Time Perception Distortion in Neuropsychiatric and Neurological Disorders

Send Orders for Reprints to reprints@benthamscience.net CNS & Neurological Disorders - Drug Targets, 2013, 12, 567-582 567 Time Perception Distortion in Neuropsychiatric and Neurological Disorders Silmar Teixeira 1,17,18, Sergio Machado 2,3,4,5, Flávia Paes 2,3, Bruna Velasques 1,6, Julio Guilherme Silva 7,8, Antonio L. Sanfim 2,3, Daniel Minc 1, Renato Anghinah 9, Luciano L. Menegaldo 10, Mohamed Salama 16, Mauricio Cagy 1, Antonio E. Nardi 2,3, Ernst Pöppel 11,12,13, Yan Bao 11,12,13, Elzbieta Szelag 12,14, Pedro Ribeiro 1,2,15 and Oscar Arias-Carrión *,19 1 Brain Mapping and Sensory Motor Integration, Institute of Psychiatry of Federal University of Rio de Janeiro (IPUB/UFRJ), Brazil 2 Laboratory of Panic and Respiration, Institute of Psychiatry of Federal University of Rio de Janeiro (IPUB/UFRJ), Brazil. 3 National Institute of Translational Medicine (INCT-TM), Brazil 4 Quiropraxia Program of Faculty of Health Sciences, Central University, Santiago, Chile 5 Institute of Philosophy, Federal University of Uberlândia, Minas Gerais, Brazil 6 Institute of Applied Neuroscience (INA), Rio de Janeiro, Brazil 7 Department of Physical Therapy, Federal University of Rio de Janeiro (IPUB/UFRJ), Brazil 8 Master of Science Rehabilitation, UNISUAM, Rio de Janeiro, Brazil 9 Department of Behavioral and Cognitive Disorders, Faculty of Medicine, University of Sao Paulo, Brazil 10 Biomedical Engineering Program, Institute for Graduate Studies and Research in Engineering, Federal University of Rio de Janeiro (COPPE/UFRJ), Brazil 11 Department of Psychology and Key Laboratory of Machine Perception (MoE), Peking University, China 12 Human Science Center, Ludwig-Maximilians-Universität (LMU), Germany 13 Institute of Medical Psychology, Ludwig-Maximilians-Universität (LMU), Germany 14 Laboratory of Neuropsychology, Nencki Institute of Experimental Biology, Warsaw, Poland 15 School of Physical Education, Bioscience Department (EEFD/UFRJ), Brazil 16 Toxicology Department, Faculty of Medicine, Mansoura University, Mansoura, Egypt 17 Department of Physical Therapy, Veiga de Almeida University (UVA), Rio de Janeiro, Brazil 18 Department of Physical Therapy, Piquet Carneiro Policlinic, State University of Rio de Janeiro, (PCC/UERJ, Brazil 19 Movement Disorders and Transcranial Magnetic Stimulation Unit, Hospital General Dr. Manuel Gea González, Secretaría de Salud, México D.F., México Abstract: There is no sense organ specifically dedicated to time perception, as there is for other senses such as hearing and vision. However, this subjective sense of time is fundamental to our conception of reality and it creates the temporal course of events in our lives. Here, we explored neurobiological relations from the clinical perspective, examining timing ability in patients with different neurological and psychiatric conditions (e.g. Parkinson s disease, depression, bipolar disorder, anxiety disorders and schizophrenia). The neural bases of present distortions in time perception and temporal information processing still remain poorly understood. We reviewed: a) how the brain is capable of encoding time in different environments and multiple tasks, b) different models of interval timing, c) brain structures and neurotransmitters associated with time perception, d) the relationship between memory and time perception, e) neural mechanisms underlying different theories in neural and mental processes, and f) the relationship between different mental diseases and time perception. Bibliographic research was conducted based on publications over the past thirteen years written in English in the databases Scielo, Pubmed/MEDLINE, ISI Web of Knowledge. The time perceptions research are executed to evaluate time perception in mental diseases and can provide evidence for future clinical applications. Keywords: Anxiety, mood, schizophrenia, Parkinson s disease, time perception, timing, interval-timing. *Address correspondence to this author at the Unidad de Investigación de Desordenes del Movimiento y Estimulación Magnética Transcraneal, Hospital General Dr. Manuel Gea González. Calzada de Tlalpan 4 800, Col Sec. XVI, Delegación Tlalpan. Código postal: 14080. México D.F. México; Tel/Fax: +52 55-85438283; E-mail: arias@ciencias.unam.mx INTRODUCTION For centuries, time has always been considered an intriguing problem for a better understanding of the human mind. Time inspired artists, philosophers and researchers, perhaps, for the fact that everyone is limited by time. Thus, 1871-5273/13 $58.00+.00 2013 Bentham Science Publishers

568 CNS & Neurological Disorders - Drug Targets, 2013, Vol. 12, No. 5 Teixeira et al. time imposes on us a limitation in many ways; in particular, since the day we are born, everyone s clock begins to run. Understanding the impact of time in our lives seems to be extremely complex, especially knowing how the central nervous system (CNS) perceives time [1]. Sometimes we perceive events that take forever, on other occasions time seems to pass very fast. If you are with friends watching the final game of a Soccer World Cup with 5 minutes to the end and your team is losing by one goal difference, these will be the longest and worst 5 minutes of your life. On the other hand, if your team is winning by 3 goals, time doesn t make a difference. In a psychological view, the way we perceive time is relative to the importance of that particular moment for each of us. Thus, subjective time is dependent on the task one is engaged in during a certain time interval [2]. Moreover, time perception is affected by impressions, attention, the impact of the fact on people and by mood; in particular, time is affected by our memories of past events [1, 3]. Time participates in every aspect of our life, such as: driving a car, stepping on the sidewalk, listening to music, dancing, or even eating peanuts at a bar. Within this context, time perception is an expression that comprises many scales ranging from milliseconds, minutes, hours, days, weeks or years. For instance, circadian rhythms (i.e., 24-hours) are responsible for controlling biochemical, physiological or behavioral processes of living organisms which persist in the absence of cues, such as the control of sleep and wakefulness, cortisol levels, growth hormones, blood pressure and several other factors. Another example is interval timing, implicated in seconds-to-minutes range, responsible for decision making, arithmetic operations and