. M., & Lees, A. J. (1991). Aging and Parkinson's disease: Substantia nigra regional

selectivity. Brain, 114(5), 2283-2301.

Ferrandez, A. M., Hugueville, L., Lehericy, S., Poline, J. B., Marsault, C., & Pouthas, V. (2003).

Basal ganglia and supplementary motor area subtend duration perception: an fMRI study.

Neuroimage, 19(4), 1532-1544. doi: 10.1016/s1053-8119(03)00159-9

Fitzgibbons, P. J., & Gordon-Salant, S. (2001). Aging and temporal discrimination in auditory

sequences. Journal of the Acoustical Society of America, 109(6), 2955-2963.

Fortin, C. (1999). Short-term memory in time interval production. International Journal of

Psychology, 34(5-6), 308-316.

156

Fortin, C., & Rousseau, R. (1998). Interference from short-term memory processing on encoding

and reproducing brief durations. Psychological Research/Psychologische Forschung,

61(4), 269-276.

Fortin, C., Rousseau, R., Bourque, P., & Kirouac, E. (1993). Time-estimation and concurrent

nontemporal processing: Specific interference from short-term-memory demands.

Perception and Psychophysics, 53(5), 536-548.

Fraisse, P. (1963). The Psychology of Time (J. Leith, Trans.). New York: Harper & Row.

Fraisse, P. (1978). Time and rhythm perception. In E. C. Carterette & M. P. Friedman (Eds.),

Handbook of Perception (Vol. 8, pp. 203-254). New York: Academic Press.

Fraisse, P. (1982). Rhythm and Tempo. In D. Deutsch (Ed.), The Psychology of Music (pp. 149-

180). New York: Academic Press, Inc.

Freeman, J. S., Cody, F. W. J., & Schady, W. (1993). The influence of external timing cues upon

the rhythm of voluntary movements in Parkinson's disease. Journal of Neurology,

Neurosurgery, and Psychiatry, 56(10), 1078-1084.

Friberg, A., & Sundberg, J. (1995). Time discrimination in a monotonic, isochronous sequence.

Journal of the Acoustical Society of America, 98(5, Pt 1), 2524-2531.

Funnell, M. G., Corballis, P. M., & Gazzaniga, M. S. (2003). Temporal discrimination in the

split brain. Brain and Cognition, 53(2), 218-222. doi: 10.1016/s0278-2626(03)00113-1

Getty, D. J. (1975). Discrimination of short temporal intervals: A comparison of two models.

Perception and Psychophysics, 18(1), 1-8.

Gibbon, J. (1977). Scalar expectancy theory and Weber's law in animal timing. Psychological

Review, 84(3), 279-325.

157

Gibbon, J., Church, R. M., & Meck, W. H. (1984). Scalar timing in memory. Annals of the New

York Academy of Sciences, 423, 52-77.

Goberman, A. M., & Coelho, C. (2002). Acoustic analysis of Parkinsonian speech II: L-Dopa

related fluctuations and methodological issues. NeuroRehabilitation, 17(3), 247-254.

Goetz, C., Koller, W., Poewe, W., Rascol, O., & Sampaio, C. (2002). Management of

Parkinson's disease: An evidence-based review. Movement Disorders, 17, S1-S166. doi:

10.1002/mds.5555

Goldstone, S., & Goldfarb, J. L. (1964a). Auditory and visual time judgment. Journal of General

Psychology, 70(2), 369-387.

Goldstone, S., & Goldfarb, J. L. (1964b). Direct comparisons of auditory and visual durations.

Journal of Experimental Psychology, 67(5), 483-485.

Grahn, J. A., & Brett, M. (2005). The role of the basal ganglia in beat-based rhythm processing.

Journal of Cognitive Neuroscience, 19(5), 893-906.

Grahn, J. A., & Brett, M. (2007). Rhythm and beat perception in motor areas of the brain.

Journal of Cognitive Neuroscience, 19(5), 1-14.

Grahn, J. A., & Brett, M. (2009). Impairment of beat-based rhythm discrimination in Parkinson's

disease. Cortex, 45(1), 54-61. doi: 10.1016/j.cortex.2008.01.005

Grahn, J. A., & McAuley, J. D. (2009). Neural bases of individual differences in beat perception.

Neuroimage, 47(4), 1894-1903. doi: 10.1016/j.neuroimage.2009.04.039

Greene, L. S., & Wiliams, H. G. (1993). Age-related differences in timing control of repetitive

movement: Application of the Wing-Kristofferson model. Research Quarterly for

Exercise and Sport, 64(1), 32-38.

158

Grondin, S. (1992). Production of time intervals from segmented and nonsegmented inputs.

Perception and Psychophysics, 52(3), 345-350.

Grondin, S. (2001a). Discriminating time intervals presented in sequences marked by visual

signals. Perception and Psychophysics, 63(7), 1214-1228.

Grondin, S. (2001b). From physical time to the first and second moments of psychological time.

Psychological Bulletin, 127(1), 22-44. doi: 10.1037//0033-2909.127.1.22

Grondin, S. (2010). Timing and time perception: A review of recent behavioral and neuroscience

findings and theoretical directions. Attention, Perception, and Psychophysics, 72(3), 561-

582. doi: 10.3758/APP.72.3.561

Grondin, S., & Girard, C. (2005). About hemispheric differences in the processing of temporal

intervals. Brain and Cognition, 58(1), 125-132. doi: 10.1016/j.bandc.2004.11.001

Grondin, S., & McAuley, J. D. (2009). Duration discrimination in crossmodal sequences.

Perception, 38(10), 1542-1559. doi: 10.1068/p6359

Grondin, S., & Plourde, M. (2007). Judging multi-minute intervals retrospectively. Quarterly

Journal of Experimental Psychology, 60(9), 1303-1312. doi:

10.1080/17470210600988976

Guehl, D., Burbaud, P., Lorenzi, C., Ramos, C., Bioulac, B., Semal, C., et al. (2008). Auditory

temporal processing in Parkinson's disease. Neuropsychologia, 46(9), 2326-2335. doi:

10.1016/j.neuropsychologia.2008.03.007

Halpern, A. R., & Darwin, C. J. (1982). Duration discrimination in a series of rhythmic events.

Perception and Psychophysics, 31(1), 86-89.

159

Halsband, U., Ito, N., Tanji, J., & Freund, H. (1993). The role of premotor cortex and the

supplementary motor area in the temporal control of movement in man. Brain, 116, 243-

266.

Handy, T. C., Gazzaniga, M. S., & Ivry, R. B. (2003). Cortical and subcortical contributions to

the representation of temporal information. Neuropsychologia, 41(11), 1461-1473.

Harrington, D. L., Boyd, L. A., Mayer, A. R., Sheltraw, D. M., Lee, R. R., Huang, M. X., et al.

(2004). Neural representation of interval encoding and decision making. Cognitive Brain

Research, 21(2), 193-205.

Harrington, D. L., & Haaland, K. Y. (1999). Neural underpinnings of temporal processing: A

review of focal lesion, pharmacological, and functional imaging research. Reviews in the

Neurosciences, 10(2), 91-116.

