Attenuation of the prepulse inhibition of the acoustic startle response within and between sessions

Biological Psychology 71 (2006) 256–263 www.elsevier.com/locate/biopsycho

Attenuation of the prepulse inhibition of the acoustic startle response within and between sessions Boris B. Quednow *, Kai-Uwe Ku¨hn, Katrin Beckmann, Jens Westheide, Wolfgang Maier, Michael Wagner Department of Psychiatry, University of Bonn, Sigmund-Freud-Straße 25, D-53105 Bonn, Germany Received 12 July 2004; accepted 8 May 2005 Available online 12 July 2005

Abstract Prepulse inhibition (PPI) and habituation of the acoustic startle response (ASR) are widely used biological markers in the study of psychiatric disorders and have been shown to be homologous across species. Previous studies in humans suggested that PPI is a stable and reliable measure between test sessions, but that PPI decreases within sessions. The purpose of this study was to explore the short- and longterm decrease in PPI as a potential confound in the measurement and interpretation of PPI. We investigated the progression of PPI and habituation of ASR in three test sessions spaced 4 weeks apart in a group of 20 healthy participants. Analysis revealed a significant decrease in the percent PPI within and between the test sessions. Nevertheless, PPI was reliable across three test sessions, indicating that the significant attenuation of PPI over time was a consistent phenomenon. These results suggest that PPI exhibits short- and long-term attenuation. # 2005 Elsevier B.V. All rights reserved. Keywords: Prepulse inhibition; Habituation; Acoustic startle response; Sensorimotor gating; Electromyography; Clinical studies; Stability; Reliability

1. Introduction The startle reflex is a fast response to a sudden, intense stimulus such as a loud sound and consists of contraction of the skeletal and facial musculature. This reflex is usually classified as a defensive response. The acoustic startle response (ASR) of mammals is mediated by a simple threesynapse neuronal circuit located in the lower brainstem. Neurons of the caudal pontine reticular nucleus are key elements of this primary ASR pathway (Davis et al., 1982; Koch, 1999). The startle reflex shows several forms of behavioral plasticity, such as prepulse inhibition (PPI) and habituation. PPI refers to the reduction of ASR magnitude when a distinctive non-startling stimulus is presented 30– 500 ms before the startling stimulus. PPI is used as an operational measure for sensorimotor gating that reflects the ability of an organism to properly inhibit sensory information (Graham, 1975; Hoffman and Ison, 1980). Habituation is a theoretical construct that refers to the reduction in * Corresponding author. Tel.: +49 228 287 5681; fax: +49 228 287 6949. E-mail address: [email protected] (B.B. Quednow). 0301-0511/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.biopsycho.2005.05.001

magnitude of ASR after repeated presentation of the startling stimulus that is not due to muscle fatigue or blunting of sensory receptor responsiveness (Groves and Thompson, 1970; Siddle and Kroese, 1985). PPI and habituation of the acoustic startle response are consistent phenomena across species and are widely used to measure sensorimotor gating and to investigate information processing (Braff et al., 1992; Geyer and Braff, 1987; Swerdlow et al., 1999). Several investigations have reported changes of habituation and/or PPI of ASR in neuropsychiatric disorders such as schizophrenia (Braff et al., 1992; Geyer and Braff, 1982; Parwani et al., 2000), schizotypal personality disorder (Cadenhead et al., 1993), obsessivecompulsive disorder (Swerdlow et al., 1993), and Huntington’s disease (Swerdlow et al., 1995). Changes in PPI and habituation may provide trait markers for psychiatric disorders with altered neurotransmitter regulation (Cadenhead et al., 1999). To enhance our knowledge of PPI deficits in patients with various neuropsychiatric disorders and to develop translational animal model studies, investigators have increasingly utilized psychopharmacological modulation of PPI in

