Effects of Background and Prepulse Characteristics on Prepulse Inhibition and Facilitation: Implications for Neuropsychiatric Research

Effects of Background and Prepulse Characteristics on Prepulse Inhibition and Facilitation: Implications for Neuropsychiatric Research Ming H. Hsieh, Neal R. Swerdlow, and David L. Braff Background: Both prepulse inhibition (PPI) and prepulse facilitation (PPF) deficits have been reported in schizophrenia patients, but the use of different experimental parameters across laboratories makes direct comparisons of results difficult. We assessed the effects of different parameters on PPI and PPF in normal subjects. Methods: Eyeblink startle was measured in 14 healthy male subjects, using 115 dB[A] white noise startle pulses and 86 dB[A] prepulses. Analyses compared the effects of: 1) background noise level (ambient 54 vs. 70 dB[A]) on PPI and PPF, 2) prepulse duration (discrete 20 msec vs. continuous) on PPF, 3) prepulse frequency (1000 Hz vs. white noise) on PPI and PPF, and 4) prepulse interval (2000 vs. 4500 msec) on PPF. Results: Compared to an experimentally delivered 70 dB[A] background, ambient 54 dB[A] background led to significantly more PPI (with discrete white noise prepulses), and more PPF (with continuous prepulses). Continuous and longer (4500 msec) prepulses induced more PPF than did discrete and shorter (2000 msec) prepulses. Conclusions: Paradigmatic differences appear likely to be responsible for divergent findings in studies of PPI and PPF in normal and schizophrenia subjects. The present study should guide investigators in the selection of parameters for assessing PPI and PPF in studies of normal subjects and schizophrenia patients. Attention to the 4 factors of 1) background noise, 2) prepulse duration, 3) frequency, and 4) interval will facilitate comparability of results across different laboratories, especially when using PPI/PPF in schizophrenia research as neural substrate probes, as biomarkers, and as endophenotypes. Key Words: Background, continuous prepulse, discrete prepulse, prepulse facilitation, prepulse inhibition

T

he startle reflex is a whole body response to an intense sudden stimulus. Startle is modified in normal human subjects when the startle-eliciting stimulus is preceded by a weak stimulus called a “prepulse” (Graham 1975; Hoffman and Fleshler 1963). Prepulse inhibition (PPI) is the reduction in startle magnitude that occurs when the prepulse interval (the interval between the onset of the prepulse and the onset of the startleeliciting stimulus or interstimulus interval [ISI]) is between 30 and 500 milliseconds. Prepulse facilitation (PPF) is the augmentation of startle magnitude that occurs when the ISI is very short (less than 20 milliseconds) or relatively long (e.g., greater than 1000 milliseconds) (e.g., Graham 1975). There is a long history of differing effects of four paradigmatic manipulations in PPI/PPF studies (background noise levels, prepulse duration, prepulse frequency, prepulse to startle stimulus duration) on both PPI and PPF (Wynn et al 2000; Graham 1975; Putnam and Vanman 1999; Hoffman and Wible 1969; Graham et al 1975). Still, there are many issues in the PPI/PPF literature that need resolution, and no study has examined these four paradigmatic manipulations simultaneously in the same cohort of normal subjects in the manner accomplished in this study. Deficits in PPI (e.g., Braff et al 1978; Bolino et al 1994; Kumari et al 1999) and more recently PPF (e.g., Ludewig et al 2003; Wynn et al 2004) have been repeatedly identified in schizophre-

From the Department of Psychiatry (MHH), National Taiwan University Hospital, Taipei, Taiwan; and Department of Psychiatry (MHH, NRS, DLB), University of California, San Diego, La Jolla, California. Address reprint requests to David L. Braff, M.D., University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0804; E-mail: dbraff@ ucsd.edu Received December 6, 2004; revised March 21, 2005; revised July 11, 2005; accepted July 21, 2005.

