Prepulse Facilitation and Prepulse Inhibition in Schizophrenia Patients and Their Unaffected Siblings Jonathan K

Prepulse Facilitation and Prepulse Inhibition in Schizophrenia Patients and Their Unaffected Siblings Jonathan K. Wynn, Michael E. Dawson, Anne M. Schell, Mark McGee, Dustin Salveson, and Michael F. Green Background: Deficits in schizophrenia patients and their first-degree relatives have been reported in prepulse inhibition (PPI), a phenomenon that measures an early stage of information processing (sensorimotor gating). It is less clear whether these information processing deficits extend to prepulse facilitation (PPF), which measures a later stage of generalized alerting or orienting. Methods: This study examined three separate issues: first, whether schizophrenia patients have deficits in PPI and PPF; second, whether the siblings of patients show deficits in these processes; and third, whether prepulse duration influences the degree of the deficits. These issues were examined in 76 schizophrenia patients, 36 of their siblings, and 41 normal control subjects. Results: Patients and siblings did not differ from control subjects in PPI, perhaps due to the use of different procedural parameters compared with other laboratories that have consistently found PPI deficits in schizophrenia patients. Patients and their siblings produced significantly less PPF than control subjects. For both PPI and PPF, prepulse duration was not a significant factor. Conclusions: These results imply that PPF deficits reveal a generalized alerting or orienting deficit that is present in both schizophrenia patients and their siblings, suggesting that this deficit may be tapping an endophenotypic vulnerability factor.

Key Words: Startle, prepulse inhibition, prepulse facilitation, ori- PPF in a passive attention paradigm is thought to reflect a classical activating effect (Graham 1975), indicating alerting or enting, schizophrenia, endophenotypic marker attention or automatically elicited generalized orienting (as op- tartle eyeblink modification (SEM) is a powerful tool for posed to specific orienting during a task that requires attention to be paid to specific stimuli; Graham and Hackley 1991). Prepulses studying information processing by examining how a star- appear to cause subjects to orient to incoming information, thus tle eyeblink is modified by a preceding, nonstartling stim- S enhancing the subsequent amount of eyeblink to the startle ulus (prepulse). When the prepulse precedes the startle pulse by burst, if the prepulse and startle burst are in the same modality a short interval (known as the lead interval), for example, less (Graham 1980). Deficits in orienting have been related to deficits than 500 msec, the startle eyeblink is inhibited, a phenomenon in allocation of processing resources to external stimuli and a known as prepulse inhibition (PPI). When the lead interval is reduction in available processing resources, traits also exhibited longer, for example, greater than 1000 msec, the startle eyeblink by schizophrenia patients (for a review, see Nuechterlein and is facilitated, particularly if the prepulse and startle stimulus are in Dawson 1984). Moreover, deficits in orienting have been shown the same sensory modality, a phenomenon known as prepulse to be stable trait markers for schizophrenia patients (Dawson et facilitation (PPF). Investigators have used SEM to study informa- al 1994; Spohn et al 1989) and at-risk populations (Simons 1981). tion processing deficits in schizophrenia for many years (Braff et Although PPI is thought to be automatic, several studies have al 1978; Cadenhead et al 1997; Dawson et al 1993; Hazlett et al shown that by directing attention toward the prepulse, PPI and 1998). More recently, PPI has been used with schizotypal indi- PPF can be enhanced (Dawson et al 1993; Filion et al 1993, 1994; viduals (Cadenhead et al 1993) and first-degree relatives of Hawk et al 2002; Hazlett et al 1998; Schell et al 2000). Some schizophrenics (Cadenhead et al 2000) as a possible endophe- investigators (e.g., Braff et al 1992; Cadenhead et al 2000) notypic marker. This article examines both PPI and PPF in interpret the PPI deficits exhibited by schizophrenia patients as schizophrenia patients, their first-degree relatives (siblings), and reflecting automatic sensorimotor gating deficits during a passive normal control subjects. It is thought that PPI and PPF reflect two separate mecha- attention task. Others (e.g., Dawson et al 1993, 2000) suggest that nisms: PPI is believed to index an automatic sensorimotor gating PPI deficits exhibited by schizophrenia patients are in the mechanism or a protection of processing mechanism (Braff et al controlled attentional modulation of PPI during an