B.B. Quednow et al. / Biological Psychology 71 (2006) 256–263

normal humans and animals (Braff et al., 2001). For a valid interpretation of pharmacological effects on a biological marker, such as in pre-post designs with a pharmacological treatment, the marker should be stable and reliable. Crossover designs are well established in studies with pharmacological modulation of PPI in rodents as well as healthy volunteers. However, clinical studies – especially with schizophrenic patients – utilizing longitudinal designs in combination with a control or a placebo group are missing so far. Graham (1975) stated that PPI may reflect a strong and stable automatic process that serves to protect the processing of the weak prepulse from disruption by the intense startle stimulus. This view was supported by the findings that PPI is present on the first presentation of a prepulse-startle stimulus pairing, does not habituate, and occurs during sleep (Graham and Hackley, 1991). At first, these findings seemed to indicate that PPI might not be influenced by processes such as learning or habituation (Lipp et al., 1994). In the meantime, several studies have shown that the amount of PPI decreases over repeated trials within a startle session (Lipp et al., 1994; Blumenthal, 1996; Lipp and Krinitzky, 1998; Lipp and Siddle, 1998). Furthermore, it could be shown that selective attention influences the extent of PPI (Filion et al., 1993, 1994; Schell et al., 1995). These findings do not confirm the notion that PPI is a strong and stable automatic process (Lipp and Krinitzky, 1998). Blumenthal (1997) remarked that a decrease of PPI within test sessions could be due to habituation of some aspect of the inhibitory processes initiated by the prepulse, or to habituation of the startle response itself. He found that the amount of PPI decreased across trials, but that the proportion of inhibition (percent PPI), measured as a difference between the average magnitude of blinks preceded by a prepulse and the average magnitude of those that were not, divided by the average magnitude of those that were not, remained constant across trials (Blumenthal, 1997). Previous habituation to the prepulse did not reduce subsequent PPI when prepulse trials were paired with startle stimuli (Blumenthal, 1997; Lipp and Krinitzky, 1998; Schell et al., 2000). In addition, the reduction in PPI across trials can be abolished after dishabituation of the startle reflex (Lipp and Krinitzky, 1998). These data supported the suggestion that the reduction of PPI seen across trials is not due to habituation of the prepulse but is related to startle reactivity in pulse-alone (control) trials, which is reduced by habituation. Some studies have been performed to investigate the reliability and stability of PPI also between test sessions (Schwarzkopf et al., 1993; Abel et al., 1998; Cadenhead et al., 1999; Swerdlow et al., 2001). In contrast to previous studies, which found a decrement of PPI within sessions, none of these studies found significant changes in PPI between two or three test sessions, and all studies have shown high test-retest reliability of PPI. Only Abel et al. (1998) reported a slight decrease of percent PPI across sessions. Test retest reliability of PPI of somatosensoric

257

elicited blink reflexes seems also to be very high (Flaten, 2002). However, the number of prepulse trials measuring PPI, as well as the sample sizes, in most of the previous studies was possibly not sufficient to investigate the longterm stability of PPI. The following listing of studies shows the number of trials of the single prepulse condition within a session which can be expected to exert the strongest effects on PPI: Schwarzkopf et al. (1993) used 24 trials in each of three sessions at 1-week intervals (n = 11), Abel et al. (1998) used eight trials in each of three sessions which were separated by a minimum of 2 h on 1 day (n = 15), and Cadenhead et al. (1999) used 12 trials in each of three sessions at 1-month intervals (n = 10). Swerdlow et al. (2001) used, however, a special study design to examine pretesting of ASR measurement in humans to diminish variability across experimental groups (n = 20). They used 5 trials in a pre-testing session and 7–10 days later they used 25 trials (5 trials in 5 identical sessions with 15 min between each session). Because of the different results between within- and between-session stability of PPI in previous studies, the aim of the present study was to investigate if there is a dissociation between short- and long-term stability of PPI. To accomplish our goal, we measured magnitude, habituation, and PPI of ASR three times during a span of 8 weeks using a startle sequence optimized for habituation effects consisting of 36 prepulse trials under a single condition (86 dB prepulse; SOA 140 ms) and 36 pulse-alone trials (116 dB) per session in 20 healthy volunteers.

2. Materials and methods 2.1. Participants PPI and habituation of ASR were measured in 20 healthy participants (8 women, 12 men, all were of Caucasian ethnicity), who were recruited by advertisement in a local newspaper. Age ranged from 18 to 63 years (38.4  12.2 (mean and standard deviation)), and the years of education ranged from 9 to 17 years (14.3  2.5). Four of the participants were smokers (one woman, three men) and one woman was post menopausal. All participants were employed at time of experiment. Exclusion criteria consisted of legitimate use of psychotropic medication and/or illicit drug use. An SCIDI interview was carried out according to DSM-IV procedures by a psychologist trained in the use of this instrument. None of the included participants had a personal or family history of any DSM-IV axis I psychiatric diagnosis. In addition, none of the participants had a history of migraine, epilepsy, or craniocerebral trauma. The study was approved by the Ethics Committee of the Medical Faculty of the University of Bonn. After receiving a written and oral description of the aim of this study, all participants gave written informed-consent statements.