0006-3223/06/$32.00 doi:10.1016/j.biopsych.2005.07.032

nia patients and in their unaffected relatives (Cadenhead et al 2000; Wynn et al 2004). However, there have also been reports of normal levels of PPI (Wynn et al 2004) and PPF (Ford et al 1999; Kumari et al 2004) in schizophrenia patients utilizing a typical neutral or uninstructed paradigm. Understanding the basis for these divergent findings is critically important, given the demonstrated utility of PPI and PPF as biomarkers and possible endophenotypes (Braff and Freedman 2002) for schizophrenia. The current study utilized normal subjects in a typical neutral or uninstructed paradigm to begin the translation of these findings to the ambiguities found in the PPI and PPF literature in normal and schizophrenia subjects. One likely explanation for the different findings across these studies is that they differed in the types of background noise and stimuli used to elicit PPI and PPF. Differences in parameters such as background noise might generate different levels of PPI and PPF that fall within different portions of the dynamic ranges of these processes relative to “floor” and “ceiling” values. To clarify these issues, this study was conducted in a cohort of normal subjects. There are already studies of some of these issues. For example, maximizing event or temporal uncertainty increases PPF in “long lead” startle modification paradigms, whereas creating more certainty has the opposite effect (Graham et al 1975). It is not fully informative to compare absolute levels of PPI across published studies performed in different laboratories due to differences in equipment, session design, testing environment, etc. In the present study, within one laboratory, we examined the impact on PPI and PPF of four specific paradigmatic characteristics that differed across some of these divergent reports. This is a study of normal subjects and we do not address issues such as medication, symptom influences, or instructional set (i.e., attend to vs. ignore the prepulse) on PPI or PPF. The major aims of this study were to assess in normal subjects the relative contributions of 1) ambient 54 dB[A] versus 70 dB[A] background noise in assessing PPI and PPF; 2) the effects of white noise versus 1000 Hz tone prepulses; 3) the effects of short and long (2000 and 4500 millisecond) ISIs; and 4) the effects of continuous versus BIOL PSYCHIATRY 2006;59:555–559 © 2005 Society of Biological Psychiatry

556 BIOL PSYCHIATRY 2006;59:555–559

M.H. Hsieh et al

Table 1. Experimental Design: Four Different Subsessions Continuous Prepulses Subsessions A B C D

Discrete Prepulses (20 milliseconds in Duration)

Background

2000 milliseconds

4500 milliseconds

2000 milliseconds

4500 milliseconds

120 milliseconds

70 dB[A] white noise Ambient background 70 dB[A] white noise Ambient background

T2000C T2000C W2000C W2000C

T4500C T4500C W4500C W4500C

T2000D T2000D W2000D W2000D

T4500D T4500D W4500D W4500D

T120D T120D W120D W120D

All prepulses are 86 dB[A], while all pulses are 40 milliseconds 115 dB[A] white noise. T/W, tone/white noise prepulses; C/D, continuous/discrete prepulses.

discrete prepulses in assessing PPF. The rationale of the study was that the results would facilitate a clearer understanding and guidelines regarding parametric influences for designing studies using neutral, uninstructed PPI and PPF paradigms in schizophrenia patients.

Methods and Materials Participants Nineteen healthy right-handed male subjects were recruited through a posted advertisement. Of these, 14 completed the study with valid data, as described below (20 –55 years old; mean age 34.1, SD ⫽ 12.8). Four subjects were excluded due to unusable/unscorable PPI/PPF data (e.g., the baseline pulse alone startle levels were too low to meet criteria), and one subject was excluded for a positive benzodiazepine toxicology screen after denying any use of this class of drugs. On the day of startle examination, all subjects were screened using the Structured Clinical Interview for DSM-IV, Nonpatient (SCID-NP) for Axis I Disorders and for Cluster A Axis II disorders, using the Structural Clinical Interview for DSM-IV Axis II Disorders (SCID-P) (First et al 1995, 1996). Subjects were excluded if they reported a history of major medical illnesses, substance abuse, past or present psychiatric disorders, head trauma, or hearing loss or a family history of mental illness, per our previously described methods (Braff et al 1999). On the day of testing, a hearing test was performed to exclude hearing impairment using a Saico SCR-2 Audiometer (Saico, Assens, Denmark) (ability to hear ⬎45 dB at 500, 1000, or 6000 Hz). Fully informed verbal and written consent was obtained (University of California San Diego [UCSD] Institutional Review Board [IRB] #031058). Urine toxicology screens were also done to confirm that subjects were free from current drug intoxication, and none of the samples were positive. Startle Response Measurement Each subject, wearing earphones, was seated comfortably in an upright position in a recliner chair in a sound-attenuated room (hence the low 54 dB[A] ambient noise levels). Acoustic startle stimuli were presented to subjects binaurally through headphones. Two miniature silver/silver chloride electrodes were positioned below and lateral to the each eye over the orbicularis oculi muscle; electrode resistances were less than 10 kOhm. A ground electrode was placed behind the left ear over the mastoid. The eyeblink component of the acoustic startle response was measured using an electromyographic (EMG) startle system (EMG-SR-LAB, San Diego Instruments, San Diego, California) for digitization and analysis. The system recorded 250 1-millisecond epochs, starting with the onset of the startle stimulus. Recorded EMG activity was band-pass filtered (100 to 1000 Hz). A 60-Hz notch filter was also used to eliminate 60-Hz interference. Sound levels were calibrated monthly by using Quest-2700 dB meter (Quest Electronics, Oconomowoc, Wisconwww.sobp.org/journal

