Prepulse inhibition as an index of attentional processing: A comparison between young and elderly

Prepulse inhibition (PPI) is assumed to index attentional processes by inhibition of the startle reflex. By directing attention towards a weak stimulus, i.e., the prepulse, PPI is increased. We investigated controlled and automatic processes related to attention in young and elderly subjects. Both groups (n=41) attended to a task where they were to judge if length of a comparison tone was shorter or longer than prepulse. Degree of PPI was assessed by different stimulus onset asynchronies (SOA) assumed to index automatic and controlled processing. We predicted firstly that the young would show established PPI values. Secondly, that normal elderly would show a increase in PPI compared to young when attending to task compared to no-task. As predicted, we found normal PPI function in the young. In the elderly, the expected hyperbolic and quadric PPI function failed to display. Thus, a straight PPI line suggests a continuously elaborating of the first prepulse and pulse in task and No-task, meaning that task made no difference in attentional processing in the elderly. The hypothesis of a global decline in inhibitory function in elderly is suitable as an explanation for reduction in PPI, and we assume that results is due to inhibitory problems in attention, as a consequence of physiological aging in cortex. Prepulse inhibition and attention 3 Introduction Common for all humans is aging. It is well established that physiological, as well as cognitive capacity, is reduced when the years go by. As cognitive resources degrade, is it natural to think that selective attention will be reduced (McDowd & Filion, 1992). Attentional Process It is common to divide attention in two processes. Automatic processes are silent in the sense that they operate outside conscious awareness, and refer to the processes of detection, analysis, and identification of stimuli. They occur at Stimulus Onset Asynchrony (SOA) of less than 120 ms (Dawson, Schell, Swerdlow & Fillion, 1997). The individual merely detects a stimuli and the process of detecting simple features like intensity, pitch and shape (Öhman, 1997) takes place. Thus, what eventually is experienced as an conscious experience by the subject, are not raw material, but elaborated sensory information that already been processed (Cook III & Turpin, 1997). This process happens unconsciously, automatic, quickly and effortless, in a way that the individual have no recollection of the process that just took place (Elden, 2002). In controlled processing, only a fraction of the incoming stimuli are chosen for further processing (Elden, 2002). It occurs relatively slowly and is dependent on conscious awareness, and occur at SOA longer than 120 ms (Dawson et al., 1997). Since it is dependent on limited attentional resources, under intentional control, and associated with conscious experience (Dawson et al., 1997), this is the processing the subject is able to recollect Attention In Aging In psychology, attention is viewed as capacity to, or energy to support cognitive processing (Lezak, 1995). A central division is between controlled and automatic processes. A deficit in selective attention may be central to explain changes in cognitive performance during aging, and there is evidence of an age deficit in selective attention (Hasher & Zacks, 1979; Kausler, 1994). Selective attention is the phenomenon that one attends to the stimuli of most interest, and filter out irrelevant stimuli. According to two-process theories of selective attention (Posner & Snyder,1975), facilitation of relevant stimuli and inhibition of irrelevant stimuli constitute different aspects of selective attention. Common for theories postulating deficits in selective attention, is the result of a deficit in inhibitory functions (Ellwanger, Geyer & Braff, 2003), and with aging selective attention declines due to a failure in inhibitory versus facilitory processes. Salthouse (1985, 1988) Prepulse inhibition and attention 4 claims that age-related differences in attentional functioning are the result of a reduction in the energy that “fuels” cognitive processing, and this fuel can be increased by arousal (Kahneman, 1973). By using Skin Conductanse Orienting Response (SCOR), McDowd & Filion (1992) postulated that elderly would show increment in inhibitory functions, and as a consequence allocate their attentional resources less efficiently than young adults. Their investigation was performed in the context of the “DualProcess model of selective attention”, a model that emphasize both selection of relevant stimuli and inhibition of irrelevant information. This model has been developed theoretically