conscious time estimation, and millisecond timing, crucial for control of motor behavior or speech patterns [4]. Perhaps all the terms above are vague concepts to talk about the deep impact of time on our lives. Whether we have to choose one concept which may better explain the meaning of words such as attention, perception or mood on time perception, timing would possibly be a reasonable one. Or at least, timing as a measure would give us insights about how the brain detects and interprets reality, i.e., scanning the environment. Timing is also related to the detection of a number of events per period of time. In other words, it corresponds to the frequency for an acquiring piece of information [5]. Timing is a fundamental component of information processing in the CNS, since it tells us how we perceive the environment, and especially how we interpret events around us. Of course, we have several other components involved in the interpretation of reality, but timing is a key element. Commonly, the term timing has been used to refer not only to the estimation of an event s duration, but also about prediction of an event onset or offset, or even whether an event occurred before or after some temporal landmark, i.e., temporal order judgment. Both types of prediction require a metrical representation of time; meanwhile timing, i.e., duration or onset, of a single event can be measured on a continuous, parametric timescale [6]. With this in mind, the present paper reviews the main aspects considered strategic in the understanding of how the CNS is able to encode time in different environments and multiple tasks. Here, we will discuss the different models of time perception to describe the phenomenon, the CNS structures and neurotransmitters associated with time perception, the relationship memory, time perception and current experimental mechanisms involved in neuropsychiatric disorders (i.e., anxiety disorders, mood, Parkinson s disease (PD), schizophrenia, or neuropsychological phenomena). In the first part, we show that temporal processing primarily about the duration, interval and perception estimation, and in the second part about sensory, perceptual and cognitive processes. MODELS OF TIME PERCEPTION The issue of time perception models can be seen by several points of view, depending on the focus of investigation. Several models have been proposed to describe time perception. According to Church and Broadbent [7] a general model of information processing contains i) a pacemaker or oscillator (send pulses), ii) an accumulator which count the pulses (working memory) and one comparator which compares the pulse number with the reference memory. Thus, the number of pulses accumulated in the working or reference memory is the number of pulses that were accumulated at the time of the reinforcement (activity). So, pulses are multiplied by one constant of memory i.e. the working or reference memory will store the number of pulses recorded in the accumulator which is multiplied by one constant of memory, and this value will be compared with the current values of the accumulator to record a decision. Moreover, Ivry and Schlerf [8] showed in a review study that the models of time perception are modular; they also showed that the pacemaker-counter is a modular model defined by accumulation of inputs from a pacemaker. Another example of the modular process is the spectral model of timing, where the interaction among oscillators or the exploitation of different patterns forms a set delay line defining different intervals. These representations are seen as specializations, characteristic of particular neural structures. According to Mauk & Buonomano [9], we may summarize them in three different approaches: i) neurobiological internal clocks; ii) spectral timing; and, iii) state dependent. Other forms focus on cognitive and neural aspects, contrasting them in terms of conceptualization and pointing evidence of heterogeneous mechanisms of temporal perception processes. Besides these aspects, there is evidence of linear and nonlinear models, as also of models capable to identify errors, apprehend and modify strategies [10, 11]. In particular, the temporal differentiation of neural processes has been referenced as needed in various models of time decoders. Studies show that in the neurobiological internal clock models, neural cells respond to stimuli with different answers timed, in other words, the stimuli activate neural cells that oscillate at different frequencies and phases [12]. This fact reflects that time points, after a stimulus, can be encoded by activity in a different subset of neurons through another subset activated in another time [12]. A second model hypothesizes that the neural connection promotes a sequential activation at different times in a stimulus chasing [13]. In addition, spectral models are described as elements that share a common application, that is, each neural unit responds selectively to different intervals of time, and is able to encode the time at the beginning of the stimulus by

Time Perception Distortion in Neuropsychiatric and Neurological Disorders CNS & Neurological Disorders - Drug Targets, 2013, Vol. 12, No. 5 569 different subsets of neural cells that are activated at different times [7]. In order to clarify these models, neuroimaging studies show a difficulty to point out which region functioned primarily as a timekeeping area. After different experiments in which several regions appeared activated during time perception tasks, the final conclusion is that disruptions in any components can result time in impairments, led a new model which assumes that instead of a specific region the timing mechanism is distributed in a neural network, within each region encoding duration [9]. Therefore, the state dependent model alters the concept of internal-clock centralized displacement to the capacity and the intrinsic property in temporal processing as a result of the neural circuits complexity [6, 13], not needing specialized areas for the temporal processing [14]. The stimulus would activate cell matrixes that would oscillate in distinct frequencies and phases, in order for the time event codification to occur in different neuron subclasses at different times [15-17]. One of the characteristics of the proposed model is that the implicit time codification is a function of the neural firing and of the circuit's intrinsic properties such as: membrane's potential, synaptic forces and stocked