active atten- 1992; Graham 1980) whereby the processing of the prepulse tion task; however, differences in the methods used to elicit PPI gates out the sensory information of the startle stimulus. Deficits and PPF can have substantial impact on the results. The methods in PPI could theoretically lead to cognitive overload and frag- used in our study differ in several ways from those of laboratories mentation, traits exhibited by some pathologic populations such that consistently find absolute PPI deficits in schizophrenia as schizophrenia patients (Braff and Geyer 1990). In contrast, patients, for example, differences in level of background noise (which makes the prepulse less easily detectable), recording electromyogram (EMG) from the right eye, and the use of different types of prepulses. From the Department of Psychiatry and Biobehavioral Sciences (JKW, MM, DS, MFG), University of California, Los Angeles; Veterans Affairs Greater This study addressed one parameter that could lead to Los Angeles Healthcare System (JKW, MM, DS, MFG); Department of differences in findings across laboratories. Specifically, laborato- Psychology (MED), University of Southern California; and Department of ries that find deficits in absolute, automatic PPI in a passive Psychology (AMS), Occidental College, Los Angeles, California. attention paradigm in schizophrenia patients and at-risk subjects Address reprint requests to Jonathan K. Wynn, Ph.D., Veterans Affairs use discrete white noise prepulses (which begin and end before Greater Los Angeles Healthcare System, 11301 Wilshire Blvd. (MIRECC startle onset), whereas those that find deficits only in attentional 210A), Building 210, Los Angeles, CA 90073. modulation of PPI and PPF use continuous pure-tone prepulses Received March 28, 2003; revised August 26, 2003; accepted October 24, 2003. (which begin and continue up to and beyond startle onset) in an

0006-3223/04/$30.00 BIOL PSYCHIATRY 2004;55:518–523 doi:10.1016/j.biopsych.2003.10.018 © 2004 Society of Biological Psychiatry J.K. Wynn et al BIOL PSYCHIATRY 2004;55:518–523 519 active attention paradigm. Our study included both types of Table 1. Demographic Data prepulses in a passive attention paradigm. In passive attention paradigms, PPI has been proposed as a Schizophrenia Siblings Controls vulnerability indicator for schizophrenia. Those presumably at n 76 36 41 risk for developing schizophrenia (psychometrically defined; Age (years)a 42.70 (8.46) 38.23 (9.61) 35.64 (7.96) e.g., Simons and Giardina 1992) and those diagnosed with Range 20–55 18–55 21–50 schizotypal personality disorder (Cadenhead et al 1993) have Sex (M:F)b 74:2 17:19 19:21 exhibited PPI deficits. Only one published study, to our knowl- Antipsychotic Medicationc 22:43 edge, examined PPI in first-degree relatives of schizophrenia (typical:atypical) patients (Cadenhead et al 2000). In this study, deficient PPI in BPRS first-degree relatives was found under a specific condition— Total 47.78 (10.67) namely, measuring PPI from the right (but not left) eye at a Thinking Disturbance 2.95 (1.15) 30-msec lead interval, but not at lead intervals of 60 or 120 msec. Withdrawal-Retardation 2.30 (1.10) Although there have been numerous studies on PPI in Baseline Startle (␮V) 17.34 (28.82) 20.12 (22.34) 12.89 (14.43) schizophrenia, schizophrenia-spectrum populations, and at-risk Data are Mean (SD). populations, the literature concerning PPF deficits is sparse. M, male; F, female; BPRS, Brief Psychiatric Rating Scale. There is only one published study, to our knowledge, that aSchizophrenia patients signiﬁcantly older than controls, p Ͻ .001. examined PPF in a passive paradigm in schizophrenia. Braff et al bLarger proportion of males in the schizophrenia group, p Ͻ .001. c (1978) examined both PPI and PPF in schizophrenia patients. For patients on known medications. Although the patients were deficient in PPI, they did not show any PPF differences from normal control subjects. Other studies schizophrenia based on the Structured Clinical Interview for have found PPF attentional modulation deficits during an active DSM-IV (SCID; First et al 1996b). All patients with useable startle attention task (i.e., failure to produce more PPF during a eyeblink data (see Subject Exclusions) were receiving antipsy- to-be-attended prepulse) in both schizophrenia patients (Daw- chotic medication at the time of testing: 43 were on atypical son et al 1993, 2000) and subjects putatively at risk based on the antipsychotics, 22 were on typical antipsychotics, and 11 were on Chapman scales (Schell et al 1995). mixed or unknown (e.g., blinded) antipsychotics. Our research examined three separate issues. First, we tested Psychiatric symptoms were assessed in the patient group whether schizophrenia