258

B.B. Quednow et al. / Biological Psychology 71 (2006) 256–263

2.2. Procedure

2.3. Statistical analysis

Four and eight weeks after the first test session, the measurement was repeated. The first ASR assessments were carried out after informed consent was given by all participants. Before each ASR assessment, all participants underwent a brief hearing screening to insure hearing within normal limits. Participants would have been excluded on the basis of hearing impairment at 40 dB (1000 Hz). Electrodes were attached while the participants were seated in a comfortable chair. All examinations were performed by the same experimenter. Participants were advised that they would hear white noise and bursts over the headphones and were instructed to keep their eyes open. The eye-blink component of the acoustic startle response was measured using an electromyographic startle system (EMGSR-LAB; San Diego Instruments, Inc., San Diego, CA). Recorded EMG activity was band-pass filtered (1–1000 Hz). A 50-Hz notch filter was also used to eliminate 50-Hz interference. The amplifier gain was kept constant for all participants, and the EMG was recorded from the onset of the acoustic startle stimulus for 250 ms with a sampling rate of 1 ms. Two silver/silver chloride electrodes were positioned below and to the outer canthus of the right eye over the orbicularis oculi muscle, and a ground electrode was placed on the glabella. All electrode resistances were less than 10 kV. Acoustic startle stimuli were presented binaurally through headphones (TDH-39-P; Maico). Each session began with a 4-min acclimation period of 70 dB background white noise that was continued throughout the session. Participants received 73 white noise sound pulses at an intensity of 116 dB (instantaneous rise/fall time), and a duration of 40-ms separated by variable inter-trial intervals between 8 and 22 s (mean 15 s). In 36 of the trials, the pulse was preceded by a 20-ms 86-dB white noise prepulse (instantaneous rise/fall time) with a stimulus onset asynchrony (SOA) of 140 ms. The initial trial was a pulse-alone trial, which was separated for further analysis. All following trials were presented in a pseudorandomized order. The entire test session lasted about 20 min. Smoking was permitted before assessment, but none of the four smokers took advantage of this. Voluntary and spontaneous blinks were excluded from further analysis using the registration parameters described by Braff et al. (1992). The latency to startle response onset was defined by a shift of 2.28 mV (six digital units) from the baseline value occurring 21–120 ms after the acoustic startle stimulus. Latency to response peak was defined as the point of maximal amplitude that occurred within 150 ms after the startle stimulus. Response rejections were made both in case of onset-to-peak latencies >95 ms and baseline shifts >34.2 mV (>90 digital units). Error trials were defined as trials in which no startle response was recorded because of a baseline shift (e.g. due to spontaneous or voluntary blinks). Two women with error trials >50% were excluded from data analysis.

The mean percent PPI of startle magnitude was calculated using the formula [% PPI = 100  (magnitude on pulsealone (PA) trials magnitude on prepulse (PP) trials)/ magnitude on PA trials] (Braff et al., 1992; Blumenthal et al., 2004). PA and PP trials were divided each in six blocks for assessing habituation of startle reactivity, and PPI was also calculated within each block for assessing attenuation of PPI. The percent habituation was calculated as the reduction in startle magnitude between the first and last block of PA and PP trials [% HAB = 100  (first block last block)/first block] (Geyer and Braff, 1982). As a further measure for habituation, the linear gradient coefficient b was calculated across six blocks of PA and PP trials within each subject [b = (nSxy (Sx)(Sy))/(nSx2 (Sx)2); x = block number, y = startle amplitude PA trial per block], according to Geyer and Braff (1982). The linear gradient coefficient b represents a measure for the rate of fall (negative ascending slope) of habituation. Startle magnitude was assessed by using the mean values of all PA trials, the first block of PA trials and the initial PA trial. PPI, habituation measures, and startle amplitude were analyzed by analysis of variance (ANOVA). ANOVA with more than one degree of freedom in the numerator were adjusted by means of the GreenhouseGeisser correction (Winer, 1971). The original degrees of freedom and the adjusted p-values are reported. Interrelations between startle reactivity and PPI were tested by Pearson’s product-moment correlation. The reliability of the different parameters over time was investigated using the intraclass correlation coefficient (ICC) model (3,3) (McGraw and Wong, 1996), which is equivalent to Cronbach’s alpha coefficient of internal consistency. The coefficient estimates the proportion of total between-subject variability due to ‘‘error-free’’ between-subject differences. The confirmatory statistical comparisons of all data were carried out at a significance level set at p < .05 (two-tailed).

3. Results 3.1. Prepulse inhibition Table 1 shows the mean percent PPI at baseline, week 4, and week 8. There was a significant decrease in mean percent PPI and a strong trend for a decrease in percent PPI in the first block across the test sessions (for statistics, see Table 1). Analysis of the contrasts with an ANOVA with repeated measures at factor session showed a significant linear trend for the decrement of mean percent PPI across sessions, F(1, 17) = 5.11, p < .05. A 6  3 (block  session) ANOVA of percent PPI revealed a significant main effect of the factor block, F(5, 85) = 2.73, p < .05, reflecting the decrement of PPI within the test sessions, and a significant main effect of the factor session, F(2, 34) = 4.91, p < .05, reflecting the progressive

B.B. Quednow et al. / Biological Psychology 71 (2006) 256–263

259

Table 1 Means and S.E. of means (in parentheses) of the initial pulse-alone (PA) the first block and the mean amplitude of PA (116 dB) and prepulse (PP) (86 dB) trials and percent prepulse inhibition (PPI) as well as habituation measures of PA and PP trials at baseline and after 4 and 8 weeks n = 18

Baseline

Week 4

Week 8

Fsessiona

df/dferra

pa

Initial pulse-alone trial (mV) First block, amplitude of pulse-alone trials (mV) Mean amplitude of pulse-alone trials (mV) First block, amplitude of prepulse trials (mV) Mean amplitude of prepulse trials (mV) First block, percent prepulse inhibition Mean percent prepulse inhibition Percent habituation of pulse-alone trials (between first and last block) Percent habituation of prepulse trials (between first and last block) Habituation of pulse-alone trials across six blocks (linear gradient coefficient b) Habituation of prepulse trials across six blocks (linear gradient coefficient b)

202.7 167.9 102.5 72.0 50.5 60.6 52.7 46.1

196.7 149.8 108.9 70.7 59.7 52.0 44.8 35.3

202.8 116.1 89.2 75.1 58.1 40.8 38.3 36.3

(29.3) (17.6) (15.5) (19.0) (14.4) (9.85) (8.91) (5.67)