and empirically after Triesman’s model of 1969 (as cited in McDowd & Fillion, 1992). The subject was instructed to attend toward a story, or some noises (75 dB, 1000 Hz) presented during low loudness of the story. The noise was presented simultaneously with the story, and the subject had to ignore noise or story. Results show that the young adults were more accurate than the elderly, and the elderly were less able to inhibit responding to irrelevant stimuli than the young. The authors stated that inhibitory processes were compromised in the elderly, and this produced a deficit in selective attention. These results are in accordance with data suggesting that older adults have reduced inhibitory control relative to young adults (McDowd & Fillion, 1992). Working Memory. Hasher & Zacks (1979, 1984, 1988) found that automatic encoding processes were presumably unaffected by a persons limited resources. They proposed a “capacity theory”, stating that working memory (WM) is reduced with aging because less efficient inhibitory process fail to prevent irrelevant information from entering, or being processed, in WM. That storage functions is reduced with aging, comes from evidence suggesting that task that make demands on the storage component, have a particularly disruptive effect on the elderly compared with younger adults (Hasher & Zacks, 1988; Light & Capp, 1986). The notion of limited capacity and different demands on that capacity, are most often embedded in the concept of working memory (Hasher & Zacks, 1988). Working memory is conceived as a limited capacity mechanism which share its resources between a storage function and a set of processing functions (i.e. attentional analysis). Demanding tasks may place a large burden on storage function in WM, thus having access to prior knowledge about the demanding task, will relive strength on the total amount of capacity available. This process can be expected to malfunction, preventing new inferences from being drawn, and preventing the representation of a general construction of the stimuli (Hasher & Zacks, 1988). The capacity model assumes Prepulse inhibition and attention 5 that there is a competition between processing and storage, and due to the decreasing supply of capacity in the elderly, the processing component of WM has higher priority than the storage compartment. The model involves two basic mechanism of selective attention, namely activation and inhibition. Inhibition is suppressing of irrelevant information so that such information is less likely to have access to WM, and irrelevant information and previously relevant information that does enter WM, is quickly removed. Attentional inhibition may also have the function of preventing the return of attention to a previously rejected item. The presence of this irrelevant information results in poor encoding, retrieval, and comprehension of incoming information (Zacks, Radvansky & Hasher, 1996). Critics argue that the major limitation of the theory, is the establishing of good measures of attentional capacity. To measure attentional capacity, startle eyeblink modification (SEM) is a common measure and have proven reliable (Fillion, Dawson & Schell, 1998). Startle Eyeblink Modification Startle eyeblink modification involves a relatively intense stimulus (e.g. a sudden loud noise), which elicits a startle eyeblink reflex. The nature of SEM have been divided in two main classes, the first is modification and latency of the reflex by lead stimuli presented up to approximately 500 ms (called short lead interval effects), and second; modification with lead stimuli presented longer than 500-800 ms prior to the startle reflex (called long lead interval effects)( Fillion et al. 1998).). Thus, to achieve inhibition, short lead interval effects of 15 to about 400 ms (Graham, 1975; Elden & Flaten, 2002, 2003) is most common. The time period between the lead stimulus (prepulse) and the startleeliciting stimulus (pulse) determines if the lead stimulus facilitates or inhibits the strength of the reflex, and is known as Stimulus Onset Asynchronies (SOA). If the lead stimulus (prepulse) inhibits the startle reflex (pulse), the paradigm is known as Prepulse Inhibition (PPI). The strength of the reflex in the presence of the lead stimulus gives an image of the attention processes taking place in the subject. A weak prepulse seems to demand or dominate automatic processing capacity and inhibit startle reflex to subsequent stimuli. The automatic process is considered to be unconscious, fast and parallel in the sense that it can be performed together with other processes (Öhman, 1997), thus the amount of PPI can be used as an involuntary, nonverbal, index of automatic processing. Prepulse