patients have deficits in PPI and PPF with using the 24-item Brief Psychiatric Rating Scale (BPRS; Ventura et the parameters used in this study. Second, we tested whether the al 1993). Interviewers were trained to a minimum intraclass siblings of patients show deficits in these processes. If SEM is a correlation of .80 on ratings. Table 1 provides the means and genetic vulnerability marker for information processing deficits standard deviations of the items (range 1–7) for the patients on in schizophrenia, patients and their nonpsychotic siblings should the thinking disturbance, or positive symptom (unusual thought show deficits in SEM. We hypothesized that both schizophrenia content, hallucinations, and conceptual disorganization) and on patients and their siblings would show reduced PPI as well as withdrawal-retardation, or negative symptom (emotional with- reduced PPF. The third issue this study addressed was whether drawal, blunted affected, and motor retardation) factors of the the use of two types of prepulses, discrete white noise pulses or BPRS (Overall et al 1967), as well as the total BPRS score. continuous pure tones, in the passive paradigm would partially Nonpatient control subjects were screened with the SCID and explain the differences in findings among laboratories. We SCID-II (First et al 1996a) and were excluded if they met criteria hypothesized that with the discrete white noise prepulses in a for any psychotic disorder, bipolar mood disorder, recurrent passive attention task patients and their siblings would show depression, substance dependence, or paranoid, schizoid, bor- reduced PPI compared with normal control subjects, and that derline, avoidant, or schizotypal personality disorders. Control with the continuous-tone prepulses patients and their siblings subjects were also excluded if there was any history of any would show normal levels of PPI compared with normal control psychotic disorder among first-degree relatives. subjects. To remove the possible effect of psychosis on SEM measures, we excluded from analysis siblings with a history of psychosis. Methods and Materials Of the original 48 siblings assessed with the SCID and SCID-II, five were excluded for a history of a psychosis (3 with schizo- Participants phrenia, 1 with schizoaffective disorder, and 1 with delusional Participants initially included 90 schizophrenia patients, 48 of disorder). None of the remaining siblings were taking any their siblings (sharing both biological parents of the schizophre- antipsychotic medication. Seven additional siblings were ex- nia proband), and 47 nonpatient control subjects. Subjects were cluded because they were nonresponsive to the startle stimulus, drawn from a larger study of Early Visual Processing in Schizo- leaving 36 siblings in all further analyses. Of these siblings, two phrenia (principal investigator: M.F. Green); schizophrenia pro- received an Axis II disorder diagnosis (not otherwise specified bands were recruited from outpatient clinics at the Greater Los and paranoid). Eight siblings had a history of drug or alcohol Angeles Veteran’s Administration (VA) Healthcare System and abuse, four a history of a depression, and four a history of both through presentations in the community. Nonpatient control substance abuse and depression. All siblings with a history of an subjects were recruited through flyers posted at the University of Axis I disorder were in remission. California-Los Angeles (UCLA) and the VA medical centers and Additional exclusion criteria for all patients and control through newspaper advertisements. Siblings were contacted with subjects included age Ͻ 18 or Ͼ 55 years, active substance use permission from the schizophrenia proband. All participants disorder in the past 6 months, identifiable neurologic disorder, gave written informed consent after the procedures were fully mental retardation, or seizure disorder. Diagnostic interviewers explained according to the procedures approved by the UCLA were trained to a minimum kappa of .75 for rating psychotic and and VA Institutional Review Boards. Patients met criteria for mood symptoms by the Diagnostic and Psychopathology Unit of