neurotransmitters. Moreover, this model is capable of discriminating temporal interval patterns in a millisecond scale and is also capable of showing that in this band, time does not behave according to a clock, with linear and sequential metric [9]. Finally, the models with oscillator have been the more used, the difference between theories is related to number and characteristics of oscillators which some authors have proposed that there one oscillator, while others, that there multiple oscillators [18]. NEUROANATOMY OF TIME PERCEPTION The issue of time perception has examined whether the CNS owns a unique internal chronometer or whether different specialized structures are part of this process, i.e., a multiple chronometer. In other words, studies aimed to verify whether neural pulses emitted at regular intervals are stored in one or more time accumulators [1-3]. These studies investigate how the CNS calculates information, arguing that the discontinuous stimulus (task with visual interrupted signal) for bottom-up system is responsible for synchronizing the milliseconds, and the continuous stimulus (task with visual continuous signal) for top-down system is responsible for synchronizing the seconds and minutes [2, 13]. In this section, we aim to discuss whether CNS structures involved in interval timing operate integrated rather than individually. The manipulation of temporal events is controlled by an attention switch, which would be responsible for the synchronization processing when the signal is scheduled to be started or suspended [19]. In this investigation model, it is assumed that the interval timing is involved in a thalamocortico-striatal circuit [20], composed of basal ganglia, prefrontal cortex and posterior parietal cortex [6], anterior cingulate gyrus [21] and suprachiasmatic nucleus (SCN) [22]. In particular, the basal ganglia have neurobiological properties, such as striatal modulation by dopamine, which are necessary for the interval timing. More specifically, the striatum has been referred to as a detector of cortical and thalamic inputs [20]. Within this context, PD patients demonstrated greater variability of time during task execution and difficulty in estimating interval timing. In other words, PD patients underestimate the time, suggesting a reduction in the internal clock speed [6, 23]. Basal ganglia would act as a counter-clock system, including the operation of dopaminergic neurons in the substantia nigra of the midbrain as a pacemaker unit [23]. The cerebellum seems to be closely related to the internal clock activity [23, 24]. In this regard, it was hypothesized that the cerebellum could be involved in the impairment of the capacity to respond appropriately to temporal discrimination simple tasks such as repetitive movements in short-duration scales. Some studies were conducted using mice with cerebellar injury, trained in temporal discrimination tasks, showing no impairment in the shortduration task performance when compared to controls [6, 11]. Thus, the cerebellum seems to operate as a millisecond controller [11] avoiding the loss of temporal precision and perception in patients with cerebellar injury [24]. Other indirectly time-related functions are also attributed to the cerebellum, such as motor learning, adaptation [25], controlling synergies [26, 27], as well as memory acquisition, storage and retrieval [28]. Welsh and Llinas attribute a role to the cerebellum in the timing of collective muscle group contractions [29]. Specifically, the cerebellum-related temporal models aim to understand how the Purkinje's cells learn to cease and to control its firings in the presence of an unconditioned stimulus, releasing its discharge from deep cells. The first delay line-model states that pontine cells are sequentially linked not only to each other [30]; in fact, they are also linked to the granule-cells, leading to a delay line. In the presence of a conditioned stimulus, pontine cells are sequentially activated, but provoked only by granule-cell firing, one by one. The second delay line-model proposes a large lag in the conduction of the parallel fiber stimuli when compared to mossy fibers [31]. The first and second delay line models are just theoretical models with no anatomical evidence regarding pontine cell connections, and do not provide much in the way of explanations regarding cerebellar timing. The spectral timing model assumed that granule-cells have constant large range distributions in time into membrane, leading to several active delays. Each granule-cell would be related to a Golgi's cell, therefore inhibiting the activation of such Golgi cells. Due to this inhibition, the granule-cells become transiently active, with several delays since the start of the conditioned stimulus appearance, resulting in peaks of spectral activity during the conditioned-unconditioned interval [32]. Besides basal ganglia and cerebellum, the supplementary motor area (SMA) has been seen as the main activated region in visual and auditory temporal discrimination tasks. Specifically, different magnitudes of time interval have been observed pre-sma, which is responsible for inferring time interval, receiving somatosensory stimuli and activating temporal attributes for information processing [21]. For instance, certain behavioral states can induce several synchronized and non-synchronized peaks of neural

570 CNS & Neurological Disorders - Drug Targets, 2013, Vol. 12, No. 5 Teixeira et al. oscillations combinations in several cortical regions happening within a specific time interval. These activity patterns are responsible for the time signals so that SMA functions are associated with other regions [20, 29]. In addition, experiments show that the brain clock, responsible for generating circadian rhythms, is located on the SCN. Its oscillations are provided by retinal ganglion cells (i.e., retinal thalamic tract) [33], which translate the visual feedback loops and synchronized environmental cycles (day/night). This process occurs through liberation of neurotransmitters in SCN receptors, promoting physiological changes in time cycle, such as the development of longduration potential [22, 34]. Some authors have proposed mathematical models based on ordinary differential equation able to simulate numerically the SCN behavior [34, 35]. NEUROPHARMACOLOGY AND TIME PERCEPTION Time perception is involved in many aspects of our lives. The knowledge of the brain structures involved in each scale (i.e., milliseconds, minutes, hours, days, weeks