inhibition and attention 6 The most reliable component of startle is the eyeblink reflex (Landis & Hunt, 1939), a robust effect considered quite reliable and occurring in 90-100 % of normal adult participants (Fillion et al., 1998). SEM has proven suitable for the investigation of a global loss of inhibitory function, mainly because startle plasticity is assumed to give direct indices of inhibition (Fillion et al, 1998). Startle Eyeblink Modification In Aging. SEM in the elderly has not previously received much attention. However, a few studies have been performed. Harbin and Berg (1983,) found inhibition of airpuff elicited reflexes at an SOA of 420 ms in both young (mean = 20 years) and elderly (mean = 68 years) subjects. In a second study (Harbin & Berg, 1986), the participants did an attention demanding visual search task. By comparing young (mean = 19) and elderly (mean = 69), they reported no significant effect of age on PPI. Nearly significant results in age by condition interaction, indicated that young participant demonstrate more PPI during a condition engaging a task versus no task. The old subjects demonstrated equal PPI in task and No-task condition (Harbin & Berg, 1986; Ellwanger et al., 2003). However, Ford et al., (1997) found that young persons (18-25 years) demonstrated greater frequency of startle blink to loud noises than elderly (58-76years). Flaten and Powell (1998) investigated whether young and older subjects would show similar or different rates of reflex facilitation as a result to previous exposure to classical conditioning. In two age groups, one group (paired group) received a classical conditioning paradigm consisting of 70 trials, where the unconditioned stimuli (US) was an airpuff to the eye. Second group (unpaired group) received equal airpuff and same number of trials, but the stimuli was presented in a way that no conditioning took place. Result showed increased conditioning in the paired group consisting of young participants, with no conditioning in the elderly group. Further, there was increased reflex amplitudes in the young group compared to the old. The elderly displayed startle facilitation when conditioned, compared to reflexes elicited alone without conditioning present. There was no difference in noise elicited eyeblink amplitudes between young and elderly subjects. To see whether reduced Conditioned responses (CRs) in the elderly could be due to decreased Unconditional Response (UR), Flaten & Friborg (2005) investigated if reduced autonomic activation might explain the impaired acquisition of eyeblink Conditioned Respons (CR). This was based on literature discussing if smaller reflex amplitude might be due to Prepulse inhibition and attention 7 decreased orbicularis oculi activation that control blink, or reduced sensory abilities. Participants were divided in two groups, young (mean 23) and elderly (mean 73). Participants was conditioned to airpuff and tones, and startle eliciting noises was used as a measure of reflex strength. The results showed a significant higher startle reflex magnitude in the young subject compared to the elderly. The young subject also showed increased startle reflex in the presence of CS compared to the elderly. In sum, significantly more frequent and larger eyeblink in young than in the elderly subject, data supporting Flaten & Powell (1998). However, startle facilitation was not related to conditioning in the old, contrary to Flaten and Powell. In a study concerning PPI in patients with Alzheimer, Hejl, Glenthøj, Mackeprang, Hemmingsen & Waldemar. (2003) used normal elderly (> 60 years) as control group. Their results in a passive paradigm and with continuous background noise, was significant PPI in SOA 30, 60, and 120 ms, compared to the pulse alone condition. Ellwanger et al. (2003) were the first to investigate the relationship between age and PPI. They hypothesized that aging would give a global decline in inhibitory functions, such as ability to ignore relatively irrelevant sensory, cognitive, or motor information. The participants consisted of four age groups spread among college students, young, middle age and old. The groups were not equally spread. All the participants were tested on part A and B of the Trail Making test, to investigate perceptual-motor speed and cognitive flexibility. In the experiment the participant’s were exposed to a “passive paradigm” (where they are not attending attention to any aspect of the experiment) with continuing background noise of 75dB. The SOA was 30 and 120 ms. Results showed that startle reflex had a slowing (latency decrease) and decline (magnitude decrement) with aging. Further, the results showed that the middle age group displayed the most PPI, and