0.04 6.40 2.76 0.05 1.05 2.64 3.89 1.47

2/34 2/34 2/34 2/34 2/34 2/34 2/34 2/34

Ns .007 .089 Ns Ns .093 .047 Ns

(25.1) (24.2) (15.2) (15.4) (9.5) (4.25) (3.94) (6.26)

(25.1) (21.6) (16.9) (12.8) (11.6) (5.69) (6.00) (6.35)

23.6 (10.9)

16.7 (10.9)

17.6 (8.83)

0.17

2/34

Ns

39.4 (9.00)

25.4 (5.72)

19.1 (5.54)

4.05

2/34

.049

0.38

2/34

Ns

14.0 (5.66)

9.97 (4.96)

8.93 (3.22)

a

ANOVA with repeated measures at factor session, adjusted by means of the Greenhouse-Geisser correction. The original degrees of freedom and the adjusted p-values are reported.

decrement of PPI between the test sessions. The interaction of factors block and session was not significant. An analysis of the contrasts showed significant linear trends for the decrement of PPI across blocks and sessions, F block(1, 17) = 4.45, p < .05; F session(1, 17) = 6.11, p < .05. Smokers and non-smokers as well as females and males did not significantly differ in PPI at the first test session or across three test sessions. The short and long-term attenuation of PPI was still significant even after exclusion of smokers or women. 3.2. Startle reactivity and habituation Table 1 shows several measures for startle reactivity and habituation at baseline, week 4, and week 8. There was a significant decrease in the mean amplitude of the first block of PA trials reflecting long-term habituation of ASR. In addition, the linear gradient coefficient b of PA trials (representing the negative linear slope of within-session habituation which was calculated across 6 blocks of each session) significantly increased between the test sessions, reflecting the progressive flattening of the habituation curve and, therefore, also long-term habituation. Percent habituation of PA trials decreased also between test sessions but, overall, this trend was not significant (for statistics, see Table 1). The mean amplitude of PA trials showed a slight but not significant increase between baseline and week 4, t(17) = 0.62, p = .54. However, at week 8, the mean amplitude of PA trials showed a weak trend for a decrease compared to baseline, t(17) = 1.69, p = .11, and a significant decrease compared to week 4, t(17) = 2.71, p < .05. Analysis of the contrasts showed significant linear trends for the decrement of mean amplitude of the first block of PA trials and for the increase of b of PA trials across sessions, F 1,block(1, 17) = 9.56, p < .01; F b(1, 17) = 4.63, p < .05.

The mean amplitude of the first block of PP trials, the mean amplitude of all PP trials, the percent habituation of PP trials, and b of PP trials did not show significant changes between the test sessions (for statistics, see Table 1). Fig. 1 show the habituation curves of PA and PP trials diagrammed as the mean amplitude of the trials in six blocks at baseline, week 4, and week 8, respectively. A 6  3  2 (block  session  pulse) ANOVA of the PA and PP trials revealed significant main effects of the factors block, F(5, 85) = 20.1, p < .001, reflecting shortterm habituation, and pulse, F(1, 17) = 28.9, p < .001, reflecting the difference in startle magnitude between PP and PA trials. The factor session was not significant. The were also significant interactions of the factor pulse with the factors block and session, respectively, F block(5, 85) = 11.38, p < .001; F session(2, 34) = 4.99, p < .05, and a significant three-way interaction of factors block, session, and pulse, F(10, 170) = 1.95, p < .05, reflecting a significantly different progression of PA and PP trials within and between sessions. A 6  3 (block  session) ANOVA of blink magnitudes from PA trials revealed a significant main effect of the factor block, F(5, 85) = 22.68, p < .001, reflecting short-term habituation of ASR, and a strong trend for a main effect of the factor session, F(2, 34) = 2.86, p = .08, reflecting longterm habituation of ASR. The interaction of factors block and session was also significant, F(10, 170) = 3.10, p < .01, indicating that short-term habituation decreases between sessions which is the equivalent of long-term habituation. A 6  3 (block  session) ANOVA of the PP trials revealed only a significant main effect of the factor block, F(5, 85) = 7.27, p < .01, reflecting a decrease in the mean startle amplitude of PP trials within the sessions. There was no significant main effect of factor session and no significant interaction.

260

B.B. Quednow et al. / Biological Psychology 71 (2006) 256–263

Fig. 1. Mean blink startle magnitude on 116-dB-pulse-alone (PA) and 86-dB-prepulse (PP) trials across six blocks of trials at baseline and after 4 and 8 weeks (means and S.E. of means).