Harrington, D. L., Haaland, K. Y., & Hermanowicz, N. (1998). Temporal processing in the basal

ganglia. Neuropsychology, 12(1), 3-12.

Harrington, D. L., Haaland, K. Y., & Knight, R. T. (1998). Cortical networks underlying

mechanisms of time perception. The Journal of Neuroscience, 18(3), 1085-1095.

Harrington, D. L., Lee, R. R., Boyd, L. A., Rapcsak, S. Z., & Knight, R. T. (2004). Does the

representation of time depend on the cerebellum? Effect of cerebellar stroke. Brain, 127,

561-574.

Hawkes, C. H., Shephard, B. C., & Daniel, S. E. (1997). Olfactory dysfunction in Parkinson's

disease. Journal of Neurology, Neurosurgery, and Psychiatry, 62(5), 436-446.

Hellstrom, A., & Rammsayer, T. H. (2004). Effects of time-order, interstimulus interval, and

feedback in duration discrimination of noise bursts in the 50-and 1000-ms ranges. Acta

Psychologica, 116(1), 1-20. doi: 10.1016/j.actpsy.2003.11.003

160

Helson, H. (1964). Adaptation-Level Theory. New York: Harper & Row.

Henderson, R., Kurlan, R., Kersun, J. M., & Como, P. G. (1992). Preliminary examination of the

comorbidity of anxiety and depession in Parkinson's disease. Journal of Neuropsychiatry

and Clinical Neurosciences, 4(3), 257-264.

Hershey, T., Revilla, F. J., Wernle, A., Gibson, P. S., Dowling, J. L., & Perlmutter, J. S. (2004).

Stimulation of STN impairs aspects of cognitive control in PD. Neurology, 62(7), 1110-

1114.

Hibbard, T. R., Migliaccio, J. N., Goldstone, S., & Lhamon, W. T. (1975). Temporal information

processing by young and senior adults and patients with senile dementia. Journal of

Gerontology, 30(3), 326-330.

Hicks, R. E. (1992). Prospective and retrospective judgments of time: A neurobehavioral

analysis. In F. Macar, V. Pouthas & W. J. Friedman (Eds.), Time, Action, and Cognition:

Towards Bridging the Gap (pp. 97-108). Norwell, MA: Kluwer.

Hicks, R. E., Miller, G. W., & Kinsbourne, M. (1976). Prospective and retrospective judgments

of time as a function of amount of information processed. American Journal of

Psychology, 89(4), 719-730.

Hinton, S. C., & Rao, S. M. (2004). "One-thousand one... one-thousand two...": chronometric

counting violates the scalar property in interval timing. Psychonomic Bulletin and

Review, 11(1), 24-30.

Hirsh, I. J., Monahan, C. B., Grant, K. W., & Singh, P. G. (1990). Studies in auditory timing. 1.

Simple patterns. Perception and Psychophysics, 47(3), 215-226.

Hoehn, M. M., & Yahr, M. D. (1967). Parkinsonism: Onset, progression and mortality.

Neurology and Clinical Neurophysiology, 17(5), 427-442.

161

Hollingworth, H. L. (1910). The central tendency of judgment. The Journal of Philosophy,

Psychology, and Scientific Methods, 7(17), 461-469.

Ivry, R., & Corcos, D. M. (1993). Slicing the variability pie: Component analysis of coordination

and motor dysfunction. In K. M. Newell & D. M. Corcos (Eds.), Variability and Motor

Control (pp. 415-447).

Ivry, R. B. (1996). The representation of temporal information in perception and motor control.

Current Opinion in Neurobiology, 6(6), 851-857.

Ivry, R. B., & Hazeltine, R. E. (1995). Perception and production of temporal intervals across a

range of durations: Evidence for a common timing mechanism. Journal of Experimental

Psychology: Human Perception and Performance, 21(1), 3-18.

Ivry, R. B., & Keele, S. W. (1989). Timing functions of the cerebellum. Journal of Cognitive

Neuroscience, 1(2), 136-152.

Ivry, R. B., Keele, S. W., & Diener, H. C. (1988). Dissociation of the lateral and medial

cerebellum in movement timing and movement execution. Experimental Brain Research,

73(1), 167-180.

Jaggi, J. L., Umemura, A., Hurtig, H. I., Siderowf, A. D., Colcher, A., Stern, M. B., et al. (2004).

Bilateral stimulation of the subthalamic nucleus in Parkinson's disease: Surgical efficacy

and prediction of outcome. Stereotactic and Functional Neurosurgery, 82(2-3), 104-114.

doi: 10.1159/000078145

Jahanshahi, M., Jones, C. R. G., Dirnberger, G., & Frith, C. D. (2006). The substantia nigra pars

compacta and temporal processing. Journal of Neuroscience, 26(47), 12266-12273. doi:

10.1523/jneurosci.2540-06.2006

James, W. (1890). The Principles of Psychology (Vol. 1). New York: Holt.

162

Jancke, L., Loose, R., Lutz, K., Specht, K., & Shah, N. J. (2000). Cortical activations during

paced finger-tapping applying visual and auditory pacing stimuli. Cognitive Brain

Research, 10(1-2), 51-66.

Jennings, J. M., & Jacoby, L. L. (1993). Automatic versus intentional uses of memory: Aging,

attention, and control. Psychology and Aging, 8(2), 283-293.

Jones, C. R. G., Malone, T. J. L., Dirnberger, G., Edwards, M., & Jahanshahi, M. (2008). Basal

ganglia, dopamine and temporal processing: Performance on three timing tasks on and

off medication in Parkinson's disease. Brain and Cognition, 68(1), 30-41. doi:

10.1016/j.bandc.2008.02.121

Jones, M. R. (1976). Time, our lost dimension: Toward a new theory of perception, attention,

and memory. Psychological Review, 83(5), 323-355.

Jones, M. R. (2004). Attention and timing. In J. G. Neuhoff (Ed.), Ecological Psychoacoustics

(pp. 49-85). San Diego: Elsevier Academic Press.

Jones, M. R., & McAuley, J. D. (2005). Time judgments in global temporal contexts. Perception

and Psychophysics, 67(3), 398-417.

Jueptner, M., Flerich, L., Weiller, C., Mueller, S. P., & Diener, H. C. (1996). The human

cerebellum and temporal information processing: Results from a PET experiment.

Neuroreport, 7(15-17), 2761-2765.

Jueptner, M., Rijntjes, M., Weiller, C., Faiss, J. H., Timmann, D., Mueller, S. P., et al. (1995).

Localization of a cerebellar timing process using PET. Neurology, 45(8), 1540-1545.

Jurica, P. J., Leitten, C. L., & Mattis, S. (2001). Dementia Rating Scale-2: Professional manual.

Lutz, FL: Psychological Assessment Resources.

163

Kagerer, F. A., Wittmann, M., Szelag, E., & Von Steinbuchel, N. (2002). Cortical involvement

in temporal reproduction: Evidence for differential roles of the hemispheres.

Neuropsychologia, 40(3), 357-366.

Kandel, E. R., Schwartz, J. H., & Jessell, T. M. (2000). Principles of Neuroscience (4th ed.).

New York: McGraw-Hill.