sin) with a 6-cc coupler in an artificial ear. Ambient noise in the sound-attenuated room was measured using an artificial ear coupler on the headphones with observed readings of 54 dB measured by the Quest-2700 dB meter. Startle Session Design Each session contained four different subsessions (A–D as outlined in Table 1). These subsessions differed in their level of background noise (background: ambient 54 dB[A] vs. 70 dB[A]) and frequency characteristics of the prepulse (1000 Hz tone vs. white noise). Prepulse duration was either discrete (20 milliseconds, followed by a return to background level) versus continuous (2000 or 4500 milliseconds, continuing until the startle pulse onset). In all cases, the pulse-alone stimulus was a 40-millisecond presentation of 115 dB[A] white noise and all the prepulses were 86 dB[A] (Table 1). Each subsession began with a 3-minute acclimatization period followed by 36 startle stimuli divided into three blocks. Block 1 consisted of three pulse-alone trials. Block 2 included 6 pulse-alone trials, 10 continuous prepulse trials (5 each with 2000 and 4500 millisecond prepulse intervals), and 10 discrete prepulse trials (5 each with 2000 and 4500 millisecond prepulse intervals). Block 3 consisted of three pulse-alone trials and four discrete prepulse trials with a 120millisecond prepulse interval. Within each block, trials were presented in a pseudorandom order with a mean intertrial interval of 15 seconds (range 11–19 seconds). Prepulse trials with a 120-millisecond interval were expected to elicit PPI; all other prepulse trials were expected to elicit PPF. As reflected in Table 1, which is a clear representation of the study design, all subjects were tested for PPI and PPF with subsessions A through D presented in counterbalanced order. Each subsession lasted about 12 minutes. Data Processing All recordings were screened to exclude spontaneous eyeblink activity before data analysis. For each participant, all blink response variables were averaged for each trial type within a block. Baseline to peak startle magnitude was calculated in arbitrary digitalized units and peak latency was calculated in milliseconds for each trial in the 20- to 100-millisecond time window following startle stimulus onset. These values were calculated using commercially available software (SRRED, San Diego Instruments, San Diego, California). Nonresponsiveness was defined as having three or more pulse-alone amplitudes of 0 anywhere in the entire session. Using this criterion, four participants had to be excluded (24 – 47 years old; mean age: 37.0, SD ⫽ 10.0). The startle measures examined were as follows: 1. Startle reactivity: The magnitude of responses on pulsealone trials in Block 1, analyzed by repeated-measures

