Attention and prepulse inhibition: the effects of task-relevant, irrelevant, and no-task conditions

Abstract

We investigated whether attentional modulation of prepulse inhibition (PPI) is due to increased protection of processing of attended lead stimuli, decreased protection of processing of ignored lead stimuli, or a combination of both processes. Task and no-task trials, pre-cued by red and blue dots on a computer screen, were randomly intermixed. College student participants were instructed to do a tone duration judgment task on trials preceded by one color (task condition) and to do nothing on trials preceded by the other color (no-task condition). On task condition trials participants were instructed to count the number of longer duration tones of a particular pitch (attended condition) and to ignore tones of a different pitch (ignored condition). White noise startle stimuli were presented at 60 ms and 120 ms lead intervals on some trials in each condition. Additional startle stimuli were presented during the inter-trial intervals to measure baseline (unmodified) startle response. PPI in the attended condition was reliably greater than that in both the ignored and no-task conditions. PPI did not differ between the ignored and no-task conditions. The results are consistent with the conclusion that attentional modulation of PPI is due to increased protection of attended stimuli and not to decreased protection of ignored stimuli. Possible reasons for robust attentional modulation at the 60 ms lead interval as well as the usual 120 ms lead interval are discussed.

Introduction

A sudden and intense stimulus, such as a brief burst of white noise, elicits a startle reflex response. In humans, this response is most commonly measured by the amplitude of the eyeblink (see Berg and Balaban, 1999 for a review of startle elicitation, recording, and quantification). If the startling stimulus is preceded by a non-startling stimulus, such as a soft tone, the amplitude of the eye-blink response is modified. The non-startling stimulus is called a lead stimulus or prepulse, and the interval between the onset of the lead and startle stimuli is called the lead interval or stimulus onset asynchrony (SOA). The specific effect of the lead stimulus on startle eyeblink depends upon the lead interval. At short lead intervals (15–400 ms), the amplitude of the startle response is decreased. This is called prepulse inhibition (PPI), which is thought to reflect a sensory gating process that protects initial processing of the lead stimulus from interference by extraneous stimuli (see Blumenthal, 1999 for a review of PPI).

The effect of the lead stimulus on startle is further modified by attention. At short lead intervals, PPI after a lead stimulus that is attended is typically greater than PPI after a lead stimulus that is not attended (Dawson et al., 1993, Dawson et al., 2000, Filion et al., 1993, Filion et al., 1994, Hawk et al., 2002, Hazlett et al., 1998, Jennings et al., 1996, Schell et al., 2000). This increase in PPI is called attentional modulation.

Most of the experiments that have investigated attentional modulation of PPI have employed a tone duration judgment task (e.g., Filion et al., 1993). In this task, the lead stimuli are high and low pitch tones of short and long duration. Participants are instructed to attend to one of the two pitches (the task-relevant, or attended, stimulus) and to count the number of longer duration tones of that pitch while ignoring tones of the other pitch (the task-irrelevant, or ignored, stimulus). Startle stimuli are presented at various lead intervals during both the attended and ignored stimuli. For auditory lead and startle stimuli the typical pattern is reliably greater PPI to attended stimuli than ignored stimuli at a 120 ms lead interval and no reliable difference in PPI between attended and ignored stimuli at lead intervals of 60 and 240 ms.

Dawson et al. (1993) interpreted these findings in terms of the time course of lead stimulus processing. They proposed that pre-attentive stimulus detection and evaluation occur at lead intervals of about 60 ms, stimulus discrimination and greater allocation of attention to the attended lead stimulus occur at lead intervals of about 120 ms, and transition from stimulus evaluation to duration judgment occurs at lead intervals of about 240 ms. They proposed that greater inhibition to attended lead stimuli at the 120 ms lead interval reflects increased sensory gating, which screens out extraneous stimuli, thus protecting the processing of the task-relevant lead stimulus.

An important question is: How does the processing of an attended lead stimulus differ from the processing of an ignored lead stimulus? Does the difference in startle inhibition between the attended and ignored conditions occur because attention increases protection of processing of the attended stimulus, decreases protection of processing of the ignored stimulus, or some combination of both processes? This issue can be operationally tested by comparing PPI following attended and ignored lead stimuli in a selective attention task with PPI following similar lead stimuli in a no-task condition. If attentional modulation reflects only enhanced protection of the attended stimulus, then PPI following the attended stimulus will be greater than that following the ignored and no-task stimuli, which will not differ. If attentional modulation reflects only suppressed protection of the ignored stimulus then PPI following the ignored stimulus will be less than PPI following the attended and no-task stimuli, which will not differ. Finally, if attentional modulation reflects both enhanced protection of the attended stimulus and suppressed protection of the ignored stimulus, then PPI following all three types of stimuli will differ.