the college group the least. The most extreme age groups did not display any significant PPI effect, a finding that does not support the original expectation of general decrease in inhibitory function with age, thus a finding consistent with Harber & Berg (1983, 1986). Though, what remains unclear is how the elderly will process PPI if they attend to a task. Animal Studies. In rodents, startle is typically measured by using the whole body-flinch that occur in response to the startle stimuli. Of the most strongly supported research findings say that startle magnitude decreases with increasing age in both mice and rodents (Ison, Bowen, Pak & Gutierrez, 1997; Ellwanger et al., 2003). Ison et al. (1997) studied mice of three ages (young, Prepulse inhibition and attention 8 middle age, and old). Results revealed that startle response decreased with increasing age, and that percent PPI was not affected by age. By using rats of four ages (young, adult, middle age, and old), Varty, Haugher & Geyer, (1998) found that when acoustic or tactile stimuli was used to elicit startle, the old group consistently had the smallest magnitude. The condition of auditory stimuli, showed no effect of age on either PPI or startle, when ratio measures was corrected for age-related changes. Neurobiology Of Startle Eyeblink Modification. The neurobiology of startle and PPI has been described. The SEM indicate central nervous system activity, and this activity can be measured as electrical activity to the muscles. The muscle activity is driven by a set of neurons in the brain stem (Elden, 2002), described in rats by Lee et al. (1996) (as cited in in Elden, 2002). The first synapse is the cochlear root in the auditory nerve, a small nucleus made of very large cells. The cochlear root axon terminates directly in the nucleus Reticularis Pontis Caudalis (nRPC), situated in the medial tegmental pons. This nucleus is known as the startle center because electrical stimulation in this area elicits a startle respons, and lesions abolish it. From nRPC motoneurons project in the spinal cord, to the facial nucleus (nerve VII), that controls the pinna and the blink reflexs (Elden, 2002). A Positron Emission Tomography (PET) -study by Pissiota, Frans, Fredrikson, Langstrøm & Flaten. (2002) confirmed that this center was activated by startling noise in humans. What Causes PPI Inhibitory functions in SEM operate in automatic and controlled processing. The different SOA for these processes, are related to two main theories. The Protection Of Processing Hypothesis According to Graham (1975, 1992) there are two parallel processes that occur when a stimulus is perceived. The first is encoding and perceptual analysis of the stimulus, and the second is a protective process that attenuates all subsequent stimuli until the encoding of the stimuli is completed. Together the processes are called “The protection of processing hypothesis”. The rationale in this theory states that a startle stimulus would be perceived less intense if the prepulse reduce the available capacity to the attention system. Simple physical features such as intensity and pitch are detected and analyzed automatically. Cohen, Hoffman & Stitt (1981) used 80 dB tone as prepulse, and a tap on the forehead Prepulse inhibition and attention 9 (glabellar tap) as pulse. They found that the presence of the prepulse decreased both the size of the eyeblink elicited by the tap as well as the estimated intensity of the tap itself. Perlstein, Fiorito, Simons & Graham (1993) obtained similar results, using a 75 dB tone as prepulse, and 110 dB tone as pulse. The theory further state that perception analysis of the prepulse will influence the response to the pulse, i.e., the startle-eliciting stimulus. The analysis of the prepulse reduces processing of the startle stimulus, and inhibits the reflex. If startle inhibition serves to protect the perceptual processing of the pulse, then perception of prepulse should be more accurate when it is effective in producing startle inhibition (Fillion et al., 1998). Norris and Blumenthal (1995) instructed participants to indicate after each trial, whether a high-pitched prepulse, a low-pitched prepulse, or no prepulse had been present. Because the tone pitches was difficult to discriminate, these investigators were able to use the number of hits and misses for the target lead stimuli as a measure of the accuracy of lead stimuli perception. Results revealed that greater startle inhibition was produced on trials in which the lead stimulus was correctly identified. Elden and Flaten (2003) asked participants to judge whether a tone prepulse was shorter or longer than a comparison tone presented before prepulse. To make the task more