Auditory P300 event-related potential: Normative data in the Indian population
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.222874
Source of Support: None, Conflict of Interest: None
Keywords: Auditory, event-related potential, oddball paradigm, P300, normative
Electrophysiological measures are being increasingly used to gain insights into the biological basis of psychiatric disorders. In the late 1960s, a careful analysis of electroencephalogram (EEG) revealed that the presentation of a stimulus produces specific changes in the brain owing to significant increase in the synaptic activity in millions of neurons simultaneously in a synchronized manner. The combined electrical responses of this neuronal population are known as event-related potentials (ERPs). P300 is a positive ERP waveform with maximum peak around 300 ms post stimulus. It is composed of two main subcomponents, namely, P3a and P3b. At the functional level, P3a is related to an “attention orienting complex,” whereas P3b is related to psychological constructs such as controlled information processing, context updating, cognitive closure, or response-related decisional stages.
Abnormalities of P300 have been demonstrated in diseases exhibiting cognitive dysfunction such as schizophrenia, Alzheimer's disease, posttraumatic stress disorder (PTSD), and depression-associated pseudodementia.,, At the present state of knowledge, it has limited routine clinical usage owing to its variability across population. Moreover, the definition of normal component values for P300 and a standard method for their acquisition has not been established. Therefore, it is imperative to develop normative data for P300 for widening its scope in clinical applications such as for the diagnosis of psychiatric disorders.
There are few studies that have described normative data for different age groups. Furthermore, P300 variations have been reported across ethnic groups, which may be related to culture and context. Normative data is not yet available for Indian population. The current study was planned to generate normative data of auditory P300 ERP for various age groups in the Indian population.
The study was carried out at the Centre for Cognitive Neurosciences at the Central Institute of Psychiatry (CIP), Ranchi, a tertiary teaching hospital in India. The study was approved by the Institute's ethics committee. The sample consisted of auditory P300 ERP of healthy control participants of studies previously conducted at CIP. A list of all previous studies, from 1999 to 2009, conducted at CIP measuring P300 was prepared. Only those studies with an auditory oddball paradigm were included. Healthy control participants in the age range of 10–50 years were selected from these studies. Healthy controls in these studies included staff and students of the Institute who were recruited through purposive sampling.
Out of the total 200 healthy controls, data from 155 was selected for the final analysis. Data from 45 subjects was excluded due to artefacts affecting a significant number of trials, leaving inadequate number of trials for averaging. Of the 155 participants, 40 (25.8%) were females. The mean age of the subject sample was 28.90 (standard deviation 8.97) years.
P300 recordings of healthy control participants of the previous studies meeting the inclusion criteria were analyzed for amplitude and latency. P300 amplitude was measured as the peak amplitude, relative to a pre-stimulus baseline, between 250–500 ms post-stimulus. The P300 latency was measured relative to the stimulus onset and was defined as the time-period between stimulus onset and peak amplitude. The amplitude and latency of P300 for Fz, Cz, and Pz were selected.
In all the studies, P300 was recorded by an auditory oddball paradigm, in which two types of tones were presented to the participants through a headphone, i.e., a frequent tone of 40 dB; and another, a rarer louder target tone at 55 dB. Participants were instructed to recognize the rarer type of the tone and press a button with the dominant hand each time they heard it. The electroencephalographic (EEG) filters were set at a high cut off of 100 Hz and a low cut off of 0.1 Hz. The frequency of the target tone was 2 KHz and presentation probability was 20%, whereas the frequency of the frequent tone was 1 KHz and the presentation probability was 80%.
Signal averaging was done using EB Neuro Galileo NT software (Firenze, Italy). Trials with significant artefacts, including eye blinks, eye-movements, muscle artefacts, and skin potentials were manually excluded before generating the averaged ERP waveforms.
Data obtained was analyzed using the Statistical Package for the Social Sciences (SPSS; IBM) version 16 for Windows. Data was grouped in to age ranges of 10–19, 20–29, 30–39, and 40–50 years. There were missing values in 9% of the data. Considering data missing at random, the mean values for that age group were imputed for the missing data. Mean and standard deviation (SD) was used to describe the normative data across four age groups. Multivariate analysis of variance (MANOVA) was used to compare P300 amplitude and latency among the four age groups followed by ANOVA for Fz, Cz, and Pz. Post hoc Ryan, Einot, Gabriel, Welsch Q (REGWQ) test was used for multiple comparisons across age ranges. Pearson's correlation coefficient (r) was calculated between age, P300 amplitude, and latencies. Significance level was kept at P < 0.05 (two-tailed).
P300 amplitude and latency
The mean and SD for the P300 amplitude and latency at Fz, Cz, and Pz for the four age groups are summarized in [Table 1]. For P300 amplitude, overall MANOVA was significant [Pillai's Trace F (9/453) = 3.46, P < 0.001]. Follow-up ANOVA showed significant difference across age groups at Fz, Cz, and Pz. Post hoc REGWQ test showed significantly higher amplitude in the 10–19 years age group as compared to all others at Fz and Pz, and higher amplitude than the 30–39 years age group at Cz [Figure 1]. For P300 latency, there was a trend towards significance for overall MANOVA [Pillai's Trace F (9/453) = 1.68, P = 0.09]. Follow-up ANOVA showed a trend towards significant difference across age groups at Fz only [Figure 2]. Jonckheere–Terpstra test for trend showed significant decreasing trend for P300 amplitude at Fz (P< 0.001), Cz (P = 0.002), and Pz (P = 0.015) across age groups, whereas no such trend was found for P300 latency. The percentile distribution for P300 amplitude and latency at Fz, Cz, and Pz for the four age groups is summarized in [Table 2].