3.3. Relationship of PPI and startle reactivity Startle reactivity (first block of PA trials/mean amplitude of PA trials) and mean percent PPI was not correlated in any test session (baseline: r = .17/ .20, week 4: r = .02/.02; week 8: r = .22/ .13), thus poor startle responders did not significantly differ in percent PPI from good responders. Mean percent PPI and habituation of startle reactivity (b) was not correlated within the single test sessions (baseline: r = .10, week 4: r = .18; week 8: r = .17). The short-term habituation of startle reactivity and the attenuation of percent PPI within sessions (differences scores block 1–6) were significantly correlated only in week 8 (baseline: r = .10, week 4: r = .22; week 8: r = .64; p < .01). The change of startle reactivity between baseline and week 8 did not correlate with the change of percent PPI within 8 weeks (difference scores: baseline week 8: r = .04). This means that long-term habituation of startle reactivity and attenuation of percent PPI between sessions were independent of each other. In addition, a 2  3 (PPI versus

PA  session) ANOVA with startle reactivity (first block of PA trials) and percent PPI across test sessions revealed an significant interaction of both factors, F(2, 34) = 5.15, p < .01, reflecting the different progression of percent PPI and PA trials across sessions. This interaction did not occur, if this analysis was repeated with percent PPI versus PP trials. 3.4. Test-retest reliability of startle measures Startle reactivity on PA and PP trials (mean amplitude and first block) was highly reliable with ICCs from .82 to .95. Habituation of PA trials (percent habituation and b) was reliable with ICCs .61 and .71, respectively, whereas habituation of PP trials was not reliable. Mean PPI was also highly reliable (ICC = .87), indicating that the attenuation of PPI over time was a consistent phenomenon (Table 2). In fact 72.2% of our participants (13 of 18) have shown an overall attenuation of PPI over time (linear gradient coefficient b of PPI across three test sessions was < 0).

Table 2 Intraclass correlations of startle measures between three test sessions n = 18

ICC (3,3)

F

df/dferr

p

Initial pulse-alone trial (mV) First block, amplitude of pulse-alone trials (mV) Mean amplitude of pulse-alone trials (mV) First block, amplitude of prepulse trials (mV) Mean amplitude of prepulse trials (mV) First block, percent prepulse inhibition Percent prepulse inhibition Percent habituation of pulse-alone trials (between first and last block) Percent habituation of prepulse trials (between first and last block) Habituation of pulse-alone trials across six blocks (linear gradient coefficient b) Habituation of prepulse trials across six blocks (linear gradient coefficient b)

.76 .91 .95 .82 .94 .49 .87 .61 .47 .71 .33

4.18 10.60 18.64 5.56 16.95 1.96 7.87 2.55 1.89 3.44 1.48

17/34 17/34 17/34 17/34 17/34 17/34 17/34 17/34 17/34 17/34 17/34

<.001 <.0001 <.0001 <.0001 <.0001 <.05 <.0001 <.01 .056 <.001 Ns

B.B. Quednow et al. / Biological Psychology 71 (2006) 256–263

4. Discussion The object of this experiment was to investigate the shortand long-term stability of percent PPI of ASR in healthy volunteers. Analysis revealed a strong and significant decrease of PPI within and between three test sessions across a period of 8 weeks, which was independent of habituation of startle reactivity. Nevertheless, PPI was reliable across the test sessions, indicating that the significant attenuation of PPI over time was a consistent phenomenon. In addition, we found short- and long-term habituation of the general startle reactivity in PA trials; however, short-term habituation of the startle reaction in PP trials was found to be weaker (but still significant) whereas long-term habituation in PP trials was absent. The significant interaction between PA and PP trials within and between sessions confirmed the different progression of startle reactivity in PA and PP trials (e.g. strong long-habituation of the startle amplitude of PA trials in the first block between the test sessions and simultaneously no alteration of the startle amplitude of PP trials in the first block between the test sessions) which results in an attenuation of percent PPI across time. Furthermore, we could not demonstrate a significant interrelation between percent PPI and startle reactivity or between attenuation of PPI and habituation of startle reactivity. Thus, the habituation of startle reactivity alone does not explain the attenuation effect of PPI; this effect occurs only in combination with the lower habituation of the PP trials. But why does the habituation of responses on PP and PA trials differ? At first we speculated that the lower reduction in startle magnitude on PP trials reflects a ‘floor effect’, that is, either absolute (i.e. the startle magnitude could not be lower) or relative (i.e. magnitude in the lower range was more resistant to reduction relative to magnitude in the higher range elicited by PA trials; Swerdlow et al., 2000). We tested this assumption by comparing participants who have high versus low startle magnitude on PP trials, but there was no indication that changes in PP trials were more evident among participants who have higher prepulse levels. Thus, a ‘floor effect’ within PP trials seems to be unlikely as an explanation of our results. In addition, we could exclude that gender, smoking or age confounded our results. The decrease of PPI within session reported here is in line with previous studies (Lipp et al., 1994; Blumenthal, 1996; Lipp and Krinitzky, 1998; Lipp and Siddle, 1998), but our results suggest that percent PPI decreases also across repeated test sessions in contrast to some previous findings (Schwarzkopf et al., 1993; Abel et al., 1998; Cadenhead et al., 1999). In a previous study with rats, PPI of ASR has also been shown to attenuate within session if prepulses close to the detection level were used (Gewirtz and Davis, 1995). Also in line with previous studies, we found a high reliability of PPI in humans (Schwarzkopf et al., 1993; Abel et al., 1998; Cadenhead et al., 1999; Flaten, 2002). Thus, as shown here, percent PPI is in fact a reliable, but not a stable measure, comparable to startle reactivity which is also