or years) is extremely important to understand the dynamics that regulate our perception. However, we know that some neurotransmitters participate in modulating the operation of these structures. The main neurotransmitters underlying time perception are dopamine (DA) and acetylcholine. These neurotransmitters are associated with different mechanisms. Other neurotransmitters as serotonin (5-HT), norepinephrine, glutamate and -aminobutyric acid are also relevant and influenced by changes in DA and acetylcholine. An important aspect in time perception is the ability to discriminate durations of time, or interval time. Some researchers have hypothesized that an optimal level of dopaminergic activity in corticostriatal circuits is modulated by 5-HT and glutamate activity [2, 36, 37]. DA also appears to modulate the effectiveness of cortical and thalamic glutamatergic excitation and may also modulate the threshold for coherent activity detection [20, 38]. These relationships among the neurotransmitters demonstrate that they work together in a system and that one interferes with the other. Although some investigations demonstrated the engagement of all the neurotransmitters cited above, the most consistent results are the related to DA. In this section, we will review these neurotransmitters and their involvement in the time perception. Most of the studies involving time perception and neurotransmitters are conducted by the administration of drugs that affect their systems. Amphetamine, methamphetamine, cocaine and haloperidol have been used to study the role of the dopaminergic system in time perception. The first two substances are dopaminergic agonists, and they have shown a speed up in the clock of rats and pigeons. The last one is a dopaminergic asntagonist and slows down the clock [19, 39, 40]. Methamphetamine is a dopaminergic agonist and a potent psychostimulant that promotes euphoria, increases mental alertness and motor activity through the increased release of DA and the modulation of norepinephrine and 5-HT [41, 42]. Methamphetamine has been used to investigate the involvement of DA on time perception. In humans and other mammals, low to moderate doses alter time perception, increasing the speed of an internal clock [6, 43, 44]. It is well-established that DA antagonists produce a slowing of the clock and that DA agonists speed up the clock [2, 23]. However, it has been seen that these drugs also influence other neurotransmitter systems, such as serotoninergic, noradrenergic and cholinergic. Other results demonstrate that the effects of drugs are related to memory and attentional effects, instead of being related to changes in the internal clock speed. For example, some studies suggest that DArelated drugs affect the level of attention to the temporal signals, and that it is not directly related to changes in the internal clock speed [17, 45]. Pharmacological administrations that alter cholinergic systems produce an effect on the storage of temporal memories, and are not directly associated with changes in the internal clock speed [1, 46]. Lesions in the basal forebrain and in the frontal cortex are a gradual shift to the right at peak time. This fact suggests modifications in the storage of temporal memories. These areas are strongly projected with cholinergic nucleus, such as the nucleus basalis magnocellularis [14]. However, lesions in another cholinergic nucleus, for example the hippocampus showed opposite directions when compared to frontal cortex lesions [46]. Although this evidence demonstrates that cholinergic projections influence the storage of temporal memories, it is uncertain how cholinergic drugs alter memory storage speed [1, 34]. Evidence demonstrates that these lesions do not impair the temporal control, but they perform an important modulatory function [14]. The administration of cholinergic-related drugs that increase the efficacy of acetylcholine shift the timing functions leftward across sessions relative to baseline, while drugs that decrease the efficacy of acetylcholine shift the timing functions rightward as training proceeds relative to baseline. If the drug administration is discontinued, then the timing functions gradually return to those which were observed during baseline, so that there is no rebound effect in this instance [47]. It is difficult to comprehend exactly what is the involvement of each neurotransmitter, because of the influence of each one on the others. DA participates in three major circuitries: nigrostriatal, which controls motor activity; mesolimbic, which regulates reinforcement systems; and mesocortical, which coordinates goal-oriented reactions [48, 49]. Directly and indirectly these circuitries participate in attentional control [50]. The involvement of DA on attention leads to an important question: whether DA influences attention or exclusively timing control. In this case, dopaminergic activation may provide a reinforcement and prediction signal for associative learning. Evidence suggests that the nigrostriatal dopaminergic circuitry communicates with the mesolimbic to modulate the temporal perception [51-54]. Previous studies demonstrated that the expected temporal characteristics of the environment are being encoded by the activity of DA neurons in the substantia nigra pars compacta. Specially, these circuitries work on the detection of time related to reward expectation. These findings suggest that dopaminergic projections modulate the production and reception of temporally informative input, providing additional support for interval timing behavior. Specifically, the striatal neurons function as a decision stage comparator of an interval timer. If the reward expected times