www.elsevier.com/locate/biopsych 520 BIOL PSYCHIATRY 2004;55:518–523 J.K. Wynn et al the UCLA Clinical Research Center for the Study of Schizophrenia Results (principal investigator: R.P. Liberman). Subject Exclusions Before any statistical analysis, subjects with blink amplitudes Procedure less than an average of 2.0 ␮V to startle-alone trials (based on the Subjects were seated in a comfortable chair. Two miniature average of all 15 ITI startle-alone trials) were considered nonre- (4-mm) Ag–AgCl electrodes were placed on the surface of the sponders and were excluded from further analysis (14 patients, 7 skin beneath the left eye, the first under the pupil and the other siblings, and 6 control subjects met this criterion). The following approximately 1 cm lateral to the first. A ground electrode was number of subjects comprised each group after exclusions: 76 (2 placed on the left mastoid. Electrical resistances between elec- female) patients, 36 (19 female) siblings, and 41 (22 female) trodes were less than 5000 Ohms. Eyeblinks were recorded as normal control subjects. EMG activity using a Coulbourn (Allentown, Pennsylvania) S75–01 bioamplifier and digitized for later analysis. The EMG Demographic Data activity was filtered with an 8-Hz high-pass filter and a 1000-Hz Descriptive data on the demographics of the schizophrenia low-pass filter, along with a 60-Hz notch filter. The data were patients, siblings, and control subjects are presented in Table 1. collected as raw EMG at 2000 Hz beginning 200 msec before The groups differed in age [F(2,150) ϭ 9.69, p Ͻ .001]; siblings startle stimulus onset and 300 msec after startle stimulus onset. and control subjects did not differ in age, nor did patients and Discrete prepulses were 20-msec presentations of 75 dB(A) siblings, but schizophrenia patients were older than the control white noise bursts (20–20,000 Hz) with a near-instantaneous subjects. Although there were significant age differences across rise–fall time. Continuous prepulses were 5000 msec presenta- groups, age, which ranged from at least 21–50 years in each tions of 75 dB(A) 1000 Hz tones, with a rise–fall time of 25 msec. group, did not correlate with any of the dependent variables in The startle stimulus consisted of a 50 msec, near instantaneous any group (p Ͼ .30 in all cases). Therefore, it appears that SEM rise–fall time, 105 dB(A) white noise burst (20–20,000 Hz). There is unrelated to age in the age range of these subjects, and it is was no imposed background noise. Auditory stimuli were digi- unlikely that age can explain any group differences. There were tized on a computer and presented through the computer’s also significant gender differences across the three groups, with sound card binaurally through headphones. Decibel levels were the schizophrenia patients having a higher proportion of men recorded with a Realistic sound level meter using a Quest than the siblings or control subjects (␹2 ϭ 48.50, p Ͻ .001). The Electronics earphone coupler. gender distributions were comparable for the siblings and the The experimental session consisted of 24 trials, presented in a control subjects. Possible effects of gender differences are dis- mixed, pseudorandom order, split into two blocks (12 trials in cussed later. There were no significant correlations between PPI each block). We assessed PPI using a lead interval of 120 msec and PPF with BPRS scores. There was also no significant effect of and PPF using a lead interval of 4500 msec. Of these 24 trials, patients’ type of medication (i.e., atypical vs. typical antipsychot- there were four startle-probed trials presented for each combi- ics) on PPI or PPF. nation of prepulse duration and lead interval, resulting in four discrete prepulses with a 120 msec lead interval, four discrete Baseline Startle Amplitude prepulses with a 4500 msec lead interval, four continuous Baseline startle amplitude and habituation for the 15 ITI prepulses with a 120 msec lead interval, and four continuous startle-alone presentations were examined to determine if the prepulses with a 4500 msec lead interval (there were four groups differed. Greenhouse–Geisser corrections (ε) were used unprobed prepulses for each prepulse type as well to make the to adjust for Type I error probabilities. We report the uncorrected presentation of prepulse plus startle unpredictable). Averages for degrees of freedom and the Greenhouse–Geisser corrected each of the two prepulse trials for each condition in each block probability levels for all analyses that have more than two levels were computed and used in the analyses. There were also 15 of repeated measures. Results of a repeated measures analysis of intertrial interval (ITI) startle probes presented throughout the variance (ANOVA) revealed no significant group differences (F experiment, varying from 8–11 sec into the ITI. Responses to the Ͻ 1; an a priori significance level of .05 was used in this and all first 8 ITI startle bursts were averaged and used to compute PPI other analyses). Results also showed a significant effect of block and PPF for the first block and the last 7 ITI startle bursts were [F(14,2100) ϭ 17.14, p Ͻ .001, ε ϭ .463], indicating that startle averaged to compute PPI and PPF for the second block. The ITIs reactivity decreased throughout the session; however, the block varied from 