difficult, a distracting airpuff was presented simultaneously with the tone prepulse. In experiment two, an airpuff was used as prepulse, and a tone was used as distracter. In the no task condition, the participants were instructed to not pay attention to the prepulse. The hypothesis was that more difficult task should increase PPI, because it would demand more attentional resources and hereby inhibit pulse compared to less demanding tasks. By presenting a distracter simultaneously with prepulse, the task difficulty should increase. Results showed that by changing the modality on prepulse, directing attention to both acoustic and tactile prepulse increased pulse according to established theory. Further, inhibition increased at SOA of 30, 60, and 420 ms.In sum, the presented experiments support the rationale presented in Protection of processing hypothesis. Results show that short lead interval effects (below 500 ms) inhibit the pulse according to automatic processes, and support the theory first and second processes. Long lead interval stimuli (above 500 ms) facilitate pulse according to controlled processing of stimuli. This supports the theory stating that analysis of the stimuli are completed and attentional resources are ready to interpret new sensory stimuli. Prepulse inhibition and attention 10 The Sensorimotor Gating Hypothesis Braff and Geyer (1990) view startle inhibition in a similar way as Protection of processing hypothesis. They stated that startle inhibition may serve as an operational measure of sensorimotor gating , “reflecting the ability to effectively buffer or screen out the potentially chaotic flow of information and sensory stimuli” (as cited in Cadenhead, Geyer & Braff, 1993). Startle inhibition is a basic function that inhibits sensory input, allowing the brain to process and elaborate the early stages of information processing. It is preattentive and automatic at very short interval (60ms), but may be controlled at longer lead interval (Elden, 2002). What differentiates this hypothesis from Protection-of-processing theory, is the suggestion that startle inhibition reflects a general ability to inhibit external stimuli (e.g. auditory, visual, tactile) as well as internal stimuli (such as thoughts and impulses)(Geyer, Swerdlow, Mansbach & Braff, 1990). In support of this theory is literature reporting deficits in startle among schizophrenia patients, obsessive-compulsive disorder, college students scoring high on psychosis-proneness scale, Huntington’s disease, and children with Attentiondeficit-disorder (as cited in Fillion et al., 1998). Relationship to attentional processing What remains unclear in both protection of processing hypothesish and sensorimotor gating hypothesis, are whether PPI occurs automatically, or are dependent of controlled attentional processing. Observation of startle inhibition in nonhuman animals, decorticated rats, in infants, sleeping adult humans, and the fact that pre-habituation does not affect startle inhibition in either humans or animals (Fillion et al., 1998), suggest that this inhibition is an automatic process. At the same time this does not exclude that inhibition could be modulated by controlled attentional processes. Delpezzo and Hoffman (1980) found that startle inhibition was greater in trials where the participants knew the location of a light in front of them. The participant’s was instructed to gaze at a grid of light located directly in front of them. In the first experiment, participants were warned on half on the trials where the light would appear, and were given no information on the other half. Results of interest are that PPI was inhibited on trials where the participants were instructed to see where the light would appear, that a focus on the prepulse increase the amount of inhibition on the pulse. In a second, experiment they altered the procedure so that participants were instructed to focus on the light in half of the presentations, and the other half to focus 40 degree left to the grid position. In this experiment the light in the central position of the grid was on continuously. Result revealed that inhibition of pulse was greater when participant was instructed to focus on the light in the Prepulse inhibition and attention 11 central position, than when they should focus to the left of the light. Results seems to support that controlled attentional processing inhibit pulse. Elden and Flaten (2002) had hypothesis that prepulse would increase PPI, and that PPI should increase on trials with correct judgement of prepulse duration. By instructing the subjects to listen to a comparison tone previous the prepulse, the task was to judge whether prepulse was shorter or longer than the comparison tone. The