Jennings et al. (1996), in an experiment using the tone duration judgment task in a between-subjects design with separate task and no-task groups, obtained results consistent with the conclusion that attentional modulation is due to increased protection of the attended stimulus compared to what would occur in a no-task condition. The stimulus sequence was identical for both the task and no-task groups; however, the instructions were different. Participants in the task group were told to count the number of longer duration tones of a specified pitch and to ignore tones of another pitch. As an incentive, they were paid up to US$5.00 for good performance on the task. Participants in the no-task group were not given a task and were not paid. PPI was reliably greater in the attended condition than the ignored condition at the 120 ms lead interval for the task group. PPI was marginally greater in the attended condition for the task group compared to matching trials for the no-task group. There was no reliable difference in PPI between the ignored condition for the task group and matching trials for the no-task group. Greater PPI in the attended than the ignored condition for the task, but not the no-task, group, and the absence of a difference between the ignored condition for the task group and matching trials for the no-task group led Jennings et al. (1996) to conclude that attentional modulation is due to increased protection of attended stimuli.

Filion and Poje (2003) carried out an experiment which used a tone duration judgment task with blocks of task and no-task trials in a within-subjects design. In the first of the two trial blocks, at both the 60 ms and 120 ms lead intervals, they found greater PPI in the attended and ignored conditions, which did not differ, compared to the no-task condition. In contrast, in the second of the two trial blocks, at only the 120 ms lead interval, they found greater PPI in the attended condition compared to the ignored and no-task conditions, which did not differ. Consistent with the operational criteria described above, the latter finding would indicate that attentional modulation reflects enhanced protection of attended stimuli only. However, Filion and Poje also noted that PPI at 120 ms following the ignored stimulus decreased significantly from the early to the late trial block, while PPI at 120 ms following the attended stimulus did not. They concluded that “It is unclear from this pattern alone whether selective attentional processes acted to enhance processing of the attended prepulse…or acted to reduce processing of the ignored prepulse…”

Consistent with the results of Filion and Poje (2003), Schell et al. (2000, experiment 2), in a study on the effect of habituation of the lead stimulus on PPI and attentional modulation, also obtained results consistent with decreased protection of ignored stimuli. They reported greater PPI to attended than ignored lead stimuli at a 120 ms lead interval only on the second of two trial blocks. However, the difference was due to decreased PPI in the ignored condition in the first trial block only, not to increased PPI in the attended condition. Schell et al. concluded that attentional modulation is due to reduced protection of a stimulus that can be ignored. However, since this study did not include a no-task condition, it could not be determined whether the protection for the ignored stimuli was the same as would have occurred for no-task stimuli, or whether there was active suppression of protection of ignored stimuli.

Finally, Hawk et al. (2002), in a study on the effect of payment for task performance, obtained results consistent with the conclusion that attentional modulation is due to both increased protection of attended stimuli and decreased protection of ignored stimuli. They reported reduced PPI at the 120 ms lead interval compared to the 60 ms lead interval for ignored lead stimuli and greater PPI at the 120 ms lead interval compared to the 60 ms lead interval for attended lead stimuli. Hawk et al. concluded that the attentional effect was due to both reduced protection of ignored lead stimuli and increased protection of attended lead stimuli. However, once again, the critical no-task condition was not available for comparison.

With four different studies and three different conclusions, the question of how attended and ignored lead stimuli affect PPI is not yet clear. In addition, for each of these studies, there are methodological issues that need to be addressed. The Jennings et al. (1996) experiment used a between-subjects design with separate task and no-task groups. The conditions for the two groups were quite different. The task group had a task to perform and had the incentive of financial reward for good performance. In contrast, the no-task group had no task to perform and no prospect of financial reward. It seems probable that motivation, therefore vigilance, was higher for the task than the no-task group. The Filion and Poje (2003) experiment used a within-subjects design with blocks of task and no-task trials. However, since some of the blocks were long (up to eight trials), it is possible that overall vigilance could have declined during the long blocks of no-task trials. The Schell et al. (2000) experiment was designed to investigate the effect of habituation of the lead stimulus on attentional modulation and the Hawk et al. (2002) experiment was designed to investigate the effect of extrinsic incentive on attentional modulation. Because they were planned for these purposes, the important no-task condition was not needed and not included. Without the no-task condition, which measures PPI without experimenter-manipulated attentional modulation, it was not possible to determine whether reduced protection of the ignored compared to the attended stimuli reflects active suppression of protection below what would occur in the absence of a task.

The purpose of the present experiment was to investigate the processes involved in attentional modulation in a within-subjects design in which task and no-task trials were interspersed in a single experimental procedure. This allows the use of the no-task condition, which indexes basic PPI without attentional modulation, as a reference for comparison with the attended and ignored trials of the task condition. This procedure controls for between-groups differences by allowing all participants to receive the same instructions and perform the same task. Since task and no-task trials occurred randomly, it was not possible for participants to adopt a task or no-task set for entire blocks of trials. Therefore, a high level of overall vigilance was required.