30–36 sec. effect did not interact with group, suggesting that all three groups Eyeblinks were scored offline using the raw, rectified EMG showed equal habituation. The mean and SD baseline startle signal. Prestartle EMG activity was defined as occurring 100 msec amplitude for each group are seen in Table 1. before startle stimulus onset. Peak startle activity was defined as a 20 msec average of the peak activity (10 msec pre-, 10 msec Prepulse Inhibition postpeak) occurring within a window of 20–120 msec after The PPI scores were analyzed with a 3 (group) ϫ 2 (prepulse startle stimulus onset. Eyeblink amplitude was the difference duration) ϫ 2 (block) mixed ANOVA. Results showed a main between the peak startle activity and the prestartle EMG activity. effect of prepulse duration [F (1,150) ϭ 49.46, p Ͻ .001], and a Startle eyeblink modification scores were then calculated as main effect of block [F(1,150) ϭ 19.93, p Ͻ .001]. There were no percent change scores: [(prepulsed startle Ϫ startle alone) / other main effects or interactions. The main effect of prepulse (startle alone) * 100]. Thus, a negative percent change score duration was due to discrete prepulses producing significantly indicates PPI, whereas a positive percent change score indicates greater prepulse inhibition (Ϫ65.44%) than continuous prepulses PPF. Percent change units are preferred over difference scores (Ϫ49.36%). The main effect of block is due to greater PPI seen in because difference scores in absolute ␮V units are correlated the first block (Ϫ63.33%) versus the second block (Ϫ51.47%). with baseline startle blink amplitude, whereas percent change Although there were few women in the patient sample, there scores are not, removing any dependence on baseline startle were adequate numbers in the sibling and control groups, and amplitude (Jennings et al 1996). gender was included as a separate group factor. Results revealed www.elsevier.com/locate/biopsych J.K. Wynn et al BIOL PSYCHIATRY 2004;55:518–523 521

Discussion

The three major findings of this paper are the following: 1) deficient PPF was found in schizophrenia patients and their siblings; 2) intact PPI was found in schizophrenia patients and their siblings, a finding contrary to many published studies finding PPI deficits in such groups; and 3) these findings were obtained regardless of the duration of the prepulse (i.e., discrete white noise vs. continuous tone). The principal new finding of this study was the deficient PPF exhibited by both schizophrenia patients and their siblings. We believe this is the first study to show PPF deficits in these groups. Both patients and siblings showed significantly less PPF than control subjects, and the patients and siblings did not differ in their amount of PPF. These results suggest that the deficit in PPF in schizophrenia is not due solely to the clinical presence of the disease. Instead, the deficient PPF in patients and siblings suggests a deficit in generalized alerting or orienting that may be a possible vulnerability marker for schizophrenia. If these find- Figure 1. Group means for prepulse inhibition (PPI) and prepulse facilitation ings are replicated, PPF may be another example of a psycho- (PPF) scores averaged across prepulse duration. *Signiﬁcant difference in Ͻ physiologic endophenotypic markers for schizophrenia, similar PPF from control subjects versus patients and siblings (p .05). Bars repre- to P50 gating (Siegel et al 1984), P300 (Turetsky et al 2000), sent standard errors of the means. mismatch negativity (Michie et al 2002), smooth-pursuit eye no significant gender main effects or interactions. Group means, tracking (Iacono et al 1992), and perhaps PPI (Cadenhead et al averaged over trial block, for PPI can be seen in Figure 1. All 2000). three groups produced PPI that was significantly different from Orienting deficits are seen widely in schizophrenia patients, zero (p Ͻ .05 in all groups). in response systems as diverse as skin conductance (Bernstein et al 1982; Dawson et al 1994) and mismatch negativity (Shelley et Prepulse Facilitation al 1991). Deficits in orienting may point to deficiencies in the The PPF scores were analyzed with a 3 (group) ϫ 2 (prepulse ability to process salient environmental stimuli (Dawson and duration) ϫ 2 (block) mixed ANOVA. Results showed a main Nuechterlein 1984); however, these deficits have rarely been effect of prepulse duration [F(1,150) ϭ 9.66, p Ͻ .01], a main studied using PPF as an index of orienting. The failure to find PPF effect of block [F(1,150) ϭ 12.94, p Ͻ .001], and a significant deficits in schizophrenia patients in the only prior study that group effect [F(2,150) ϭ 5.74, p Ͻ .01]. There were no significant examined PPF (Braff et al 1978) may have been due to