261

highly reliable, but not stable due to habituation. This attenuation effect of PPI could be a confounding factor in startle measurement if repeated measures are used. Furthermore, we were able to replicate the findings of short- and long-term habituation of ASR (Ornitz and Guthrie, 1989). Cadenhead et al. (1999) used a rather similar design to explore reliability and stability of PPI in healthy male subjects. They reported a high stability and reliability of PPI across three test sessions spaced one month apart. The two major differences to our design are that we investigated female and male subjects, and that we used a startle sequence with more trials of a single PP condition. Since we found no differences in PPI and attenuation of PPI between genders, we attribute the varying results to the differences in the startle sequence. Due to the augmentation of trials in our study we found much less variance than Cadenhead et al. (1999) in PPI: e.g. the standard deviations of mean PPI for the 86 dB/140 ms SOA condition at baseline were 16.7 compared to approximately 30 for the comparable 86 dB/120 ms SOA condition, respectively. Thus, the improved signal-to-noise-ratio with regard to PPI in our study could have revealed the reported attenuation effect of PPI. Blumenthal (1999) suggested that the decrease of PPI may be due to one of two mechanisms: ‘‘(1) habituation of either the sensory input to the lead stimulus inhibition center or the neural signal projecting from that inhibition center to the startle center or (2) a decrease in the magnitude of the startle response itself (p. 61)’’. Whereas some findings supported the view that the extent to which a lead stimulus can inhibit startle might be partially determined by the magnitude of the startle response itself (Blumenthal, 1996; Lipp and Krinitzky, 1998), our results rather support the notion that PPI could habituate independent of habituation of startle reactivity, because we did not find a consistent relationship between PPI and startle reactivity or between attenuation of PPI and habituation of startle reactivity. However, when PPI ‘‘habituates’’, this is another affirmation for the assumption that PPI is not an strong automatic process and, therefore, not ‘‘hardwired’’ as supposed by Graham et al. (1975). Previous studies had already shown that selective attention influences the extent of PPI (Filion et al., 1993, 1994; Schell et al., 1995), which is also not compatible with the view of PPI as stable automatic process. With respect to the categorization of perceptual processing acts by Kahneman and Treisman (1984) into those that are strongly automatic, partially automatic, and occasionally automatic, Schell et al. (2000) concluded from their own data that PPI is a partially automatic process. They have shown that the demonstrated attentional modulation of PPI at a lead interval of 120 ms involves controlled processes. In contrast, at a lead interval of 60 ms, PPI seemed to be a strongly automatic process. Therefore, the authors suggested that PPI could be processed automatically, but could also be altered by controlled attentional modulation (Schell et al.,

262

B.B. Quednow et al. / Biological Psychology 71 (2006) 256–263

2000). Thus, if PPI is not a strongly automatic process, it may also be influenced by simple forms of learning like habituation. In our study we used only a single lead interval condition with an SOA of 140 ms for the prepulse trials, because maximal PPI could be expected at a SOA range of approximately 100–150 ms (Blumenthal, 1999). In the schizophrenia literature, prepulses with SOAs of 30 and 60 ms have consistently revealed the strongest deficits in PPI compared to normal controls (for review see Braff et al., 2001). Thus, it would be interesting for further research to investigate whether or not the short- and long-term reductions of PPI reported here are also found if shorter SOAs are used. The result of the present study, namely that PPI attenuates independent of habituation of the startle reflex, is consistent with previous studies with rodents. These did not find a relationship between PPI and habituation of startle reactivity even after extended presentation of startle stimuli (Cory and Ison, 1979; Hoffman et al., 1984). Blumenthal et al. (2004) compared several methods used to quantify PPI of ASR. The study revealed that the ‘‘proportion of the difference from control’’ (differences between reactivity on prepulse and control trials, divided by that on control trials) was the method least affected by startle reactivity and was therefore recommended as the preferred method to use when quantifying PPI. Proportion of the difference is equivalent to what we called percent PPI. Thus, the findings of Blumenthal et al. (2004) support our results that habituation of startle reactivity is not responsible for the attenuation of percent PPI, because percent PPI is largely independent of startle reactivity. What are the implications of these new findings for future studies? In clinical research an important question is if the well-known PPI deficits of schizophrenic patients are normalized by antipsychotic drugs (Swerdlow et al., 1994; Swerdlow and Geyer, 1998). The results of previous studies, which have predominately used cross-sectional designs, were contradictory (Hamm et al., 2001; Kumari and Sharma, 2002). Until now, only two studies have implemented longitudinal designs to explore the effect of antipsychotics on PPI in schizophrenics and both found that impaired gating persists despite symptomatic improvements associated with antipsychotic medication (Mackeprang et al., 2002; Duncan et al., 2003). However, the control groups of healthy volunteers in both studies were not tested repeatedly. In a pre-post design with schizophrenic patients, the potentiation of PPI as a result of the antipsychotic treatment and the time-dependent decrease of PPI would be antidromic, so that a treatment effect could possibly be masked because of the attenuation of PPI over time reported here. In order to control this attenuation of PPI in future studies that involve a pharmacological startle modification and repeated measurements in humans, our data suggests the importance of including a placebo and a healthy control group.