Time Perception Distortion in Neuropsychiatric and Neurological Disorders CNS & Neurological Disorders - Drug Targets, 2013, Vol. 12, No. 5 571 arises, the efflux of DA decreases, depressing the ability of striatal firing [6, 50]. Reinforcement contributes for the timing function s return to the baseline. For example, Meck demonstrated the effects produced after withdrawing methamphetamine from rats [36]. A timing function rebound was observed after methamphetamine withdrawal, this phenomenom being explained because of slowdown of the internal clock when compared to the drug condition. On the other hand, the remembered times originated from a reinforcement remained in line with the internal clock s previously faster time. These facts demonstrated that the internal clock maintains the criterion of duration at a later time relative to baseline conditions until normalization/relearning occurs again. The rebound effect is interpreted as a result of the relearning during training with the drug, and of an attention effect [16, 47, 55]. Buhusi and Meck demonstrated that the stop/reset mechanism is affected by acute administration of dopaminergic drugs: methamphetamine (dopaminergic agonist) tends to increase the shift in peak time toward resetting timing and haloperidol (dopaminergic antagonist) tends to decrease the shift in peak time toward stopping timing [19]. This evidence suggests that clock and attentional effects might be related to the affinity of dopaminergic drugs for the D1 and D2 DA receptors [35, 55]. Thus, drugs that activate both receptors are expected to determine both effects, whereas drugs that selectively affect the D2 DA receptor are expected to determine only the attentional effect [56]. Buhusi and Meck found an effect in clock and attention produced by methamphetamine ingestion, whereas changes produced by haloperidol ingestion indicated a predominantly attentional effect rather than a clock-speed effect [6]. They conclude that these results might have been due to the activation of the noradrenergic or serotonergic systems produced by the drugs [57, 58]. Although temporal processing time over one second is presumably mediated cognitively, the short duration process (less than one second) seems mediated by non-cognitive mechanisms [47, 58]. Rammsayer showed that temporal processing of intervals in the range of milliseconds depends on the effective level of dopaminergic activity in the basal ganglia [59]. The author showed that in humans temporal processing of long durations was significantly impaired by 3 mg of haloperidol and 11 mg of benzodiazepine, whereas processing of extremely brief intervals was only affected by haloperidol. These findings show that temporal processing of intervals in the range of milliseconds appears to depend on the dopaminergic activity in the basal ganglia. A separate line of evidence suggests that the dopaminergic system influences the temporal resolution by affecting the track of significant events [12, 30, 34]. Specifically, DA seems to play a role in modulating the threshold for coherent activity detection, affecting the attentional system. Despite the amount of information about the modulation of internal clock speed by DA, the influence on temporal dynamics remains unclear. The participation of other neurotransmitters has been deeply investigated, particularly to better understand which projections are involved specifically in clock speed and time perception. SHORT-TERM MEMORY AND TIME PERCEPTION Although humans have the ability to perceive, estimate and reproduce time [6], the role that memory plays in the temporal dynamics is still not too clear. There are two models that relate memory to time perception. The first is the Scalar Expectancy Theory that postulates memory as one of three stages of the information-processing model, i.e., clock, memory and decision [60]. The memory stage receives temporal information obtained by the clock stage and stores it in the working memory. Therefore, the stored information will be compared to previous data in order to make a choice in the decision stage [20]. The second model is the Multiple Time Scale proposing that animals may learn to estimate time in interval-timing tasks through the strengthening of the memory trace [61]. It will be produced by events such as food reinforcement or intertrial interval. This can be explained by habituation process and its rate of sensitivity [62, 63]. In brief, Scalar Expectancy Theory and Multiple Time Scale suggest distinct roles for memory: the first suggests memory as a comparator component; and the second suggests memory as a clock [61]. Moreover, animal models of brain lesions provide important findings about the functional role of frontal cortex (FC) and hippocampus in temporal memory. Rats with lesions in the FC and nucleus basalis magnocellularis overestimated the expected time of reinforcement observed by an increase in the peak time during the trials [46, 64]. On the other hand, lesions in the medial septal area and fimbriafornix of hippocampus led to changes in task performance, such as resetting during gap trials, failure of the time criterion regulation during successive trials and reduction in peak time during peak-trials [3, 40]. Taken together, the results show two different links between memory and time perception. The first composed of the FC and nucleus basalis magnocellularis, related to changes in the content of the current memory that stores information about the current time. The second consists of medial septal area and fimbriafornix that play a role in working memory, comparing its content to the previous data, due to the failure of time criterion regulation and the resetting behavior [65]. In addition, neuropharmacological studies using anticholinesterase drugs (physostigmine and neostigmine) and cholinergic receptor blockers (atropine and methylatropine) elucidated the role of the neurotransmitter acetylcholine for temporal memory. The administration of atropine in rats during a 20-second peakinterval procedure led to an overestimation and an increased variability of remembered times of reinforcement, while physostigmine had the opposite effect (underestimation and decreased variability) [64]. Other experiments demonstrated that prenatal choline supplementation increases the precision of timing and temporal memory and facilitates simultaneous temporal processing in mature and aging rats [66, 67]. These findings suggest that long-term changes in cholinergic mechanisms are responsible for these changes by improving the quality of clock control and memory processes [68]. TIME PERCEPTION IN NEURONAL AND MENTAL PROCESSES Time perception is a subjective experience of how fast time passes or how much time passed after the occurrence of