BIOL PSYCHIATRY 2006;59:555–559 557

M.H. Hsieh et al analysis of variance (ANOVA) with subsession (A–D) as a within-subject factor. 2. Habituation: The change in pulse-alone magnitude from Blocks 1 to 3, analyzed by repeated-measures ANOVA of pulse-alone magnitude with block as a within-subject factor. 3. Prepulse inhibition in the 120-millisecond interval condition of Block 3: %PPI was calculated for Block 3 as [1 (mean startle magnitude on prepulse trials/mean startle magnitude on pulse-alone trials) ⫻ 100]. Two subjects each exhibited no scorable startle responses on individual trials in Block 3; these total of two missing PPI values for the two subjects were replaced with mean values for the respective subsession. A repeated-measures ANOVA was performed on %PPI, with background as a within-subject factor. 4. Prepulse facilitation: %PPF was calculated for Block 2 as [1 ⫺ (mean startle magnitude on prepulse trials/mean startle magnitude on pulse alone trials) ⫻ 100]. A negative value indicates the presence of prepulse facilitation. For data analysis, a minimum of ⫺100% inhibition (or 100% facilitation) was used to prevent outliers from having a disproportionate effect on group means. A repeated-measures ANOVA was performed on %PPF, with frequency, background, and prepulse interval (2000 vs. 4500 milliseconds) as within-subject factors. Eye side (right vs. left) was initially included as a withinsubject factor in all measures but yielded no significant main effects or meaningful interactions, and data were then collapsed across eyes. In terms of analyzing the results, the GreenhouseGeisser correction might be recommended for all repeatedTable 2. Results of Repeated-Measures ANOVA for the Percentage of PPI and PPF Prepulse Characteristics PPI Frequency Background Frequency ⫻ Background PPF Frequency Background Interval Duration Frequency ⫻ Background Frequency ⫻ Interval Background ⫻ Interval Frequency ⫻ Background ⫻ Interval Frequency ⫻ Duration Background ⫻ Duration Frequency ⫻ Background ⫻ Duration Interval ⫻ Duration Frequency ⫻ Interval ⫻ Duration Background ⫻ Interval ⫻ Duration Frequency ⫻ Background ⫻ Interval ⫻ Duration

df

F

Figure 1. Effect of different subsessions on PPI. This figure illustrates the high levels of PPI that are elicited when the 86 dB[A] white noise prepulse “arises” from a lower level of ambient noise (54 dB[A]) versus a 70 dB[A] background (n ⫽ 14, **p ⬍ .01). Bars represent standard errors of the means. PPI, prepulse inhibition.

measure analyses to correct for inflated degrees of freedom regardless of whether the sphericity assumption was violated. But, the Greenhouse-Geisser correction is appropriate for use in mixed-model ANOVAs. The ANOVAs in this study, as described in Methods and Materials, are not mixed-model because there is no between-subject comparison. With an ANOVA that is entirely within-subject, all diagonal elements are precisely equal and thus epsilon equals 1.0. Because of this, degrees of freedom are not adjusted but rather are correct as currently listed in Table 2. For each prepulse variable, Fisher’s least significant difference post hoc analysis was also conducted with alpha ⫽ .05.

p

Results 1, 13 1, 13 1, 13

.15 3.45 6.86

.703 .086 .021a

Results are seen in Table 2 and Figures 1 through 4 and are discussed below.

1, 13 1, 13 1, 13 1, 13 1, 13 1, 13 1, 13 1, 13 1, 13 1, 13 1, 13 1, 13 1, 13 1, 13

2.67 1.62 19.62 19.62 1.36 .49 2.31 2.37 21.78 23.12 .61 2.21 .24 .82

.126 .226 .001b .001b .264 .495 .153 .147 ⬍.001b ⬍.001b .447 .161 .629 .381

Startle Reactivity There was a main effect of background noise in the initial level of startle reactivity [F (1,13) ⫽ 5.10, p ⫽ .042] with higher startle magnitudes in those subsessions with 70 dB[A] background (subsessions A and C) than those with only 54dB[A] ambient background (subsessions B and D).

1, 13

2.52

.136

Background: Ambient versus 70 dB[A] white noise background. Duration: Continuous versus discrete prepulses. Frequency: 1000 Hz tone prepulses versus white-noise prepulses. Interval: 2000 versus 4500 millisecond prepulses. ANOVA, analysis of variance; PPI, prepulse inhibition; PPF, prepulse facilitation. a p ⬍ .05. b p ⬍ .01.