Katzenschlager, R., Sampaio, C., Costa, A., & Lees, A. (2003). Anticholinergics for

Symptomatic Management of Parkinson's Disease. Cochrane Database Syst Rev(3).

Koch, G., Brusa, L., Caltagirone, C., Oliveri, M., Peppe, A., Tiraboschi, P., et al. (2004).

Subthalamic deep brain stimulation improves time perception in Parkinson's disease.

Neuroreport, 15(6), 1071-1073. doi: 10.1097/01.wnr.0000123645.28957f9

Koch, G., Brusa, L., Oliveri, M., Stanzione, P., & Caltagirone, C. (2005). Memory for time

intervals is impaired in left hemi-Parkinson patients. Neuropsychologia, 43(8), 1163-

1167. doi: 10.1016/j.neuropsychologia.2004.11.017

Koch, G., Oliveri, M., Carlesimo, G. A., & Caltagirone, C. (2002). Selective deficit of time

perception in a patient with right prefrontal cortex lesion. Neurology, 59(10), 1658-1659.

Koch, G., Oliveri, M., Torriero, S., & Caltagirone, C. (2003). Underestimation of time

perception after repetitive transcranial magnetic stimulation. Neurology, 60(11), 1844-

1846.

Koch, G., Oliveri, M., Torriero, S., Salerno, S., Lo Gerfo, E., & Caltagirone, C. (2007).

Repetitive TMS of cerebellum interferes with millisecond time processing. Experimental

Brain Research, 179(2), 291-299. doi: 10.1007/s00221-006-0791-1

Kolers, P. A., & Brewster, J. M. (1985). Rhythms and responses. Journal of Experimental

Psychology: Human Perception and Performance, 11(2), 150-167.

164

Konczak, J., Ackermann, H., Hertrich, I., Spieker, S., & Dichgans, J. (1997). Control of

repetitive lip and finger movements in Parkinson's disease: Influence of external timing

signals and simultaneous execution on motor performance. Movement Disorders, 12(5),

665-676.

Kringelbach, M. L., Owen, S. L. F., & Aziz, T. Z. (2007). Deep-brain stimulation. Future

Neurology, 2(6), 633 - 646.

Kudo, K., Miyazaki, M., Kimura, T., Yamanaka, K., Kadota, H., Hirashima, M., et al. (2004).

Selective activation and deactivation of the human brain structures between speeded and

precisely timed tapping responses to identical visual stimulus: an fMRI study.

Neuroimage, 22(3), 1291-1301. doi: 10.1016/j.neuroimnage.2004.03.043

Laatu, S., Revonsuo, A., Pihko, L., Portin, R., & Rinne, J. O. (2004). Visual object recognition

deficits in early Parkinson's disease. Parkinsonism and Related Disorders, 10(4), 227-

233.

Labelle, M. A., Graf, P., Grondin, S., & Gagne-Roy, L. (2009). Time-related processes in time-

based prospective memory and in time-interval production. European Journal of

Cognitive Psychology, 21(4), 501-521. doi: 10.1080/09541440802031000

Lang, A. E., & Lozano, A. M. (1998a). Parkinson's disease--First of two parts. New England

Journal of Medicine, 339(15), 1044-1053.

Lang, A. E., & Lozano, A. M. (1998b). Parkinson's disease--Second of two parts. New England

Journal of Medicine, 339(16), 1130-1143.

Lange, K. W., Tucha, O., Steup, A., Gsell, W., & Naumann, M. (1995). Subjective time

estimation in Parkinson's disease. Journal of Neural Transmission-Supplement(46), 433-

438.

165

Large, E. W., & Jones, M. R. (1999). The dynamics of attending: How people track time-varying

events. Psychological Review, 106(1), 119-159.

Larsson, J., Gulyas, B., & Roland, P. E. (1996). Cortical representation of self-paced finger

movement. Neuroreport, 7(2), 463-468.

Laskowska, I., Dalla Bella, S., Rolinska, P., Litwinowicz, A., Gorzelanczyk, E., & Harat, M.

(2007). Effects of pallidotomy on sensorimotor synchronization in Parkinson's Disease.

Supplement of the Journal of Cognitive Neuroscience, 151.

Lee, K. H., Egleston, P. N., Brown, W. H., Gregory, A. N., Barker, A. T., & Woodruff, P. W. R.

(2007). The role of the cerebellum in subsecond time perception: Evidence from

repetitive transcranial magnetic stimulation. Journal of Cognitive Neuroscience, 19(1),

147-157.

Lees, A. (2005). Alternatives to levodopa in the initial treatment of early Parkinson's disease.

Drugs and Aging, 22(9), 731-740.

Lejeune, H. (1998). Switching or gating? The attentional challenge in cognitive models of

psychological time. Behavioural Processes, 44(2), 127-145.

Lewis, P. A., & Miall, R. C. (2002). Brain activity during non-automatic motor production of

discrete multi-second intervals. Neuroreport, 13(14), 1731-1735.

Lewis, P. A., & Miall, R. C. (2003a). Brain activation patterns during measurement of sub- and

supra-second intervals. Neuropsychologia, 41(12), 1583-1592.

Lewis, P. A., & Miall, R. C. (2003b). Distinct systems for automatic and cognitively controlled

time measurement: Evidence from neuroimaging. Current Opinion in Neurobiology,

13(2), 250-255.

166

Lewis, P. A., & Miall, R. C. (2006). A right hemispheric prefrontal system for cognitive time

measurement. Behavioural Processes, 71(2-3), 226-234. doi:

10.1016/j.beproc.2005.12.009

Lewis, P. A., Wing, A. M., Pope, P. A., Praamstra, P., & Miall, R. C. (2004). Brain activity

correlates differentially with increasing temporal complexity of rhythms during

initialisation, synchronisation, and continuation phases of paced finger tapping.

Neuropsychologia, 42, 1301–1312.

Loher, T. J., Burgunder, J. M., Weber, S., Sommerhaider, R., & Krauss, J. K. (2002). Effect of

chronic pallidal deep brain stimulation on off period dystonia and sensory symptoms in

advanced Parkinson's disease. Journal of Neurology Neurosurgery and Psychiatry, 73(4),

395-399.

Lozano, A. M., Dostrovsky, J., Chen, R., & Ashby, P. (2002). Deep brain stimulation for

Parkinson's disease: Disrupting the disruption. Lancet Neurology, 1(4), 225-231.

Lustig, C. (2003). Grandfather's clock: Attention and interval timing in older adults. In W. H.

Meck (Ed.), Functional and Neural Mechanisms of Interval Timing. Boca Raton, FL:

CRC Press.

Lustig, C., & Meck, W. H. (2001). Paying attention to time as one gets older. Psychological

Science, 12(6), 478-484.

Lustig, C., & Meck, W. H. (2002). Age deficits and sparing in simultaneous temporal processing.

Journal of Cognitive Neuroscience, E5.

Macar, F., Grondin, S., & Casini, L. (1994). Controlled attention sharing influences time-

estimation. Memory and Cognition, 22(6), 673-686.

167

Macar, F., & Vidal, F. (2009). Timing processes: An outline of behavioural and neural indices

not systematically considered. Canadian Journal of Experimental Psychology, 63(3),

227-239.