572 CNS & Neurological Disorders - Drug Targets, 2013, Vol. 12, No. 5 Teixeira et al. a certain event. The ability of an individual to estimate both objective and physical time has been dealt with a stable function, varying in certain brain pathologies, psychiatric disorders or toxicological situations [1, 69]. The way a person estimates time is subjective and involves the participation of an internal clock to measure an objective time with no external cues. So, we showed different mental diseases and its relations with time perception (i.e. anxiety and mood disorders, schizophrenia and PD). Anxiety Disorders Several studies have investigated abnormal time experience, showing different findings, e.g., slow and fast time estimation, and also changes in time estimation in depressive patients [70-73]. For instance, Elsass et al. [74] observed that the time estimation varied between bipolar patients treated with lithium and controls. Thus, studies comparing its symptoms to the symptoms of major depressive disorder indicate that more attention should be paid to differentiating between these two conditions. In the clinic, anxiety and depression occur together and have similarities, indicating that further research is necessary to develop differential criteria for these conditions [75]. Within this context, anxiety involves as much physiological as psychological symptoms, with a neural circuit where thalamus is the main manager of distribution of environmental stimuli. The latter are acquired by sensory receptors and transmitted to the thalamus, which consequently sends them to the amygdala and sensory cortical areas. However, questions related to the timing of amygdala response to novel relative to neutral faces and controversies as to whether the amygdala response to emotional stimuli is automatic or depends on attentional load remains unclear. On this basis, Blackford et al. [76] investigated patients with social anxiety disorder (SAD) in order to verify the timing of amygdala response characterized by latency, duration and peak using event-related functional magnetic resonance imaging when patients compared novel to familiar neutral faces. In this study, patients were first familiarized to a set of six faces using the procedure from the Schwartz et al. [77] study, with each picture of novel and newly familiarized neutral faces being randomly presented 16 times. Face stimuli were composed of black and white human neutral expression images, chosen from two sets of standard emotional expressions [78], and were randomly selected for the novel or familiar group, balanced across gender and stimulus set. SAD is a well-characterized phenotype and its patients often suffer from fear and avoidance of social situations [79, 80]. Some studies have shown that there is an association between SAD and inhibited temperament. The latter is defined as the predisposition for an individual to respond carefully to novelty, such as new places, things, situations and people or through avoidance behaviors [81, 82]. With regard to the amygdala response to faces, it may be particularly significant for inhibited temperament, given the increased risk for SAD. Face stimuli powerfully draw amygdala responses [83, 84], and numerous studies indicating that this amygdala behavior is guided by individual differences in inhibited temperament [80, 85, 86]. For instance, the study of Schwartz and colleagues showed that young adults who had been identified with inhibited temperament during childhood demonstrated greater amygdala activation to novel relative to newly familiar faces, compared with the individuals who with uninhibited temperament [80]. These findings support that amygdala acts as a mediator in the generation of different temperaments; however, it does not deal with whether there are variations in the timing of the amygdala response. Individual differences in the timing of emotional response, or "affective chronometry", have been suggested as an important component of affective style [88]; for example, this pattern of emotional response may be described as a "quick temper" or may be capable of recovering rapidly from negative emotions. Individual differences during affective responses seem to be related to both introversion and extraversion [89] and depression [87, 88]. These findings apparently reveal differences in the timing of perception regarding emotional responses, supporting the amygdala involvement. Thus, the findings of Blackford et al. [76] showed amygdala responses of patients with inhibited temperament were more rapid to novel compared with familiar faces. These results provided initial evidence for temperamental differences in the timing of the amygdala response to faces. Based on the rationale of the timing of amydala response to emotional stimuli, Luo et al. [89] used magnetoencephalography (MEG) combined with the advanced source analysis technique synthetic aperture magnetometry in 16 healthy young participants. Luo and colleagues aimed to investigate the automaticity of the amygdala response to emotional stimuli applying the modified paradigm of Erthal and colleagues that involves high and low attentional loads and fearful and neutral distracters [90]. In this task, participants responded to fearful and neutral faces bracketed between two lines by judging, via bottom press, whether the lines were parallel or not. The emotional load of the task was manipulated by varying angular differences of the lines on nonparallel trials. Participants received an initial stimulus lasting 300 millisecond that was a face given in brackets by two lines. After the initial stimulus presentation, participants were exposed to a blank screen lasting 200 millisecond before the response window appearance for 1500 millisecond, where the individuals had to respond by pressing the left (S) or the right (D) button after seeing the cue presentation S or D. After the response window, a new blank screen lasting 600 millisecond was presented to participants. There is a hypothesis that early amygdala response may be relatively independent of attentional control, and a later amygdala response to the same emotional stimuli may be subject to significant attentional modulation. The amygdala responds to emotional stimuli through a fast but common subcortical pathway (thalamus amygdala) in parallel to a slower, familiarized cortical pathway [91-93]. It is hypothesized that there is also a dual route in humans, the first suggesting that the amygdala s response to emotional stimuli is largely automatic and possibly is independent of attention [94-97]. The second, suggesting that unlikely the representations of emotional stimuli are independent of representational competition and that the increase in