several interactions. important experimental differences, such as lead interval (2000 The significant block effect was due to greater PPF in the later msec as opposed to 4500 msec) and how startle was measured (24.76%) versus earlier trial block (1.59%). During the first block, (by potentiometer as opposed to EMG). Our study demonstrates PPF was not significant (i.e., significantly different from zero) that it is possible to detect deficits in PPF in schizophrenia overall or in any individual group. This effect was expected patients, even in the absence of PPI deficits. It also demonstrates based on a block effect in a previous study of normal college the need to gain a better understanding of the experimental students (Schell et al 2000) that used the same procedures as our conditions under which PPF deficits can be observed and study. Therefore, the data were reanalyzed only for the second replicated and whether PPF reflects deficits in generalized ori- block, where significant overall PPF occurred. There was a enting and alertness. For instance, we observed significant PPF significant main effect of prepulse duration [F(1,150) ϭ 6.93, p Ͻ among control subjects only in the second trial block, which .01] and a significant group effect [F(2,150) ϭ 5.14, p Ͻ .01]. The does not seem consistent with the view that PPF reflects gener- prepulse main effect was due to continuous prepulses generating alized orienting or alertness processes. significantly larger prepulse facilitation (36.60%) than discrete Despite the notable findings with PPF, we were unable to find prepulses (12.92%). The significant group effect can be seen in PPI deficits in schizophrenia patients and their siblings using Figure 1. Using Duncan’s Range test for multiple comparisons, either discrete white noise or continuous-tone prepulses. Other the normal control subjects produced significantly greater PPF studies with schizophrenia patients have found PPI deficits at the (47.00%) than either the patients (9.21%) or the siblings (18.06%); 120-msec lead interval, as well as shorter intervals (e.g., Braff et patients and siblings were not significantly different. Only the al 1992, 2001; Cadenhead et al 2000; Kumari et al 2000, 2002). normal control subjects [t(40) ϭ 3.49, p Ͻ .05] and siblings [t(35) The failure to find PPI deficits in schizophrenia patients does not ϭ 2.15, p Ͻ .05] produced PPF that was significantly different seem to be due to major differences in subject characteristics from zero. As with the PPI data, gender was added as a group (e.g., age and illness duration) between our studies and others variable; there were no main effects or interactions with gender. that have yielded positive results (e.g., Braff et al 1999, 2001); To further explore possible effects of gender, we analyzed the however, as mentioned previously, many of the studies reporting data of the second block for the siblings and control subjects only positive findings of PPI deficits in patients, and the one study of because these groups had comparable gender distributions with siblings, used experimental procedures that differed from those larger group ϫ gender sample sizes. We found no significant used in our experiment. We measured EMG from the left eye, gender effects or interactions with gender. Therefore, it is whereas most studies that have found PPI deficits in schizophre- unlikely that the group differences observed are due to a gender nia measured startle eyeblink from the right eye (e.g., Braff et al effect. 1992, 1999; Cadenhead et al 2000; Kumari et al 2000 but for an

www.elsevier.com/locate/biopsych 522 BIOL PSYCHIATRY 2004;55:518–523 J.K. Wynn et al exception see Grillon et al 1992). Also, we did not use any Cadenhead KS, Geyer MA, Braff DL (1993): Impaired startle prepulse inhibi- imposed background noise, whereas studies that have reported tion and habituation in patients with schizotypal personality disorder. PPI deficits in schizophrenia patients (Braff et al 1991, 2001; Am J Psychiatry 150:1862–1867. Cadenhead et al 2000) have typically used a relatively strong Cadenhead KS, Geyer MA, Butler RW, Perry W, Sprock J, Braff DL (1997): Information processing deficits of schizophrenia patients: Relationship background noise (i.e., 70 dB) with short duration prepulses (20 to clinical ratings, gender and medication status. Schizophr Res 28:51–62. msec) that usually were 8–16 dB increases over the imposed Cadenhead KS, Swerdlow NR, Shafer KM, Diaz M, Braff DL (2000): Modulation background noise. Based on studies from our laboratory and of the startle response and startle laterality in relatives of schizophrenic others (e.g., Ford et al 1999), it appears that background noise patients and in subjects with schizotypal personality disorder: Evidence