results from this first experiment, as an between-subject design, showed that judgment of the duration of the prepulse increased PPI. It was hypothesized that the amount of attention directed to the prepulse, the higher was the probability that they would judge correct. Result revealed that paying attention to the prepulse increased PPI, but performing well on the task did not accentuate PPI further. In a second experiment the participants, as a part of a within-subject design, did the same task as in the first experiment. Results replicated finding from experiment one, that the reflex is inhibited by attention to the prepulse. Overall findings are that PPI increase over different SOA if the prepulse are attended to. To investigate whether different SOA in long lead interval was affected, Schell, Dawson, Hazlett & Fillion (1995) investigated attentional modulation. The participants were instructed to perform a triple task: 1) listen to a series of high and low pitch tones: 2) to silently count the longer than usual high pitched tones; and 3) to ignore the low pitched tones. The standard length of the tones was 5 seconds, and the longer than usual tones was 7 seconds in duration. Results revealed that control participants had greater inhibition of startle blink at 120 ms and greater facilitation at 2000 ms during the to-be-attended task. Filion et al. (1993, 1994) did an experiment following the same experimental procedure, where subject was presented with 5 and 7 seconds intermixed tones of different pitch. The task was to keep count of 7 seconds tone of one pitch, and to ignore the others. Similar to Hazlett & Fillion (1995), PPI at SOA 2000 ms was facilitated with greater facilitation when attending to task. Jennings et al. (1996) did a follow up study, investigating SOA 2000, 4500 and 6000 ms. They found that greatest facilitation was observed in 6000 and 4500 ms compared to 2000 ms, and that attending to task gave greater facilitation compared to not attending to task. Jennings et al. (1996) claim that the degree of facilitation appeared to reflect task difficulty, though the results indicate that the amount of startle facilitation at long lead interval is a function of attentional processing. In summary, the present experiments support the notion that controlled attention processes inhibits pulse processing. This is in accordance to both protective of processing theory, and sensory gating theory. Thus, both theories can explain how short lead interval Prepulse inhibition and attention 12 inhibits, and long lead interval facilitates, the startle reflex. However, Elden and Flaten (2002, 2003) argue that short SOA at 30 90 ms, increase PPI due to preparatory attention. Preparatory attention is directed at a predictable upcoming event, and makes information processing more efficient with anticipation (Bastiansen, Koen & Brunia , 2001), i.e. if a prepulse is not present, the anticipation of the prepulse might alter and inhibit the pulse. In this way, controlled attention may modulate automatic processing of prepulse. The aim of this study is to investigate how attentional resources are reduced in the elderly by using the paradigm of PPI. In elderly, several theories state that increasing age reduces cognitive resources compared to young adults (Lezak, 1995). By use of “Protection of processing theory” (Graham, 1975), we have a background to understand how PPI is processed in the brain. Elden & Flaten (2002) investigated how automatic processing differ between young subjects when they did a task, compared to a notask. In this study, by following the same procedure as Elden & Flaten (2002), we assume that the young group will display similar results as their experiment. For the elderly, we expect to see if they display the same results as the young when it comes to attentional resources in task versus no-task. Ellwanger et al. (2003) found no difference between the elderly and the young. We assume that the elderly will display increased PPI on task trials compared to notask trials. By following rationale of automatic versus controlled attention, we assume that the elderly will display decreased PPI processing on SOA between 120 – 2000 ms. This is based on prediction that cognitive aging have compromised the resources available to task in the elderly, and the necessary resources to accomplish task will be reduced. As a consequence, the startle will be more inhibited in the elderly compared to the young In no-task we assume that there should be no difference between the elderly and young subjects, in line with Ellwanger et al. (2003). By using cognitive neuropsychological screening test, we assure that none of the elderly have reduced cognitive capacity on a diagnostic level.