Madison, G. (2001). Variability in isochronous tapping: Higher order dependencies as a function

of intertap interval. Journal of Experimental Psychology: Human Perception and

Performance, 27(2), 411-422. doi: 10.1037//0096-1523.27.2.2411

Malapani, C., Deweer, B., & Gibbon, J. (2002). Separating storage from retrieval dysfunction of

temporal memory in Parkinson's disease. Journal of Cognitive Neuroscience, 14(2), 311-

322.

Malapani, C., Rakitin, B., Levy, R., Meck, W. H., Deweer, B., Dubois, B., et al. (1998). Coupled

temporal memories in Parkinson's disease: A dopamine-related dysfunction. Journal of

Cognitive Neuroscience, 10(3), 316-331.

Malapani, C., & Rakitin, B. C. (2003). Interval timing in the dopamine-depleted basal ganglia:

From empirical data to timing theory. In W. H. Meck (Ed.), Functional and Neural

Mechanisms of Interval Timing (pp. 485-514). Boca Raton, FL: CRC Press.

Mangels, J. A., Ivry, R. B., & Shimizu, N. (1998). Dissociable contributions of the prefrontal and

neocerebellar cortex to time perception. Cognitive Brain Research, 7(1), 15-39.

Manyam, B., & Sanchez-Ramos, J. (1999). Traditional and complementary therapies in

Parkinson's disease. Advances in Neurology, 80, 565-574.

Maricq, A. V., Roberts, S., & Church, R. M. (1981). Methamphetamine and time estimation.

Journal of Experimental Psychology: Animal Behavior Processes, 7(1), 18-30.

Mates, J., Muller, U., Radil, T., & Poppel, E. (1994). Temporal integration in sensorimotor

synchronization. Journal of Cognitive Neuroscience, 6(4), 332-340.

168

McAuley, J. D. (1995). Perception of time as phase: Toward an adaptive oscillator model of

rhythmic pattern processing. Unpublished Dissertation, Indiana University,

Bloomington.

McAuley, J. D., & Jones, M. R. (2003). Modeling effects of rhythmic context on perceived

duration: A comparison of interval and entrainment approaches to short-interval timing.

Journal of Experimental Psychology: Human Perception and Performance, 29(6), 1102-

1125. doi: 10.1037/0096-1523.29.6.1102

McAuley, J. D., Jones, M. R., Holub, S., Johnston, H. M., & Miller, N. S. (2006). The time of

our lives: Life span development of timing and event tracking. Journal of Experimental

Psychology: General, 135(3), 348-367. doi: 10.1037/0096-3445.135.3.348

McAuley, J. D., & Kidd, G. R. (1998). Effect of deviations from temporal expectations on tempo

discrimination of isochronous tone sequences. Journal of Experimental Psychology:

Human Perception and Performance, 24(6), 1786-1800.

McAuley, J. D., & Miller, N. S. (2007). Picking up the pace: Effects of global temporal context

on sensitivity to the tempo of auditory sequences. Perception and Psychophysics, 69(5),

709-718.

McCormack, T., Brown, G. D. A., Maylor, E. A., Darby, R. J., & Green, D. (1999).

Developmental changes in time estimation: Comparing childhood and old age.

Developmental Psychology, 35(4), 1143-1155.

Meck, W. H. (1983). Selective adjustment of the speed of internal clock and memory processes.

Journal of Experimental Psychology: Animal Behavior Processes, 9(2), 171-201.

169

Meck, W. H. (1986). Affinity for the dopamine D2 receptor predicts neuroleptic potency in

decreasing the speed of an internal clock. Pharmacology, Biochemistry, and Behavior,

25(6), 1185-1189.

Meck, W. H. (1996). Neuropharmacology of timing and time perception. Cognitive Brain

Research, 3(3-4), 227-242.

Meck, W. H., & Benson, A. M. (2002). Dissecting the brain's internal clock: How frontal-striatal

circuitry keeps time and shifts attention. Brain and Cognition, 48(1), 195-211. doi:

10.1006/brcg.2001.1313

Meck, W. H., Hinton, S. C., & Matell, M. S. (1998). Coincidence-detection models of interval

timing: Evidence from fMRI studies of cortico-striatal circuits. NeuroImage, 7(4), S281.

Meck, W. H., & Malapani, C. (2004). Neuroimaging of interval timing. Cognitive Brain

Research, 21(2), 133-137. doi: 10.1016/j.cogbrainres.2004.07.010

Merchant, H., Luciana, M., Hooper, C., Majestic, S., & Tuite, P. (2008). Interval timing and

Parkinson's disease: Heterogeneity in temporal performance. 184(2), 233-248.

Michon, J. A. (1964). Studies on subjective duration: I. Differential sensitivity in the perception

of repeated temporal intervals. Acta Psychologica, 22, 441-450.

Middleton, F. A., & Strick, P. L. (2000a). Basal ganglia and cerebellar loops: Motor and

cognitive circuits. Brain Research Reviews, 31(2-3), 236-250.

Middleton, F. A., & Strick, P. L. (2000b). Basal ganglia output and cognition: Evidence from

anatomical, behavioral, and clinical studies. Brain and Cognition, 42(2), 183-200.

Miller, N. S., & McAuley, J. D. (2005). Tempo sensitivity in isochronous tone sequences: The

multiple-look model revisited. Perception and Psychophysics, 67(7), 1150-1160.

Mishima, J. (1956). On the factors of mental tempo. Japanese Psychological Research, 4, 27-38.

170

Miyake, Y., Onishi, Y., & Poppel, E. (2004). Two types of anticipation in synchronization

tapping. Acta Neurobiologiae Experimentalis, 64(3), 415-426.

Montgomery, E. B. (2005). Effect of subthalamic nucleus stimulation patterns on motor

performance in Parkinson's disease. Parkinsonism and Related Disorders, 11(3), 167-

171. doi: 10.1016/j.parkreldis.2004.12.002

Montgomery, E. B., & Gale, J. T. (2007). Neurophysiology and neurocircuitry. In R. Pahwa &

K. E. Lyons (Eds.), Handbook of Parkinson's Disease (4th ed.). New York: Informa

Healthcare USA.

Montgomery, E. B., & Gale, J. T. (2008). Mechanisms of action of deep brain stimulation

(DBS). Neuroscience and Biobehavioral Reviews, 32(3), 388-407. doi:

10.1016/j.neubiorev.2007.06.003

Morris, M. E., Iansek, R., Matyas, T. A., & Summers, J. J. (1996). Stride length regulation in

Parkinson's disease normalization strategies and underlying mechanisms. Brain, 119,

551-568.

Nichelli, P., Alway, D., & Grafman, J. (1996). Perceptual timing in cerebellar degeneration.

Neuropsychologia, 34(9), 863-871.

Nichelli, P., Clark, K., Hollnagel, C., & Grafman, J. (1995). Duration processing after frontal

lobe lesions. In J. Grafman, K. J. Holyoak & F. Boller (Eds.), Structure and Functions of

the Human Prefrontal Cortex (Vol. 769, pp. 183-190). New York: New York Academy

of Sciences.

Nichelli, P., Venneri, A., Molinari, M., Tavani, F., & Grafman, J. (1993). Precision and accuracy

of subjective time estimation in different memory disorders. Brain Research: Cognitive

Brain Research, 1(2), 87-93.