Habituation As expected, there was a significant block effect [F (2,104) ⫽ 27.069, p ⬍ .001, ε ⫽ .760] with lower startle magnitudes in the last block than in earlier blocks, reflecting the phenomenon of significant within-session habituation. PPI Analysis of variance revealed a nonsignificant (p ⫽ .086) main effect of background [F (1,13) ⫽ 3.45] and a significant interaction of background ⫻ prepulse frequency [F (1,13) ⫽ 6.86, p ⫽ .021] (Figure 1). These findings reflected the fact that prepulses arising from lower level 54 dB[A] ambient background induced more PPI than did prepulses arising from a background of 70 dB[A] and that this effect was most evident with white noise versus 1000 Hz tone prepulses. www.sobp.org/journal

558 BIOL PSYCHIATRY 2006;59:555–559

Figure 2. Effect of different stimulus parameters on PPF (n ⫽ 14, **p ⬍ .01). This figure illustrates that PPF is enhanced when continuous versus discrete prepulses are used. Also, 4500-millisecond versus 2000-millisecond continuous prepulses are more effective in eliciting PPF. Negative percent scores indicate higher levels of PPF. Bars represent standard errors of the means. PPF, prepulse facilitation.

PPF Analysis of variance revealed that trials with continuous prepulses induced greater levels of PPF than did those with discrete prepulses [F (1,13) ⫽ 19.62, p ⫽ .001]. In addition, trials with 4500-millisecond prepulse intervals generated more PPF than did those with 2000-millisecond prepulse intervals [F(1,13) ⫽ 19.62, p ⫽ .001] (Figure 2). There was no main effect of background noise level, but there was a two-way interaction of background noise level ⫻ duration [F (1,13) ⫽ 23.12, p ⬍ .001], reflecting a PPF increase with ambient versus 70 dB[A] background that was evident with continuous but not discrete stimuli (Figure 3). Similarly, a two-way interaction of frequency ⫻ duration [F (1,13) ⫽ 21.78, p ⬍ .001] reflected greater PPF with

Figure 3. Interaction of background ⫻ prepulse duration (continuous versus discrete) on PPF. This figure shows that with ambient background, continuous prepulses produced more PPF than did discrete prepulses. Bars represent standard errors of the means. **p ⬍ .001. PPF, prepulse facilitation.

www.sobp.org/journal

M.H. Hsieh et al

Figure 4. Interaction of prepulse frequency (tone versus white noise) ⫻ duration (continuous versus discrete) on PPF. This figure shows that continuous tone prepulses produced more PPF than did continuous white noise prepulses. Bars represent standard errors of the means. *p ⬍ .05. PPF, prepulse facilitation.

1000 Hz tones versus white noise prepulses that was evident with continuous but not discrete stimuli (Figure 4). In addition, for the discrete condition at 2000 milliseconds, PPF was not significantly different from zero (one sample sign test, p ⫽ .424).

Discussion These findings demonstrate, within one test session in one laboratory, the clear differences in startle modulation produced by experimental conditions in normal subjects. These fundamental paradigmatic influences undoubtedly accounted for part of the variance in divergent reports of PPI and PPF deficits in normal and schizophrenia subjects in neutral, uninstructed PPI and PPF paradigms commonly used in schizophrenia research. Prepulse inhibition generated with white noise prepulses and a 54 dB[A] ambient background noise (subsession D) exceeds that generated with a 70 dB[A] background (subsession C) by nearly 28%. As predicted, this is not surprising: in the ambient condition, prepulse salience can be viewed as being 32 dB (86 54) versus 16 dB (86 - 70) in the 70 dB condition. It is possible that in studies using ambient noise comparing schizophrenia patients versus control subjects, “powerful” prepulses may have driven PPI levels to their physiological ceiling and eliminated group differences obtained with prepulses that were less intense relative to background noise levels (Braff et al 1992, 1978; Cadenhead et al 1993, 2000; Grillon et al 1992; Kumari et al 2000; Ludewig et al 2003). Prepulse facilitation is also significantly affected by background noise. As with PPI, PPF generated by continuous prepulses with a relatively low (54 dB) level of ambient background noise exceeds that generated with a 70 dB[A] background–in this case by a dramatic factor of 4 (see Figure 2). These findings support the earlier report by Graham (1975) of greater PPF with continuous versus discrete prepulses and of more pronounced PPF with longer versus shorter prepulse intervals (4500 vs. 2000 milliseconds in the present study). Thus, it may follow that the shorter 1000-millisecond prepulse intervals used by Kumari et al (2004) failed to discriminate PPF in control subjects versus schizophrenia patients due to a floor effect caused by relatively brief, weak, and therefore less salient prepulses, in addition to

