Data on deviance predictability in the assessment of mismatch negativity in patients with schizophrenia

We investigated the MMN at electrode Fz to 12% temporally predictable or unpredictable duration decrement deviant stimuli in 29 healthy controls and 31 schizophrenia patients. With a stimulus onset asynchronicity of 500 ms in the regular predictable condition, a deviant occurred every 4 s while it varied randomly in the unpredictable condition. Here we report detailed data tables and multivariate analysis of variance results (MANOVA) on MMN, P3a and standard ERP data including details on follow-up analyses. An extended figure shows MMN difference curves and averages to standard and deviant stimuli in both experimental conditions and subject groups.


a b s t r a c t
We investigated the MMN at electrode Fz to 12% temporally predictable or unpredictable duration decrement deviant stimuli in 29 healthy controls and 31 schizophrenia patients. With a stimulus onset asynchronicity of 500 ms in the regular predictable condition, a deviant occurred every 4 s while it varied randomly in the unpredictable condition.
Here we report detailed data tables and multivariate analysis of variance results (MANOVA) on MMN, P3a and standard ERP data including details on follow-up analyses. An extended figure shows MMN difference curves and averages to standard and deviant stimuli in both experimental conditions and subject groups.  We show an extended figure demonstrating how the mismatch negativity is generated by evoked response potential curves to standard and to deviant stimuli in the predictable and unpredictable deviance conditions.
We report data on ERP components to standard and deviant stimuli in patients and in controls which are not part of the MMN but which may be of specific interest for those doing evoked potential research in schizophrenia patients or other clinical samples.

Data
The data give details of analyses of the dependent variable mismatch negativity (MMN) and other evoked response potential (ERP) components in relation to the experimental factors. We assessed two groups (patients with schizophrenia and controls) and used two experimental conditions (with and without predictability of the deviant stimulus). Fig. 1 shows grand mean ERP curves to standard and deviant stimuli and the resulting MMN difference wave in both groups and both conditions (Tables 1-12).

Experimental design, materials and methods
We assessed 31 patients with schizophrenia and 29 control subjects. Groups did not differ in years in education, amount of smokers and a proxy of verbal IQ (MWT-B) [5]. Exclusion criteria were age exceeding 18-55 yrs, alcohol or drug abuse or dependency or past dependencies less than 1 year ago, acute neurological or DSM-IV axis-I disorders other than schizophrenia or schizoaffective disorder and current benzodiazepine medication.

Clinical assessments
Diagnoses were verified by means of Structured Clinical Interview for DSM-IV (SCID) [7] in patients and short diagnostic interviews for DSM-IV diagnoses in controls (Mini-DIPS) [6]. Clinical symptom assessments comprised the Positive and Negative Syndrome Scale (PANSS) [4] and the Global Assessment of Functioning Scale (GAF) [1].

Stimuli and experimental design
Auditory stimuli presented with a 500 ms stimulus onset asynchrony. Sine-wave tones were 1 kHz, 80 dBA and 80 ms with rise and fall times of 10 ms. Deviant tones had a duration of 40 ms with 5 ms rise/ fall time. Deviant probability was 12% in both conditions. In the fixed predictable condition the fourth stimulus was a duration deviant stimulus "D", the other stimuli were standards "S" resulting in a series of "SSSDSSSS" stimuli. In the unpredictable random condition, the duration deviant stimulus occurred randomly at the second to eight position in the train of eight stimuli. No deviants occurred in direct succession. Stimuli were presented in two runs using "Presentation" (V.14.1, Neurobehavioral Systems Inc.) software while participants watched a silent nature film (visual angle 5°).

EEG recording and analysis
EEG was recorded at Fz, Fcz, Cz, C3 and C4. Linked earlobes were used as reference and AFz as ground. EOG was recorded to monitor vertical and horizontal eye movements. EEG was recorded with a band pass filter of 1.5-250 Hz and a digitisation rate of 500 Hz. Trials with artefacts were excluded and data were passed through an IIR Butterworth zero 30 Hz low pass filter (48 dB/oct) [2]. Eye blinks were corrected using a regression method [3]. Segments were computed from 100 ms before to 400 ms after stimulus onset and baseline-corrected. Epochs exceeding 750 mV were rejected from further analysis. Single subject averages were computed for stimulus types and conditions. MMN difference waves were obtained and peak amplitudes and latencies were determined at electrode site Fz, which showed the largest MMN amplitude in this study. MMN mean amplitudes comprising data points 750 ms around individual peak amplitudes were computed. P3a was assessed from MMN difference curves for the largest positive deflection following the MMN. Additionally, we measured amplitudes and latencies of standard ERP components P50, N100 and P200 in ERP average waveforms to standard stimuli and the P50 component from ERP average waveforms to deviant stimuli.