171

Nutt, J. G., & Fellman, J. H. (1984). Pharmacokinetics of levodopa. Clinical

Neuropharmacology, 7(1), 35-49.

O'Boyle, D. J., Freeman, J. S., & Cody, F. W. J. (1996). The accuracy and precision of timing of

self-paced, repetitive movements in subjects with Parkinson's disease. Brain, 119, 51-70.

Obeso, J. A., Olanow, C. W., & Nutt, J. G. (2000). Levodopa motor complications in Parkinson's

disease. Trends in Neurosciences, 23(10), S2-S7.

Papagno, C., Allegra, A., & Cardaci, M. (2004). Time estimation in Alzheimer's disease and the

role of the central executive. Brain and Cognition, 54(1), 18-23.

Pastor, M. A., Artieda, J., Jahanshahi, M., & Obeso, J. A. (1992). Time estimation and

reproduction is abnormal in Parkinson's disease. Brain, 115 (1), 211-225.

Pastor, M. A., Jahanshahi, M., Artieda, J., & Obeso, J. A. (1992). Performance of repetitive wrist

movements in Parkinson's disease. Brain, 115(3), 875-891.

Penhune, V. B., Zatorre, R. J., & Evans, A. C. (1998). Cerebellar contributions to motor timing:

A PET study of auditory and visual rhythm reproduction. Journal of Cognitive

Neuroscience, 10(6), 752-765.

Penhune, V. B., Zatorre, R. J., & Feindel, W. H. (1999). The role of auditory cortex in retention

of rhythmic patterns as studied in patients with temporal lobe removals including

Heschl's gyrus. Neuropsychologia, 37(3), 315-331.

Penney, T. B., Gibbon, J., & Meck, W. H. (2000). Differential effects of auditory and visual

signals on clock speed and temporal memory. Journal of Experimental Psychology:

Human Perception & Performance, 26(6), 1770-1787.

Penney, T. B., & Vaitilingam, L. (2008). Imaging Time. In S. Grondin (Ed.), Psychology of Time

(pp. 261-294). Bingley, U.K.: Emerald Group.

172

PentonVoak, I. S., Edwards, H., Percival, A., & Wearden, J. H. (1996). Speeding up an internal

clock in humans? Effects of click trains on subjective duration. Journal of Experimental

Psychology-Animal Behavior Processes, 22(3), 307-320.

Perbal, S., Deweer, B., Pillon, B., Vidailhet, M., Dubois, B., & Pouthas, V. (2005). Effects of

internal clock and memory disorders on duration reproductions and duration productions

in patients with Parkinson's disease. Brain and Cognition, 58(1), 35-48. doi:

10.1016/j.bandc.2005.02.003

Perbal, S., Droit-Volet, S., Isingrini, M., & Pouthas, V. (2002). Relationships between age-

related changes in time estimation and age-related changes in processing speed, attention,

and memory. Aging Neuropsychology and Cognition, 9(3), 201-216.

Perlmutter, J. S., & Mink, J. W. (2006). Deep brain stimulation. Annual Review of Neuroscience,

29, 229-257. doi: 10.1146/annurev.neuro.29.051605.112824

Peters, M. (1989). The relationship between variability of intertap intervals and interval duration.

Psychological Research/Psychologische Forschung, 51(1), 38-42.

Petrusic, W. M. (1984). Explicit counting and time-order errors in duration discrimination.

Annals of the New York Academy of Sciences, 423(May), 630-633.

Ponsen, M. M., Stoffers, D., Booij, J., van Eck-Smit, B. L., Wolters, E. C., & Berendse, H. W.

(2004). Idiopathic hyposmia as a preclinical sign of Parkinson's disease. Annals of

Neurology, 56(2), 173-181.

Poppel, E. (2004). Lost in time: a historical frame, elementary processing units and the 3-second

window. Acta Neurobiologiae Experimentalis, 64(3), 295-301.

Posner, M. I. (1976). Chronometric Explorations of Mind. Hillsdale, NJ: Erlbaum.

173

Pouthas, V., George, N., Poline, J. B., Pfeuty, M., VandeMoorteele, P. F., Hugueville, L., et al.

(2005). Neural network involved in time perception: An fMRI study comparing long and

short interval estimation. Human Brain Mapping, 25(4), 433-441. doi:

10.1002/hbm.20126

Predebon, J. (1996). The effects of active and passive processing of interval events on

prospective and retrospective time estimates. Acta Psychologica, 94(1), 41-58.

Rakitin, B. C., Gibbon, J., Penney, T. B., Malapani, C., Hinton, S. C., & Meck, W. H. (1998).

Scalar expectancy theory and peak-interval timing in humans. Journal of Experimental

Psychology: Animal Behavior Processes, 24(1), 15-33.

Rakitin, B. C., Scarmeas, N., Li, T. N., Malapani, C., & Stern, Y. (2006). Single-dose levodopa

administration and aging independently disrupt time production. Journal of Cognitive

Neuroscience, 18(3), 376-387.

Rakitin, B. C., Scarmeas, N., Malapani, C., & Stern, Y. (2002). Is an age-related time production

deficit related to that found in Parkinson's disease patients? Journal of Cognitive

Neuroscience, D118.

Rammsayer, T. (1989). Dopaminergic and serotoninergic influence on duration discrimination

and vigilance. Pharmacopsychiatry, 22, 39-43.

Rammsayer, T. (2008). Neuropharmacological approaches to human timing. In S. Grondin (Ed.),

Psychology of Time (pp. 295-320). Bingley, U.K.: Emerald Group.

Rammsayer, T., & Classen, W. (1997). Impaired temporal discrimination in Parkinson's disease:

Temporal processing of brief durations as an indicator of degeneration of dopaminergic

neurons in the basal ganglia. International Journal of Neuroscience, 91(1-2), 45-55.

174

Rammsayer, T. H. (1997). Are there dissociable roles of the mesostriatal and mesolimbocortical

dopamine systems on temporal information processing in humans? Neuropsychobiology,

35(1), 36-45.

Rammsayer, T. H. (2001). Ageing and temporal processing of durations within the psychological

present. European Journal of Cognitive Psychology, 13(4), 549-565.

Rammsayer, T. H., & Vogel, W. H. (1992). Pharmacologic properties of the internal clock

underlying time perception in humans. Neuropsychobiology, 26(1-2), 71-80.

Rao, S. M., Harrington, D. L., Haaland, K. Y., Bobholz, J. A., Cox, R. W., & Binder, J. R.

(1997). Distributed neural systems underlying the timing of movements. Journal of

Neuroscience, 17(14), 5528-5535.

Rao, S. M., Mayer, A. R., & Harrington, D. L. (2001). The evolution of brain activation during

temporal processing. Nature Neuroscience, 4(3), 317-323.

Repp, B. H. (2003). Rate limits in sensorimotor synchronization with auditory and visual

sequences: The synchronization threshold and the benefits and costs of interval

subdivision. Journal of Motor Behavior, 35(4), 355-370.

Repp, B. H. (2005a). Rate limits of on-beat and off-beat tapping with simple auditory rhythms: 1.