Time Perception Distortion in Neuropsychiatric and Neurological Disorders CNS & Neurological Disorders - Drug Targets, 2013, Vol. 12, No. 5 573 attention relative to nonemotional stimulus features should notably decrease the amygdala response to emotional stimulus features [98-102]. The authors found an early amygdala response to emotional information (40 140 millisecond) that was independent (i.e., unaffected) of attentional modulation due to the low emotional load and a later amygdala response (290 410 millisecond) that showed significant attentional modulation due to high emotional load. The results suggest that there is a relationship between degree of automaticity of the amygdala response and time, in other words, the automaticity of amygdala response seems to be a function of time. These findings indicate that attentional load cannot modulate early amygdala response because this process occurs in advance to task-related activity observed in lateral frontal and parietal cortices. On the other hand, the significant interaction between emotion and task load seems to be responsible for later amygdala activity, occurring after task-related activity within the same regions. The latter finding, in fact, may reflect the involvement of top-down attentional mechanisms. In particular, the stronger representation observed during trials of high-load via topdown attention is related to task-relevant information and consequently should lead to a decrease in representation of emotional distracter information provoked by representational competition [103, 104]. With respect to the issue about timing of the brain's response, these findings provide initial evidence for the temporal dynamics of the amygdala response related to faces and emotional perception. Expanding the knowledge regarding individual differences about how our brain behaves when someone responds to stimuli may lead to new questions concerning the underlying mechanisms respect to emotional perception involved in the development of anxiety disorders. Mood Disorders Abnormality of subjective time experience is wellrecognized in psychiatric illness, such as depression and bipolar disorder [105]. In this regard, different findings are found regarding subjective experiences related to time perception. Depression is known to affect several cognitive functions, but little is known about the relationship between depression and time perception [106]. With respect to bipolar disorder, theoretical and empirical evidence suggests an impaired time perception related to internal timing mechanisms [107]. Many studies suggested that slowed time experience in depression could be a psychomotor deficit, while another studies have argued that this disturbance could be nonspecifically linked to the global severity of the depression [70]. For instance, the studies of Blewett and Mahlberg et al. [70, 108] reported a reduction in the time estimation in depressive patients, while the opposite was observed in maniac patients. These findings show a close relationship between time perception and cognition, due to the influence exerted by each emotional state on time perception [109]. In line with this, the study of Sevigny et al. [74] compared depressed and non-depressed individuals in attentional and temporal processing tasks and demonstrated that attention influences temporal processing. In this experiment, depressed individuals made more omissions, but not more erroneous responses, than non-depressed individuals. In addition, discrimination of relatively long intervals was poorer for depressed individuals, in contrast to discrimination of brief durations. In addition, a significant difference between depressed and non-depressed individuals regarding the variability of 1- or 10-second interval productions was observed in a continuous series of finger taps. As well, attention demands to long-interval processing seem to be a critical factor in depression-induced deficits of temporal processing. Other evidence came from the study of Kuhs and colleagues, who examined alterations of time experience in depressive patients by time estimation tasks [71]. Depressive patients underestimated prospectively a 30-second interval by 6 seconds, while healthy individuals overestimated this interval by more than 10 seconds during eight successive measurements over a period of 2 days. Depressive patients had a more pronounced time estimation error accompanied by a feeling of being unwell comparing to a relatively good state of well-being. A few studies were performed regarding time reproduction. For instance, the experiment of Mahlberg found smaller reproduced time-intervals in maniac patients when compared to the depressive ones [108]. Furthermore, while the maniac patients reproduced short and correct timeintervals and underestimated long intervals, the depressive ones reproduced long intervals correctly and overreproduced the short time-intervals. With regard to the time estimation and production in mood disorders, some experiments found an overestimation of the time duration in depression [71, 110, 111], while other investigations found an underestimation of time duration [112-114]. However, these pioneer studies were much criticized due to the lack of reliable diagnostic and methodological procedures. In this manner, Bschor compared estimation time and production tasks in depressive and maniac patients, and demonstrated that depressive and maniac patients overestimated only tasks of longer time production, this overestimation being more prominent in maniac patients [115]. With respect to bipolar disorder, theoretical and empirical evidence suggests that internal timing mechanisms are disrupted in bipolar disorder, regardless of symptom status [114]. This can be due to a failure in the neural circuitries involved in time perception leading to deficits in internal timing mechanisms. Cerebellum, basal ganglia and prefrontal cortex that are involved in sub-second timing, have also been implicated in the pathophysiology of bipolar disorder. Nevertheless, sub-second timing intervals have rarely been investigated in bipolar disorder. With this in mind, bipolar and matched-controls were asked to perform a finger-tapping task with dominant and non-dominant fingers in time with a paced (500-millisecond intertap interval) auditory stimulus (synchronization), and with no auditory stimulus trying to preserve the same pace (continuation). Bipolar individuals demonstrated greater timing variability compared to matched-controls regardless of stimulus or dominant fingers. Deterioration in timing variance into internal clock in opposition to motor implementation elements observed by the Wing-Kristofferson model is evidence for higher clock variability in bipolar individuals relative to matched-controls, since there were no differences