M.H. Hsieh et al other obviously important factors (e.g., demographic factors, symptoms, medication). In terms of startle magnitude, Hoffman (1999) and Hoffman and Fleshler (1963) first demonstrated that contrary to its possible “masking” effect, background noise can increase rather than decrease acoustic startle magnitudes. The present findings replicate this outcome: Block 1 pulse-alone amplitude was higher in subsessions with 70 dB[A] background than those with 54 dB[A] ambient background noise. The present findings take a helpful first step in integrating and resolving some differences in published comparisons of prepulse characteristics on PPI in a neutral, uninstructed paradigm that have been conducted across different laboratories and different conditions including different levels of background noise. Wynn et al (2000) reported the influence of prepulse frequency and duration under ambient background conditions. They found that discrete white noise prepulses produced greater PPI than did continuous white noise, discrete tone, or continuous tone prepulses. Braff et al (2001a) examined the influence of prepulse characteristics on PPI utilizing a 70 dB[A] white noise background and found that discrete white noise prepulses were most effective in eliciting PPI deficits in schizophrenia patients versus normal subjects. The present study permitted a direct “head-to-head” comparison of prepulse effects on PPI and PPF, using variations of parameters that have been used by different laboratories that reached divergent conclusions in comparisons of schizophrenia versus control subjects. Of course, the variance in startle measures attributable to overt attentional manipulation (e.g., Dawson et al 1993), gender, medication, and symptom effects (cf. Braff et al 2001b) are also potentially important independent influences on PPI and PPF and merit further scrutiny. The examination of these factors is outside the purview of this article, but studies isolating the effects of background noise in the attend/ignore paradigm would be informative and extend the results of this study. From the data in this study, we conclude that stimulus characteristics that differed across previous reports (i.e., background noise, prepulse frequency, and duration) are a likely important source of a significant amount of variance in the cited divergent findings. The effects of stimulus characteristics examined in this report generate dramatically different levels of PPI and PPF that are within very different areas of the dynamic ranges of these physiological processes. Future studies with normal subjects and schizophrenia patients should carefully account for and specify the reported differing effects of paradigm design (e.g., background noise) to optimize and control study designs and allow for comparisons of different results. This approach will be especially useful when PPI and PPF are used as neural substrate probes in brain imaging studies, as biomarkers in clinical trials, and as endophenotypes in genetic studies. This work was supported in part by grants from the National Institute of Mental Health, MH042228 and MH065571 (the Consortium on the Genetics of Schizophrenia [COGS]), and the Department of Veteran Affairs, VISN 22 MIRECC (Mental Illness Research, Education and Clinical Center). NRS was supported by MH01436. Bolino F, DiMichele V, DiCicco L, Manna V, Daneluzzo E, Casacchia M (1994): Sensorimotor gating and habituation evoked by electro-cutaneous stimulation in schizophrenia. Biol Psychiatry 36:670 – 679.