Qualitative observations. Music Perception, 22(3), 479-496.

Repp, B. H. (2005b). Sensorimotor synchronization: A review of the tapping literature.

Psychonomic Bulletin and Review, 12(6), 969-992.

Repp, B. H., & Penel, A. (2002). Auditory dominance in temporal processing: New evidence

from synchronization with simultaneous visual and auditory sequences. Journal of

Experimental Psychology: Human Perception and Performance, 28(5), 1085-1099. doi:

10.1037//0096-1523.28.5.1085

175

Repp, B. H., & Penel, A. (2004). Rhythmic movement is attracted more strongly to auditory than

to visual rhythms. Psychological Research/Psychologische Forschung, 68(4), 252-270.

doi: 10.1007/s00426-003-0143-8

Richard, I. H., Schiffer, R. B., & Kurlan, R. (1996). Anxiety and Parkinson's disease. Journal of

Neuropsychiatry, 8(4), 383-392.

Riesen, J. M., & Schnider, A. (2001). Time estimation in Parkinson's disease: Normal long

duration estimation despite impaired short duration discrimination. Journal of Neurology,

248(1), 27-35.

Rimoldi, H. J. A. (1951). Personal tempo. Journal of Abnormal and Social Psychology, 46(3),

283-303.

Rodriguez-Girones, M. A., & Kacelnik, A. (2001). Relative importance of perceptual and

mnemonic variance in human temporal bisection. Quarterly Journal of Experimental

Psychology Section A: Human Experimental Psychology, 54(2), 527-546.

Rubia, K., Overmeyer, S., Taylor, E., Brammer, M., Williams, S., Simmons, A., et al. (1998).

Prefrontal involvement in "temporal bridging" and timing movement. Neuropsychologia,

36(12), 1283-1293.

Salthouse, T. A. (1996). The processing-speed theory of adult age differences in cognition.

Psychological Review, 103(3), 403-428.

Schapira, A. H. V., Bezard, E., Brotchie, J., Calon, F., Collingridge, G. L., Ferger, B., et al.

(2006). Novel pharmacological targets for the treatment of Parkinson's disease. Nature

Reviews Drug Discovery, 5(10), 845-854. doi: 10.1038/nrd2087

176

Schubotz, R. I., Friederici, A. D., & von Cramon, D. Y. (2000). Time perception and motor

timing: A common cortical and subcortical basis revealed by fMRI. Neuroimage, 11(1),

1-12.

Schubotz, R. I., & von Cramon, D. Y. (2001). Functional organization of the lateral premotor

cortex: fMRI reveals different regions activated by anticipation of object properties,

location and speed. Cognitive Brain Research 11(1), 97-112.

Schulze, H. H. (1989). The perception of temporal deviations in isochronic patterns. Perception

and Psychophysics, 45(4), 291-296.

Semjen, A., Vorberg, D., & Schulze, H. H. (1998). Getting synchronized with the metronome:

Comparisons between phase and period correction. Psychological

Research/Psychologische Forschung, 61(1), 44-55.

Simuni, T., & Hurtig, H. (Eds.). (2002). Levodopa: 30 Years of Progress. New York: Demos

Medical Publishing, Inc.

Small, A. M., & Campbell, R. A. (1962). Temporal differential sensitivity for auditory stimuli.

American Journal of Psychology, 75(3), 401-410.

Smith, J. G., Harper, D. N., Gittings, D., & Abernethy, D. (2007). The effect of Parkinson's

disease on time estimation as a function of stimulus duration range and modality. Brain

and Cognition, 64(2), 130-143. doi: 10.1016/j.bandc.2007.01.005

Smoll, F. L., & Schutz, R. W. (1978). Relationships among measures of preferred tempo and

motor rhythm. Perceptual and Motor Skills, 46(3), 883-894.

Spence, C., & Driver, J. (1997). Audiovisual links in exogenous covert spatial orienting.

Perception and Psychophysics, 59(1), 1-22.

177

Spencer, R. M. C., & Ivry, R. B. (2005). Comparison of patients with Parkinson's disease or

cerebellar lesions in the production of periodic movements involving event-based or

emergent timing. Brain and Cognition, 58(1), 84-93. doi: 10.1016/j.bandc.2004.09.010

Spencer, R. M. C., Ivry, R. B., & Zelaznik, H. N. (2005). Role of the cerebellum in movements:

Control of timing or movement transitions? Experimental Brain Research, 161(3), 383-

396. doi: 10.1007/s00221-004-2088-6

Spencer, R. M. C., Zelaznik, H. N., Diedrichsen, J., & Ivry, R. B. (2003). Disrupted timing of

discontinuous but not continuous movements by cerebellar lesions. Science, 300(5624),

1437-1439.

SPSS. (2008). SPSS version 17.0 for Windows Chicago, IL: SPSS, Inc.

Starkstein, S. E., Preziosi, T. J., Bertheir, M. L., Bolduc, P. L., Mayberg, H. S., & Robinson, R.

G. (1989). Depression and cognitive impairments in Parkinson's disease. Brain 112(5),

141-153.

Starr, P. A., Vitek, J. L., & Bakay, R. A. E. (1998). Ablative surgery and deep brain stimulation

for Parkinson's disease. Neurosurgery, 43(5), 989-1013.

Stein, M. B., Heuser, I. J., Juncos, J. L., & Uhde, T. W. (1990). Anxiety disorders in patients

with Parkinson's disease. The American Journal of Psychiatry, 147(2), 217-220.

Steinberg, R., & Raith, L. (1985). Music psychopathology: I. Musical tempo and psychiatric

disease. Psychopathology, 18(5-6), 254-264.

Stevens, S. S. (1957). On the psychological law. Psychological Review, 64, 153-181.

Takeshita, S., Kurisu, K., Trop, L., Arita, K., Akimitsu, T., & Verhoeff, N. (2005). Effect of

subthalamic stimulation on mood state in Parkinson's disease: Evaluation of previous

178

facts and problems. Neurosurgical Review, 28(3), 179-186. doi: 10.1007/s10143-005-

0387-4 ten Hoopen, G., Boelaarts, L., Gruisen, A., Apon, I., & et al. (1994). The detection of

anisochrony in monaural and interaural sound sequences. Perception and Psychophysics,

56(1), 110-120.

Tissingh, G., Bergmans, P., Booij, J., Winogrodzka, A., van Royen, E. A., Stoof, J. C., et al.

(1998). Drug-naive patients with Parkinson's disease in Hoehn and Yahr stages I and II

show a bilateral decrease in striatal dopamine transporters as revealed by [I-123]beta-CIT

SPECT. Journal of Neurology, 245(1), 14-20.

Tracy, J. I., Faro, S. H., Mohamed, F. B., Pinsk, M., & Pinus, A. (2000). Functional localization

of a "time keeper" function separate from attentional resources and task strategy.

Neuroimage, 11(3), 228-242.

Tregellas, J. R., Davalos, D. B., & Rojas, D. C. (2006). Effect of task difficulty on the functional

anatomy of temporal processing. Neuroimage, 32(1), 307-315. doi:

10.1016/j.neuroimage.2006.02.036

Treisman, M. (1963). Temporal discrimination and the indifference interval: Implications for a

model of the "internal clock". Psychological Monographs: General and Applied, 77(13,

Whole No. 576), 1-31.