Furthermore, the current finding indicates that care should be taken when examining patients and healthy volunteers, which have been previously tested for PPI. Repeated startle testing of patients or controls could lead to a pseudo-PPIdeficit in these participants. In this study, we tested the stability and reliability of percent PPI in healthy subjects within and between three test sessions spaced 4 weeks apart. In contrast to previous studies, we found a significant long-term attenuation of PPI, which could have some implications for future clinical studies. Furthermore, we confirmed previous findings on a decrease of PPI within sessions. The present results suggest that the inhibitory mechanism underlying PPI may habituate and that PPI may therefore reflect only a partially automatic process.

References Abel, K., Waikar, M., Pedro, B., Hemsley, D., Geyer, M., 1998. Repeated testing of prepulse inhibition and habituation of the startle reflex: a study in healthy human controls. Journal of Psychopharmacology 12, 330– 337. Blumenthal, T.D., 1996. Inhibition of the human startle response is affected by both prepulse intensity and eliciting stimulus intensity. Biological Psychology 44, 85–104. Blumenthal, T.D., 1997. Prepulse inhibition decreases as startle reactivity habituates. Psychophysiology 34, 446–450. Blumenthal, T.D., 1999. Short lead interval startle modification. In: Dawson, M.E., Schell, A.M., Bo¨hmelt, A.H. (Eds.), Startle Modification. Cambridge University Press, Cambridge, UK, pp. 51–71. Blumenthal, T.D., Elden, A., Flaten, M.A., 2004. A comparison of several methods used to quantify prepulse inhibition of eyeblink responding. Psychophysiology 41, 326–332. Braff, D.L., Geyer, M.A., Swerdlow, N.R., 2001. Human studies of prepulse inhibition of startle: normal subjects, patient groups, and pharmacological studies. Psychopharmacology 156, 234–258. Braff, D.L., Grillon, C., Geyer, M.A., 1992. Gating and habituation of the startle reflex in schizophrenic patients. Archives of General Psychiatry 49, 206–215. Cadenhead, K.S., Carasso, B.S., Swerdlow, N.R., Geyer, M.A., Braff, D.L., 1999. Prepulse inhibition and habituation of the startle response are stable neurobiological measures in a normal male population. Biological Psychiatry 45, 360–364. Cadenhead, K.S., Geyer, M.A., Braff, D.L., 1993. Impaired startle prepulse inhibition and habituation in patients with schizotypal personality disorder. American Journal of Psychiatry 150, 1862–1867. Cory, R.N., Ison, J.R., 1979. Persistent effect of noise on the acoustic startle reflex in rats. Animal Learning & Behavior 7, 367–371. Davis, M.T., Parisi, D.S., Gendelman, M., Tischler, J.H., 1982. Habituation and sensitization of startle reflexes elicited electrically from the brainstem. Science 218, 688–690. Duncan, E.J., Szilagyi, S., Efferen, T.R., Schwartz, M.P., Parwani, A., Chakravorty, S., Madonick, S.H., Kunzova, A., Harmon, J.W., Angrist, B., Gonzenbach, S., Rotrosen, J.P., 2003. Effect of treatment status on prepulse inhibition of acoustic startle in schizophrenia. Psychopharmacology 167, 63–71. Filion, D.L., Dawson, M.E., Schell, A.M., 1993. Modification of the acoustic startle-reflex eyeblink: a tool for investigating early and late attentional processes. Biological Psychology 35, 185–200. Filion, D.L., Dawson, M.E., Schell, A.M., 1994. Probing the orienting response with startle modification and secondary reaction time. Psychophysiology 31, 68–78.