may make the auditory prepulses difficult to detect, especially of inhibitory deficits. Am J Psychiatry 157:1660–1668. when they are short-duration discrete changes. The possibility Dawson ME, Hazlett EA, Filion DL, Nuechterlein KH, Schell AM (1993): Atten- that PPI deficits in schizophrenia are seen primarily when tion and schizophrenia: Impaired modulation of the startle reflex. J Ab- norm Psychol 102:633–641. prepulses are near threshold levels was suggested by Judd et al Dawson ME, Nuechterlein KH (1984): Psychophysiological dysfunctions in (1992). the developmental course of schizophrenic disorders. Schizophr Bull The study by Cadenhead et al (2000) found PPI deficits in 10:204–232. siblings under one specific condition, while using a 30-msec lead Dawson ME, Nuechterlein KH, Schell AM, Gitlin M, Ventura J (1994): Auto- interval, measuring EMG from the right eye, and using a constant nomic abnormalities in schizophrenia: State or trait indicators? Arch Gen background noise. Our study only had one early lead interval Psychiatry 51:813–824. (120 msec), measured EMG from the left eye, and had no Dawson ME, Schell AM, Hazlett EA, Nuechterlein KH, Filion DL (2000): On the clinical and cognitive meaning of impaired sensorimotor gating in background noise. Cadenhead et al (2000) suggested that mea- schizophrenia. Psychiatry Res 96:187–197. suring PPI from the right eye taps into a left temporal lobe deficit Filion DL, Dawson ME, Schell AM (1993): Modification of the acoustic startle- that is seen in other different measures (e.g., Gur and Chin 1999; reflex eyeblink: A tool for investigating early and late attentional pro- Salisbury et al 1998). It is entirely possible that we would have cesses. Biol Psychol 35:185–200. found PPI deficits in patients and their siblings had we used the Filion DL, Dawson ME, Schell AM (1994): Probing the orienting response with same experimental procedures as Cadenhead and colleagues. startle modification and secondary reaction time. Psychophysiology These results indicate the value of further explorations using 31:68–78. First MB, Gibbons M, Spitzer RL, Williams JBW (1996a): Structured Clinical PPI and PPF to study schizophrenia and as endophenotypic Interview for DSM-IV Axis II Personality Disorders. New York: Biometrics markers for schizophrenia. We found for the first time deficits in Research Department. New York State Psychiatric Institute. PPF in schizophrenia patients and their siblings. These results First MB, Gibbons M, Spitzer RL, Williams JBW (1996b): Users guide for the point to the possibility that PPF deficits may be an additional Structured Clinical Interview for DSM-IV Axis I Disorders-Research Version index of a vulnerability factor for schizophrenia. It is also (SCID-1, Version 2.0, February 1996 Final Version). New York: Biometrics important to note that PPF deficits, similarly to PPI deficits, may Research Department. New York State Psychiatric Institute. not be unique to schizophrenia, and these findings should be Ford JM, Roth WT, Menon V, Pfefferbaum A (1999): Failures of automatic and strategic processing in schizophrenia: Comparisons of event-related studied in other psychiatric control subjects. It is possible, indeed brain potential and startle blink modification. Schizophr Res 37:149–163. likely, that PPF is seen only under certain experimental condi- Graham FK (1975): The more or less startling effects of weak prestimulation. tions, similar to PPI. This study strongly encourages further Psychophysiology 12:238–248. examination of the specific experimental conditions under which Graham FK (1980): Control of reflex blink excitability. In Thompson RF, Hicks PPI and PPF deficits are seen in schizophrenia and at-risk LH, Shryrkov VB, editors. Neural Mechanisms of Goal Directed Behavior populations. and Learning. New York: Academic Press, 511–519. Graham FK, Hackley SA (1991): Passive and active attention to input. In Jennings JR, Coles MGH, editors. Handbook of Cognitive Psychophysiol- This research was supported by Grant Nos. MH-46433 (MED) ogy: Central and Autonomic Nervous System Approaches. Sussex, England: and MH-43292 (MFG) and by the Department of Veterans Wiley, 251–356. Affairs, Veterans Integrated Services Network 22 Mental Illness Grillon C, Ameli R, Charney DS, Krystal J, Braff D (1992): Startle gating deficits Research, Education, and Clinical Center. JKW was supported by occur across prepulse intensities in schizophrenic patients. Biol Psychia- a National Institute of Mental Health National Research Service try 32:939–943. Gur RE, Chin S (1999): Laterality in functional brain imaging studies of schizo- Award Training Grant No. MH14584 (principal investigator: phrenia. Schizophr Bull 25:141–145. Keith H. Nuechterlein) during preparation