Truong, D. D., Bhidayasiri, R., & Wolters, E. (2008). Management of non-motor symptoms in

advanced Parkinson disease. Journal of the Neurological Sciences, 266(1-2), 216-228.

doi: 10.1016/j.jns.2007.08.015

179

Vanneste, S., Perbal, S., & Pouthas, V. (1999). Estimation of duration in young and aged

subjects: The role of attentional and memory processes. Annee Psychologique, 99(3),

385-414.

Vanneste, S., & Pouthas, V. (1999). Timing in aging: The role of attention. Experimental Aging

Research, 25(1), 49-67.

Vanneste, S., Pouthas, V., & Wearden, J. H. (2001). Temporal control of rhythmic performance:

A comparison between young and old adults. Experimental Aging Research, 27(1), 83-

102.

Vidalaki, V. N., Ho, M. Y., Bradshaw, C. M., & Szabadi, E. (1999). Interval timing performance

in temporal lobe epilepsy: Differences between patients with left and right hemisphere

foci. Neuropsychologia, 37(9), 1061-1070.

Visser-Vandewalle, V., Temel, Y., van der Linden, C., Ackermans, L., & Beuls, E. (2004). Deep

brain stimulation in movement disorders: The applications reconsidered. Acta

Neurologica Belgica, 104(1), 32-36.

Vorberg, D., & Wing, A. M. (1996). Modeling Variability and Dependence in Timing. In S.

Keele & H. Heuer (Eds.), Handbook of Perception and Action (Vol. 2). New York:

Academic Press.

Walker, J. T., & Scott, K. J. (1981). Auditory-visual conflicts in the perceived duration of lights,

tones, and gaps. Journal of Experimental Psychology: Human Perception and

Performance, 7(6), 1327-1339.

Wearden, J. H. (1999). "Beyond the fields we know ...": exploring and developing scalar timing

theory. Behavioural Processes, 45(1-3), 3-21.

180

Wearden, J. H. (2003). Applying the scalar timing model to human time psychology: Progress

and challenges. In H. Helfrich (Ed.), Time and Mind II: Information Processing

Perspectives (pp. 21-39). Cambridge, MA: Hogrefe & Huber Publishers.

Wearden, J. H. (Ed.). (2005). The wrong tree: Time perception and time experience in the

elderly. Oxford: Oxford University Press.

Wearden, J. H., & Culpin, V. (1998). Exploring scalar timing theory with human subjects. In V.

DeKeyser, G. dYdewalle & A. Vandierendonck (Eds.), Time and the Dynamic Control of

Behavior (pp. 33-49).

Wearden, J. H., Edwards, H., Fakhri, M., & Percival, A. (1998). Why "sounds are judged longer

than lights": Application of a model of the internal clock in humans. Quarterly Journal of

Experimental Psychology Section B-Comparative and Physiological Psychology, 51(2),

97-120.

Wearden, J. H., & Lejeune, H. (2008). Scalar properties in human timing: Conformity and

violations. Quarterly Journal of Experimental Psychology, 61(4), 569-587. doi:

10.1080/17470210701282576

Wearden, J. H., & McShane, B. (1988). Interval production as an analog of the peak procedure:

Evidence for similarity of human and animal timing processes. Quarterly Journal of

Experimental Psychology Section B-Comparative and Physiological Psychology, 40(4),

363-375.

Wearden, J. H., Smith-Spark, J. H., Cousins, R., Edelstyn, N. M. J., Cody, F. W. J., & O'Boyle,

D. J. (2008). Stimulus timing by people with Parkinson's disease. Brain and Cognition,

67(3), 264-279. doi: 10.1016/j.bandc.2008.01.010

181

Wearden, J. H., Wearden, A. J., & Rabbitt, P. M. A. (1997a). Age and IQ effects on stimulus and

response timing. Journal of Experimental Psychology: Human Perception and

Performance, 23(4), 962-979.

Wearden, J. H., Wearden, A. J., & Rabbitt, P. M. A. (1997b). Age and IQ effects on stimulus and

response timing. Journal of Experimental Psychology: Human Perception &

Performance, 23(4), 962-979.

Wing, A. M. (1980). The long and short of timing in response sequences. In G. E. Stelmach & J.

Requin (Eds.), Tutorials in motor behavior (pp. 469-486). Amsterdam: North-Holland.

Wing, A. M., & Kristofferson, A. B. (1973). Response delays and the timing of discrete motor

responses. Perception and Psychophysics, Vol. 14(1), 5-12.

Woodrow, H. (1930). The reproduction of temporal intervals. Journal of Experimental

Psychology 13(6), 473-499.

Woodrow, H. (1951). Time perception. In S. S. Stevens (Ed.), Handbook of Experimental

Psychology (pp. 1224-1236). New York: Wiley.

Yahalom, G., Simon, E. S., Thorne, R., Peretz, C., & Giladi, N. (2004). Hand rhythmic tapping

and timing in Parkinson's disease. 10(3), 143-148.

Zakay, D. (1993a). The impact of time perception processes on decision making under time

stress. In O. Svenson & A. J. Maule (Eds.), Time Pressure and Stress in Human

Judgment and Decision Making. New York: Plenum Press.

Zakay, D. (1993b). Time estimation methods: Do they influence prospective duration estimates?

Perception, 22(1), 91-101.

Zakay, D., & Block, R. A. (1997). Temporal cognition. Current Directions in Psychological

Science, 6(1), 12-16.

182

Zakay, D., & Block, R. A. (2004). Prospective and retrospective duration judgments: an

executive-control perspective. Acta Neurobiologiae Experimentalis, 64(3), 319-328.

183

APPENDIX B: FOOTNOTES

1The discovery of the neurotoxin 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) has been a significant advancement toward understanding PD. Administration of MPTP results in selective DA cell loss in the substantia nigra and creates disordered movement resembling the motor symptoms of PD. Interestingly, the effects of this neurotoxin were discovered by accident, when a group of synthetic-heroin users reported to several California emergency rooms with PD- like symptoms that developed over a period of days (Langston, 2008). All cases were traceable to one batch of heroin that, mistakenly, contained MPTP. Currently, MPTP is used to create animal models of PD, which have informed current understanding of the pathology of PD, and have served as a useful tool for testing a variety of pharmacological treatments.

2Several techniques have been used to estimate cell loss in the BG. Two techniques are postmortem studies and single photon emission computed tomography (SPECT). Postmortem studies have collected BG cell counts from a large sample of controls and individuals exhibiting a range of PD symptomology. A regression equation, controlling for DA cell loss due to general aging, is used to estimate cell loss prior to symptom presentation (Fearnley & Lees, 1991).

SPECT is a neuroimaging technique also used to estimate DA cell loss within the BG due to PD.

The tracer, [123I]β-CIT, binds to DA transporters within the BG after injection into an individual.

The concentration of radioactivity from the tracer binding in the BG is used to estimate neurodegeneration of DA cells (Tissingh, et al., 1998).