B.B. Quednow et al. / Biological Psychology 71 (2006) 256–263 Flaten, M.A., 2002. Test-retest reliability of the somatosensory blink reflex and its inhibition. International Journal of Psychophysiology 45, 261– 265. Geyer, M.A., Braff, D.L., 1982. Habituation of the Blink reflex in normals and schizophrenic patients. Psychophysiology 19, 1–6. Geyer, M.A., Braff, D.L., 1987. Startle habituation and sensorimotor gating in schizophrenia and related animal models. Schizophrenia Bulletin 13, 643–668. Gewirtz, J.C., Davis, M., 1995. Habituation of prepulse inhibition of the startle reflex using an auditory prepulse close to background noise. Behavioral Neuroscience 109, 388–395. Graham, F.K., 1975. The more or less startling effects of weak prestimulation. Psychophysiology 12, 238–248. Graham, F.K., Hackley, S.A., 1991. Passive and active attention to input. In: Jennings, J.R., Coles, M.G.H. (Eds.), Handbook of Cognitive Psychophysiology. Wiley, New York, pp. 251–356. Graham, F.K., Putnam, L.E., Leavitt, L.A., 1975. Lead-stimulation effects of human cardiac orienting and blink reflexes. Journal of Experimental Psychology: Human Perception & Performance 104, 175–182. Groves, P.M., Thompson, R.F., 1970. Habituation: a dual-process theory. Psychological Review 77, 419–450. Hamm, A.O., Weike, A.I., Schupp, H.T., 2001. The effect of neuroleptic medication on prepulse inhibition in schizophrenia patients: current status and future issues. Psychopharmacology 156, 259–265. Hoffman, H.S., Cohen, M.E., Corso, C., 1984. Reflex modification during habituation of the startle response. Bulletin of the Psychonomic Society 22, 574–576. Hoffman, H.S., Ison, J.R., 1980. Reflex modification in the domain of startle. I. Some empirical findings and their implications for how the nervous system processes sensory input. Psychological Review 87, 175–189. Kahneman, D., Treisman, A., 1984. Changing views of attention and automaticity. In: Parasuraman, R., Davies, D.R. (Eds.), Varieties of Attention. Academic Press, New York, pp. 29–61. Koch, M., 1999. The neurobiology of startle. Progress in Neurobiology 59, 107–128. Kumari, V., Sharma, T., 2002. Effects of typical and atypical antipsychotics on prepulse inhibition in schizophrenia: a critical evaluation of current evidence and directions for future research. Psychopharmacology 162, 97–101. Lipp, O.V., Arnold, S.L., Siddle, D.A., Dawson, M.E., 1994. The effect of repeated prepulse-blink reflex trials on blink reflex modulation at short lead intervals. Biological Psychology 38, 19–36. Lipp, O.V., Krinitzky, S.P., 1998. The effect of repeated prepulse and reflex stimulus presentations on startle prepulse inhibition. Biological Psychology 47, 65–76. Lipp, O.V., Siddle, D.A., 1998. The effects of prepulse-blink reflex trial repetition and prepulse change on blink reflex modification at short and long lead intervals. Biological Psychology 47, 45–63.

263

Mackeprang, T., Kristiansen, K.T., Glenthoj, B.Y., 2002. Effects of antipsychotics on prepulse inhibition of the startle response in drug-naive schizophrenic patients. Biological Psychiatry 52, 863–873. McGraw, J.O., Wong, S.P., 1996. Forming inferences about some intraclass correlation coefficients. Psychological Methods 1, 30–46. Ornitz, E.M., Guthrie, D., 1989. Long-term habituation and sensitization of the acoustic startle response in the normal adult human. Psychophysiology 26, 166–173. Parwani, A., Duncan, E.J., Bartlett, E., Madonick, S.H., Efferen, T.R., Rajan, R., Sanfilipo, M., Chappell, P.B., Chakravorty, S., Gonzenbach, S., Ko, G.N., Rotrosen, J.P., 2000. Impaired prepulse inhibition of acoustic startle in schizophrenia. Biological Psychiatry 47, 662– 669. Schell, A.M., Dawson, M.E., Hazlett, E.A., Filion, D.L., 1995. Attentional modulation of startle in psychosis-prone college students. Psychophysiology 32, 266–273. Schell, A.M., Wynn, J.K., Dawson, M.E., Sinaii, N., Niebala, C.B., 2000. Automatic and controlled attentional processes in startle eyeblink modification: effects of habituation of the prepulse. Psychophysiology 37, 409–417. Schwarzkopf, S.B., McCoy, L., Smith, D.A., Boutros, N.N., 1993. Testretest reliability of prepulse inhibition of the acoustic startle response. Biological Psychiatry 34, 896–900. Siddle, D.A., Kroese, B.S., 1985. Orienting, habituation, and short-term memory. Psychophysiology 22, 535–544. Swerdlow, N.R., Benbow, C.H., Zisook, S., Geyer, M.A., Braff, D.L., 1993. A preliminary assessment of sensorimotor gating in patients with obsessive compulsive disorder. Biological Psychiatry 33, 298–301. Swerdlow, N.R., Braff, D.L., Geyer, M.A., 1999. Cross-species studies of sensorimotor gating of the startle reflex. Annals of the New York Academy of Sciences 877, 202–216. Swerdlow, N.R., Braff, D.L., Geyer, M.A., 2000. Animal models of deficient sensorimotor gating: what we know, what we think we know, and what we hope to know soon. Behavioural Pharmacology 11, 185–204. Swerdlow, N.R., Braff, D.L., Taaid, N., Geyer, M.A., 1994. Assessing the validity of an animal model of deficient sensorimotor gating in schizophrenic patients. Archives of General Psychiatry 51, 139–154. Swerdlow, N.R., Eastvold, A., Uyan, K.M., Ploum, Y., Cadenhead, K., 2001. Matching strategies for drug studies of prepulse inhibition in humans. Behavioural Pharmacology 12, 45–52. Swerdlow, N.R., Geyer, M.A., 1998. Using an animal model of deficient sensorimotor gating to study the pathophysiology and new treatments of schizophrenia. Schizophrenia Bulletin 24, 285–301. Swerdlow, N.R., Paulsen, J., Braff, D.L., Butters, N., Geyer, M.A., Swenson, M.R., 1995. Impaired prepulse inhibition of acoustic and tactile startle response in patients with Huntington’s disease. Journal of Neurology, Neurosurgery, and Psychiatry 58, 192–200. Winer, B.J., 1971. Statistical Principles in Experimental Design. McGrawHill, Kogakusha, Tokyo.