of this article. We Hawk LW Jr, Redford JS, Baschnagel JS (2002): Influence of a monetary thank Bill Troyer for technical assistance and programming. incentive upon attentional modification of short-lead prepulse inhibition and long-lead prepulse facilitation of acoustic startle. Psychophysi- Bernstein AS, Frith CD, Gruzelier JH, Patterson T, Straube E, Venables PH, ology 39:674–377. Zahn TP (1982): An analysis of the skin conductance orienting response Hazlett EA, Buchsbaum MS, Haznedar MM, Singer MB, Germans MK, Schnur in samples of American, British and German schizophrenics. Biol Psychol DB, et al (1998): Prefrontal cortex glucose metabolism and startle eye- 14:155–211. blink modification abnormalities in unmedicated schizophrenia pa- Braff D, Stone C, Callaway E, Geyer M, Glick I, Bali L (1978): Prestimulus effects tients. Psychophysiology 35:186–198. on human startle reflex in normals and schizophrenics. Psychophysiology Iacono WG, Moreau M, Beiser M, Fleming JA, Lin TY (1992): Smooth-pursuit 15:339–343. eye tracking in first-episode psychotic patients and their relatives. Braff DL, Geyer MA (1990): Sensorimotor gating and schizophrenia: Human J Abnorm Psychol 101:104–116. and animal model studies. Arch Gen Psychiatry 47:181–188. Jennings PD, Schell AM, Filion DL, Dawson ME (1996): Tracking early and late Braff DL, Geyer MA, Light GA, Sprock J, Perry W, Cadenhead KS, Swerdlow NR stages of information processing: Contributions of startle blink reflex (2001): Impact of prepulse characteristics on the detection of sensorimo- modification. Psychophysiology 33:148–155. tor gating deficits in schizophrenia. Schizophr Res 49:173–180. Judd LL, McAdams L, Budnick B, Braff DL (1992): Sensory gating deficits in Braff DL, Grillon C, Geyer MA (1992): Gating and habituation of the startle schizophrenia: New results. Arch Gen Psychiatry 149:488–493. reflex in schizophrenic patients. Arch Gen Psychiatry 49:206–215. Kumari V, Soni W, Mathew VM, Sharma T (2000): Prepulse inhibition of Braff DL, Swerdlow NR, Geyer MA (1999): Symptom correlates of prepulse the startle response in men with schizophrenia: Effects of age of inhibition deficits in male schizophrenic patients. Am J Psychiatry onset of illness, symptoms, and medication. Arch Gen Psychiatry 57:609– 156:596–602. 614. www.elsevier.com/locate/biopsych J.K. Wynn et al BIOL PSYCHIATRY 2004;55:518–523 523

Kumari V, Soni W, Sharma T (2002): Prepulse inhibition of the startle re- Shelley AM, Ward PB, Catts SV, Michie PT, Andrews S, McConaghy N (1991): sponse in risperidone-treated patients: Comparison with typical anti- Mismatch negativity: An index of a preattentive processing deficit in psychotics. Schizophr Res 55:139–146. schizophrenia. Biol Psychiatry 30:1059–1062. Michie PT, Hamish I-B, Todd J, Jablensky AV (2002): Duration mismatch Siegel C, Waldo M, Mizner G, Adler L, Freedman R (1984): Deficits in negativity in biological relatives of patients with schizophrenia spec- sensory gating in schizophrenic patients and their relatives: Evidence trum disorders. Biol Psychiatry 52:749–758. obtained with auditory evoked responses. Arch Gen Psychiatry 41:607– Nuechterlein KH, Dawson ME (1984): A heuristic vulnerability/stress model 612. of schizophrenic episodes. Schizophr Bull 10:300–312. Simons RF (1981): Electrodermal and cardiac orienting in psychometrically Overall JE, Hollister LE, Pichot P (1967): Major psychiatric disorders: A four- defined high-risk subjects. Psychiatry Res 4:347–356. dimensional model. Arch Gen Psychiatry 16:146–151. Simons RF, Giardina BD (1992): Reflex modification in psychosis-prone Salisbury DF, Shenton ME, Sherwood AR, Fischer IA, Yurgelun-Todd DA, young adults. Psychophysiology 29:8–16. Tohen M, McCarley RW (1998): First-episode schizophrenic psychosis Spohn HE, Coyne L, Wilson JK, Hayes K (1989): Skin-conductance orienting differs from first-episode affective psychosis and controls in P300 ampli- response in chronic schizophrenics: The role of neuroleptics. J Abnorm tude over left temporal lobe. Am J Psychiatry 55:173–180. Psychol 98:478–486. Schell AM, Dawson ME, Hazlett EA, Filion DL (1995): Attentional modulation Turetsky BI, Cannon TD, Gur RE (2000): P300 subcomponent abnormalities in of startle in psychosis-prone college students. Psychophysiology 32:266– schizophrenia: III. Deficits in unaffected siblings of schizophrenic pro- 273. bands. Biol Psychiatry 47:380–390. Schell AM, Wynn JK, Dawson ME, Sinaii N, Niebala CB (2000): Automatic and Ventura J, Lukoff D, Nuechterlein KH, Liberman RL, Green MF, Shaner A controlled attentional processes in startle eyeblink modification: Effects (1993): Brief Psychiatric Rating Scale (BPRS). Int J Meth Psychiatr Res of habituation of the prepulse. Psychophysiology 37:409–417. 3:227–243.