BIOL PSYCHIATRY 2006;59:555–559 559 Braff DL, Freedman R (2002): Endophenotypes in studies of the genetics of schizophrenia. In: Davis KL, Charney DS, Coyle JT, Nemeroff C, editors. Neuropsychopharmacology: The Fifth Generation of Progress. Philadelphia: Lippincott Williams & Wilkins, 703–716. Braff DL, Geyer MA, Light GA, Sprock J, Perry W, Cadenhead K, et al (2001a): Impact of prepulse characteristics on the detection of sensorimotor gating deficits in schizophrenia. Schizophr Res 49:171–178. Braff DL, Geyer MA, Swerdlow NR (2001b): Human studies of prepulse inhibition of startle: Normal subjects, patient groups, and pharmacological studies. Psychopharmacology (Berl) 156:234 –258. Braff DL, Grillon C, Geyer MA (1992): Gating and habituation of the startle reflex in schizophrenic patients. Arch Gen Psychiatry 49:206 –215. Braff DL, Stone C, Callaway E, Geyer MA, Glick I, Bali L (1978): Prestimulus effects on human startle reflex in normals and schizophrenics. Psychophysiology 15:339 –343. Braff DL, Swerdlow NR, Geyer MA (1999): Symptom correlates of prepulse inhibition deficits in male schizophrenia patients. Am J Psychiatry 156: 596 – 602. Cadenhead KS, Geyer MA, Braff DL (1993): Impaired startle prepulse inhibition and habituation in patients with schizotypal personality disorder. Am J Psychiatry 150:1862–1867. Cadenhead KS, Light GA, Geyer MA, Braff DL (2000): Sensory gating deficits assessed by the P50 event-related potential in subjects with schizotypal personality disorder. Am J Psychiatry 157:55–59. Dawson ME, Hazlett EA, Filion DL, Nuechterlein KH, Schell AM (1993): Attention and schizophrenia: Impaired modulation of the startle reflex. J Abnormal Psych 102:633– 641. First MB, Spitzer RL, Gibbon M, Williams JBW (1995): Structured Clinical Interview for DSM-IV Axis II disorders, Patient Edition (SCID-P), Version 2. New York: New York State Psychiatric Institute, Biometrics Research. First MB, Spitzer RL, Gibbon M, Williams JBW (1996): Structured Clinical Interview for DSM-IV Axis I disorders, Non-Patient Edition (SCID-1/NP), Version 2. New York: New York State Psychiatric Institute, Biometrics Research. Ford JM, Roth WT, Menon V, Pfefferbaum A (1999): Failures of automatic and strategic processing in schizophrenia: Comparisons of event-related brain potential and startle blink modification. Schizophr Res 37:149 –163. Graham FK (1975): The more or less startling effects of weak prestimulation. Psychophysiology 12:238 –248. Graham FK, Putnam LE, Leavitt LA (1975): Lead stimulation effects on human cardiac orienting and blink reflexes. J Exp Psychol 104:161–169. Grillon C, Ameli R, Charney DS, Krystal J, Braff DL (1992): Startle gating deficits occur across prepulse intensities in schizophrenic patients. Biol Psychiatry 32:939 –943. Hoffman HS (1999): A historical note on the “discovery” of startle modification. In: Dawson ME, Schell AM, Boehmelt AH, editors. Startle Modification: Implications for Neuroscience, Cognitive Science, and Clinical Science. New York: Cambridge University Press, 1–5. Hoffman HS, Fleshler M (1963): Startle reaction: Modification by background acoustic stimulation. Science 141:928 –930. Hoffman HS, Wible BL (1969): Temporal parameters in startle facilitation by steady background signals. J Acoust Soc Am 45:7–12. Kumari V, Aasen I, Sharma T (2004): Sex differences in prepulse inhibition deficits in chronic schizophrenia. Schizophr Res 69:219 –235. Kumari V, Soni W, Mathew VM, Sharma T (2000): Prepulse inhibition of the startle response in men with schizophrenia: Effects of age of onset of illness, symptoms, and medication. Arch Gen Psychiatry 57:609 – 614. Kumari V, Soni W, Sharma T (1999): Normalization of information processing deficits in schizophrenia. Am J Psychiatry 156:1046 –1051. Ludewig K, Geyer MA, Vollenweider FX (2003): Deficits in prepulse inhibition and habituation in never-medicated, first-episode schizophrenia. Biol Psychiatry 54:121–128. Putnam L, Vanman E (1999): Long lead interval startle modification. In: Dawson ME, Schell AM, Boehmelt AH, editors. Startle Modification: Implications for Neuroscience, Cognitive Science, and Clinical Science. New York: Cambridge University Press, 72–92. Wynn JK, Dawson ME, Schell AM (2000): Discrete and continuous prepulses have differential effects on startle prepulse inhibition and skin conductance orienting. Psychophysiology 37:224 –230. Wynn JK, Dawson ME, Schell AM, McGee M, Salveson D, Green MF (2004): Prepulse facilitation and prepulse inhibition in schizophrenia patients and their unaffected siblings. Biol Psychiatry 55:518 –523.

www.sobp.org/journal