574 CNS & Neurological Disorders - Drug Targets, 2013, Vol. 12, No. 5 Teixeira et al. between them on motor implementation variability. In this regard, this increased clock variability in bipolar disorder may be associated with abnormalities in cerebellar function [116]. Therefore, the findings about the relationship between mood disorders and time perception propose that the brain could represent time in a distributed manner, telling the time by detecting the coincidental activation of different neural populations [6]. Schizophrenia Study showed that schizophrenia may be related with disturbance in the temporal coordination of information processing. This relationship may lead clinical symptoms such as hallucinations, delusion, psychomotor poverty and poverty of speech [117, 118]. Despite the growing interest and the time-dependent conceptualizations with regard to the pathophysiology of schizophrenia, there is a shortage of research directly examining timing perception in schizophrenia. In addition, the few studies which have been conducted revealed different findings due to different methodologies and inconsistent definitions. For instance, the experiment of Johnson showed that schizophrenic patients compared to controls were not significantly inferior to controls in reporting their own age, and in addition, with respect to time estimation, a temporal disturbance was noted in brief intervals [119]. In another experiment, Densen examined time perception of schizophrenic patients compared to normal individuals using an estimation method composed of four intervals ranging from 5 to 120 seconds, where each individual was asked to execute a task with paper, scissors, 20 pennies and one deck of playing cards; when the experimenter asked to stop the task, the individuals answered the time spent on the task, i.e. estimating the time duration of the task [120]. There were significant differences between schizophrenic patients and normal individuals, i.e. schizophrenic patients were not simply less accurate in time estimation tasks, as they also tended to overestimate time intervals (only for judging 5- second intervals). Wahl employed several methods of time estimation and maintained a consistent definition of overestimation and underestimation across three types of time-estimation, comparing schizophrenic with control individuals [121]. The first task was longer interval estimation, involving judgment about how much time had passed. The second task was verbal estimation, requiring an individual s judgment about the length of brief intervals signaled by the experimenter. The last task was operative estimation where the individual indicated when a specified number seconds had passed. Schizophrenic patients were notably more inaccurate than controls in the verbal and operative estimation tasks, i.e. overestimation when the individual judging that more time passed than in fact occurred, both verbal and operative estimation outcomes showed schizophrenic patients were notably more liable to overestimate. A significant difference was observed only in long interval estimation task; however, the non-specific nature of the temporal judgments related to the task made it a less valid indicator of ability to estimate. More recently, Carroll et al. [118] examined timing in schizophrenia using a well-established task of time perception. They requiring individuals to make temporal judgments related to hearing (sound made by birds) and visual stimuli (silhouette of a flying bird) presented ranging from 300 to 600 milliseconds. Schizophrenic patients and control individuals showed greater visual timing variability than when hearing. Moreover, the experimenter did not note a difference between schizophrenic patients and controls in the visual modality. On the other hand, schizophrenic patients had less temporal hearing precision when compared to the control group. Besides the inconsistent findings, the data suggest that schizophrenics may have a disturbed sense of time, with real (clock) time experienced as passing more slowly than normal individuals. Moreover, it is hypothesized that the corticostriatal circuit may be related to this process, specifically in the dysfunction of explicit codification in temporal representation, since the activation of the right caudate and putamen is strictly associated with the time-interval codification [122]. This association produces an augmentation in the processing of the internal-clock producing a subjective slowness time when compared to real time [118, 123]. Parkinson s Disease Time estimation is a subjective process that requires the use of an internal clock to measure objective time with no assistance of cues from external clocks. Due to the difficulty to find the internal clock in the brain, the issue of timing and time perception has resisted to define a better understanding and to better separate itself from the study of other cognitive processes, e.g., attention and memory. In addition, an important issue has been raised, in that interval timing in the seconds to minutes range may be derived from other cognitive processes (e.g., attention, memory, planning, decision making) and may not have its own essential features or neural networks [124, 125]. With respect to interval timing, researchers recently recognized it as an essential component of cognition, taking into account the identification of neural mechanisms responsible for encoding the stimuli duration [20, 21, 126-131] and because of the identification of interval-timing dysfunctions in a variety of neurological diseases [132-136]. The main source of agreement, however, is that movement control and its related variation can be divided into the perceptual aspects of interval timing, consequently describing a specialized process considered as the internal clock. Within this context PD, a neurodegenerative disease, is characterized by dysfunction of the basal ganglia loop as a result of a degenerative process in the nigrostriatal pathway, with progressive cell death in the substantia nigra pars compacta leading to DA reduction in the striatum, causing indirectly cortical dysfunctions [137]. PD is accompanied by a series of motor and cognitive deficits of processes that require accurate timing, such as disruption in reaction time, inter onset latency, rhythm maintenance, and temporal organization of speech. These processes are well correlated with the clinical phenomenon of bradykinesia in PD [138-140]. Actually,