Time course of regional brain activation associated with onset of auditory/verbal hallucinations

The time course of brain activation prior to onset of auditory/verbal hallucinations was characterised using functional magneticresonance imaging in six dextral patients with schizophrenia.Composite maps of pre-hallucination periods revealed activationin the left anterior insula and in the right middle temporalgyrus, partially replicating two previous case reports, aswell as deactivation in the anterior cingulate and parahippocampal gyri. These findings may reflect brain events that triggeror increase vulnerability to auditory/verbal hallucinations.

INTRODUCTION

TOP

INTRODUCTION

Auditory/verbal hallucinations of spoken speech, which occurin 60–80% of people with schizophrenia, are associatedwith high levels of distress and functional disability. Althoughneuroimaging studies have associated regional brain activationwith such hallucinations, the data generally cannot differentiateactivation reflecting their production from downstream consequencesof these events such as shifts in arousal or registration ofhallucinations in verbal memory. Two reports of time courseof brain activation associated with auditory/verbal hallucinationscharacterised using functional magnetic resonance imaging (fMRI)are therefore of special interest. Lennox et al described activationin the right middle temporal gyrus detected 3 s prior to hallucinationonset followed by bitemporal activation in one patient.¹ Shergill et al detected left inferior frontal and right middle temporalgyral activation arising a full 6–9 s prior to hallucinationonset in two patients.² The pre-hallucination timing of theseactivations suggests neural events triggering or increasingvulnerability to auditory/verbal hallucinations. We attemptedto replicate and extend these reports by studying a larger patient sample.

Method

TOP

Method

Participants
Six right-handed patients with either schizophrenia or schizo-affective disorder diagnosed using DSM–IV³ criteria were studied.These patients were treated with clozapine (n=3), haloperidol(n=1) or olanzapine (n=1), or no antipsychotic drug (n=1).Informed written consent was obtained from all participantsprior to the study.

Imaging
We used a 1.5 T Signa LX scanner (General Electric, Milwaukee,Wisconsin, USA) to obtain T₁-weighted structural images of14 contiguous 6 mm slices acquired parallel to the anterior–posterior commissural line. These images were used for anatomical identificationand coincided with slices in the functional runs. For functionalruns, participants were instructed to depress a button withtheir right hand to mark the onset of each auditory/verbalhallucination and to release the button at termination of theevent; six or seven of these runs lasting 4 min 6 s each wereused for each participant. Functional runs – 14 axial–oblique slices of single-shot echoplanar imaging: repetition time (TR)1500 ms, echo time (TE) 60 ms, flip angle 60°, 64x64 acquisitionmatrix, voxel size 3.125 mmx3.125 mmx6 mm – were motion-correctedusing the statistical parametric mapping algorithm and subjectedto a spatial Gaussian filter with 2 pixels full width at halfmaximum. The time course of button-presses in each scan wasconvolved with a standard model of the haemodynamic responsefunction. For each individual and each scan, blood oxygen leveldependent (BOLD) signal fluctuations in all pixels composinga given slice were correlated with the reference time courseat temporal shifts ranging from –4.5 s to 4.5 s in 1.5-ssteps, after removing the mean time course across the slicefrom each pixel time course and from the reference time course.For each time lag and run, correlations were transformed toan approximately normal Gaussian distribution. Gaussian-transformedmaps were averaged across runs, yielding a map representingthe strength of correlations to the reference button-press time course in terms of standardised z-values. The seven z-maps of correlations from each participant for each time lag werethen transformed to Talairach coordinates. At each Talairachpixel and time lag, a t-test was used to assess significanceof deviation of z-values from zero. This correlation-basedanalysis can accommodate shorter inter-hallucination intervalseven though the BOLD signal may not return fully to baseline– irregular temporal spacing of hallucinations can improvestatistical power, just as `jittering' allows closer event spacing in event-related designs.

Results

TOP

Results

Positive BOLD signal correlations with hallucination time coursewere detected selectively at negative time intervals (timelags ranging from –4.5 s to –1.5 s) in the leftinsula and a right middle temporal region, as well as in leftpre-central areas (online Fig. DS1). Positive correlationsalso emerged in the superior temporal gyrus bilaterally at negativetime lags, peaking approximately at zero time lag (online Fig.DS1). Negative BOLD signal correlations with hallucinationtime course were detected in the right ventral anterior cingulateand left parahippocampal gyri at negative time lags.

Discussion

TOP

Discussion

Two sites exhibiting positive BOLD signal correlations withhallucination time course exclusively at negative time lagslargely replicated pre-hallucination activation sites describedpreviously.

Right middle temporal site
The right middle temporal site (Brodmann area 21) was locatedremarkably close to middle temporal gyrus sites of pre-hallucinationactivation reported by Lennox et al and Shergill et al.^1,2 Activation in the bilateral superior temporal gyrus (Brodmannarea 22) emerged at negative time lags also, but peaked laterand was broadly distributed over both positive and negativetime lags. Our correlation-based method for mapping BOLD signaltime course will produce temporal smearing that broadens with increasing neural activation. The temporal pattern of our datasuggests therefore that the more robust bilateral activationin the superior temporal gyrus arose somewhat later –perhaps at hallucination onset – than the middle temporalgyrus activation. Bilateral activation of the latter regionhas been associated with aspects of verbal comprehension duringspeech processing distinct from acoustic feature detectionreferable to the superior temporal gyrus,⁴ whereas non-dominantmiddle temporal gyral activation has been associated with detectingprosodic features of spoken speech.⁵ One plausible accountof our findings is that pre-hallucination activation in the middle temporal gyrus reflecting verbal content and/or prosodyis subsequently propagated to the superior temporal gyrus viatop-down processing, which generates (hallucinated) acousticrepresentations.

Left anterior insula
The left anterior insula was close to a site of activation inthe left inferior frontal gyrus 9 s prior to hallucinationonset identified by Shergill et al,² who reported expandedactivation incorporating the left insula at later times. Leftinsula activation has been associated with speech articulation,⁶ imagining spoken speech of others,^7,8 and focused auditory attention,⁴ suggesting that pre-hallucination insula activationreflects inner speech or auditory imagery generation as previously hypothesised,^7,8 or enhanced auditory attention. However,pre-hallucination insula activation might instead reflect motormovement required to signal these events. This possibilityis suggested by the fact that simple generation of finger movementsis preceded by activation in the adjacent Broca's area, whichhas been postulated to reflect mental preparation.⁹

Other sites
Evidence of right ventral anterior cingulate and left parahippocampal deactivation preceding hallucination onset was detected. Co-occurring deactivations in these regions have also been linked to heightened vigilance/attention,¹⁰ suggesting a shift in cognitive statepreceding auditory/verbal hallucinations. Along these lines,Arieti described a `listening attitude' that predisposes peoplewith schizophrenia to hear `voices'.¹¹ Pre-central activationemerging prior to hallucination onset in our study could reflecteither inner speech generation or signalling hallucinationsby finger movement.

In summary, activation detected as BOLD signal changes correlatedwith auditory/verbal hallucination time course at negativetime lags may reveal complex brain processes triggering theseexperiences. Future studies of this type would be advancedby controlling for effects of motor behaviour required to signalhallucination occurrences.⁹

ACKNOWLEDGMENTS

TOP

ACKNOWLEDGMENTS

This study was funded by the NIMH grant R01MH067073 (R.E.H.),the Dana Foundation (R.E.H.) and NIH/NCRR/GCRC program grantRR00125.

REFERENCES

TOP