3Interestingly, anticholinergics may have been the first generally accepted pharmacological treatment for Parkinsonism. Their use dates back to ancient India, when they were used to treat a neurological disorder believed to be Parkinsonism (Manyam & Sanchez-

Ramos, 1999).

184

4A collection of oscillators is likely to exist. But, for the sake of brevity, only one oscillator is discussed.

5Interested readers are directed to Block, Zakay, and Hancock (1998) for a brief review of some early psychological ideas regarding theories of time experience in older adulthood.

Interestingly, age-related hypotheses about time experience date back as far as 1877.

6 Readers interested in theoretical predictions of timing in young adults from an interval approach are directed to Droit-Volet and Wearden (2001), Lustig (2003), and Wearden (2005).

Theoretical predictions from an entrainment perspective are provided by Drake, Jones, & Baruch

(2000), Jones (1976), and McAuley et al. (2006).

7Interpretation of the results of the W & K model and slope analysis can be problematic when the models are applied to groups other than young adults. Application of these models to children, older adults, and individuals with neurological impairments is often reported to result in negative motor variance. Aside from the implausibility of negative motor variance, these results are difficult to interpret. Corrections for these violations have been suggested, such as setting motor variability equal to zero, but such methods require the assumption that all variability is due to the clock (Ivry & Keele, 1989). Therefore, one must carefully interpret the estimates of these models when applied to groups other than young adults. While an more in depth discussion of these issues and violations is beyond the scope of this dissertation, interested readers are directed to McAuley et al. (2006) for a discussion of these concerns.

8Pastor et al. (1992) report the results of a task that they consider to be a time estimation task. Participants were trained to accurately reproduce 1 second. Following training, they were asked to estimate durations of 3-, 9-, and 27-seconds while internally counting at a rate of 1 second. They found that individuals with PD were more variable, overall, compared to older

185 adult controls, but the largest amount of variability was found for the 27-second duration.

Furthermore, reductions in variability were observed for all durations when the PD group was on medication. Based on the counting requirement of this task, it appears to be more of a production task, as opposed to an estimation task. Interpretation of the results of this study as a production task is complicated by the data reported in the article. Participants are reported to underestimate the duration, but no data regarding group estimations/productions is provided. Furthermore, all error scores are reported in terms of absolute error, rendering assessment of the direction of errors impossible.

9A few studies have examined both the perception of isolated time intervals and sequences of time intervals in individuals with PD using durations shorter than those discussed in this dissertation (e.g., < 100-ms). Two studies have used tasks similar to the isolated-interval duration discrimination task described in Chapter 3, however very short IOIs (< 100-ms) were investigated (Guehl, et al., 2008; Rammsayer & Classen, 1997). Due to the time scale difference, these studies were not included in the review. Other studies have used a gap discrimination task, which has participants judge whether a brief time gap is present in a tone (Artieda, Pastor,

Lacruz, & Obeso, 1992; Guehl, et al., 2008). Although papers frequently cite Artieda, Pastor, and

Lacruz (1992) as evidence that individuals with PD show improvements in duration discrimination while on medication, there has been some question as to whether the gap discrimination task they used is truly only a measure of time discrimination, or also tests the ability to perceive fast intensity fluctuations created by the gap that separates the tones (Guehl, et al., 2008). Hence, this study was not included in this review.

10Overall, individuals with PD performed well on the DRS-2. One score approached the exclusion criteria. However, this score reflects speech problems associated with PD, as opposed

186 to a cognitive impairment. The low score is primarily driven by the participant having difficulty with the spoken portions of the DRS-2. Additionally, the participant had some difficulty with abstract thought. Over the remainder of the testing sessions, the participant showed no signs of a cognitive impairment. On the other hand, one individual fell below the exclusion criteria. In addition to a DRS-2 score of 117, this participant also had difficulty comprehending the CES-D and MEQ.

187

APPENDIX C: DESCRIPTION OF SCREENING MEASURES

Dementia Rating Scale (DRS-2)

The original DRS (Mattis, 1998) was developed to objectively measure overall cognitive functioning in individuals with impaired cognitive abilities. Because the scale is sensitive to lower levels of cognitive ability, it is commonly used in neuropsychological assessment (Jurica,

Leitten, & Mattis, 2001). The DRS-2 remains relatively unchanged from the original version

(Jurica, et al., 2001). The current version still separates 36 tasks into five separate subscales that reflect specific cognitive abilities: attention (8 items), initiation/perseveration (11 items), construction (6 items), conceptualization (6 items) and memory (5 items). Individual scores on the separate subscales are summed for a raw score. The maximum DRS-2 total score is a 144.

Often, studies use a cut-off score of 123, as scores below this value are taken to suggest more than a mild cognitive impairment.

Morningness-Eveningness Questionnaire (MEQ)

The MEQ is a 19-question scale used to identify an individual‘s peak time in the circadian rhythm (Horne & Ostberg, 1976). Questions range from when one would prefer to perform intense physical labor to waking habits. Individuals select their preferred time to perform the listed activity, or select their awakeness/tiredness during the time of day described in the question. Each response is given a numerical score. The total sum of scores is used to categorize an individual‘s Morningness-Eveningness type on the following scale:

70-86 Definitely Morning Type 59-69 Moderately Morning Type 42-58 Neither Type 31-41 Moderately Evening Type 16-30 Definitely Evening Type

188

Center for Epidemiologic Studies Depression scale (CES-D)

The CES-D is a 20-question self-report depression screening (Radloff, 1977). The scale reflects the following dimensions of depression: depressed mood, feelings of guilt and worthlessness, feelings of helplessness and hopelessness, psychomotor retardation, loss of appetite, and sleep disturbance (Radloff, 1977). The questionnaire contains a list of feelings or behaviors. Individuals rate whether they have ―felt or behaved this way in the past week‖ on a scale of 0 (rarely or none of the time) – 3 (most or all of the time). Possible scores range from 0 –

60. Typically, scores above 16 are taken to suggest high depressive symptoms (Radloff, 1977).

Unified Parkinson Disease Rating Scale (UPDRS)

The UPDRS scale is used to track the progression of PD (Fahn & Elton, 1987). The scale assesses individuals with PD on three separate categories: (1) Mentation, Behavior, and Mood,

(2) Activities of Daily Living, and (3) Motor Examination. Numerical scores are given for each of the three categories.

Only the Motor Examination subscale was assessed for this dissertation. Twenty-seven motor symptoms of PD were rated on a 5-point scale (0 = Normal/Absent; 4 =

Marked/Severe/Unable to Perform). Anchors for each scale varied, based on the specific motor symptom being assessed. Possible scores on the UPDRS-Motor range from 0 (no motor symptoms) – 108 (severe motor symptoms).

Self-Reports of PD Symptom Severity and Medication Effectiveness

Individuals with PD were asked to self-report both their PD-related symptoms and medication effectiveness throughout Epochs 2 and 3. This measure was collected to validate the

UPDRS scoring and to assess whether participants may have experienced any medication fluctuation (i.e., OFF state) during testing. Parkinson symptoms were rated on a seven-point

189 scale with the following anchors: 1 = extreme; 4 = moderate; 7 = none. Medication effectiveness was rated on a seven-point scale with the following anchors: 1 = not at all; 4 = moderate; 7 = perfectly.