Correlation between P300 amplitude, latency, and age
There was a significant negative correlation between age and P300 amplitude at Fz, Cz, and Pz, whereas a significant negative correlation was seen with P300 latency at Fz only. There was significant positive correlation of P300 amplitude at Fz with Pz and Cz and Pz, whereas, no correlation was seen between Cz and Fz. Significant positive correlation was noted between P300 latency at all the three locations. There was significant negative correlation between P300 amplitude at Fz with latency at Fz, Cz, and Pz; however, no correlation was found between amplitude at Cz and Pz with latencies at Fz, Cz, and Pz [Table 3].
Our study generated a P300 amplitude and frequency normative database at Fz, Cz, and Pz for the Indian population, which will provide reference values for studies defining abnormities in various psychiatric disorders. The mean P300 amplitude and latencies across age ranges were found to be higher than the previous studies reported by Polich et al., and Schiff et al. who used a similar odd ball paradigm. In contrast, the mean P300 latencies across age ranges were comparable to the findings observed in the study by Anderer et al., whereas the mean P300 amplitudes were lower in our sample across age groups. There was a decreasing trend in P300 amplitude across age groups at all the three sites that were most prominent at Pz, whereas no such trend was found for P300 latency. Some variability in P300 amplitude and latencies across studies could be due to several factors including food intake, body temperature, and handedness. Further, study parameters in eliciting P300 including the task, paradigm, stimulus factors, and software may also affect the component values. Thus, it may be desirable to have normative data for all study centres. It would be also worthwhile to have standardized protocols for conducting ERP studies to make the comparison with reference values meaningful.
Identification of biomarkers may aid in the diagnosis of psychiatric disorders because clinical diagnosis can be highly error-prone, especially when it is based on cross-sectional information or short follow up. Recent evidence suggests that ERP P300 is an actionable biomarker in schizophrenia and can be utilized for differential diagnosis, predicting conversion to psychosis in clinically high risk population, predicting treatment response, and new treatment development.,, Furthermore, the use of ERP P300 in the detection of Alzheimer's dementia has also shown promising results. However, the application of ERP P300 in the real world setting has been limited due to lack of validation in different populations. Our normative data can be used to develop diagnostic biomarkers for psychiatric disorders.
The P300 amplitude was found to increase from anterior (Fz) to posterior (Pz) areas across all the four age groups studied. There was maximum P300 amplitude at Pz electrode, which is consistent with the previous data showing maximum P300 activity in parietocentral areas. Verleger et al., found P300 amplitude being affected by temporal–parietal junction integrity because its absence significantly reduces the component size over the parietal area, suggesting that it represents memory operations in temporal-parietal areas, activated by attentional resources.
Previous studies have found that ERP P300 latency increases from the anterior to the posterior scalp areas, i.e., from Fz (frontal), to Cz (central), and Pz (parietal) electrode sites. This anteroposterior increasing trend was observed only in the age group of 40–50 years in our study, whereas in all the other three age groups, there was a decreasing trend. Thus, the present study supports the view that P300 generation stems from frontal and temporal/parietal activations. P300 latency represents the index of classification speed, i.e., the processing time required before response generation, and is a measure of neural activity underlying the process of attention allocation and immediate memory.
Our study also found significant negative correlation of P300 amplitude with increasing age; i.e., peak amplitude was found to diminish with advancing age. Similar observations have been reported in the previous studies by Johnson  and Courchesne  who found a significant decrease in P300 amplitude over time in early and middle adolescence. This decrease in amplitude seen in late adulthood has been interpreted as reflecting a slowing of central information processing time, which occurs with decreased cortical arousal as well as inhibition associated with old age. The smaller P300 amplitude seen in the elderly has been attributed to a degenerative process in the brainstem. Moreover, the age-related decline of cholinergic neurotransmission may explain the smaller P300 amplitude in the elderly because acetylcholine has been found to increase the P300 amplitude.
There was significant negative correlation between age and P300 latency only at the Fz electrode, but not in other areas. This may be because of relative stability of P300 latency in later ages, as shown in the Johnson  study that found relatively abrupt changes around the age of 12 years, after which P300 latencies changed slightly and were essentially at the adult level. Our sample did not include children where changes in P300 latencies are expected. Furthermore, Anderer et al., found that the P300 latency and age relationship was curvilinear with an accelerated latency increase in the elderly subjects above 60 years of age. Moreover, the developmental trajectories of frontal and parietal P300 have been found to differ across lifespan, with declining parietal P300 with age and unaffected frontal P300. The P300 latency has been reported to increase with normal aging., However, our findings show a decrease in the frontal P300 latency with age, which implies an increase in the mental function speed. One possible reason could be the compensatory increased brain activity in frontal areas with increasing age, which is also known as compensation-related utilization of neural circuits hypothesis (CRUNCH).
The sample had underrepresentation of female subjects, which precluded generating separate normative data for males and females. Due to the retrospective nature of the study, it was not possible to control for several factors such as the circadian rhythm, season, food intake, and exercise, which could potentially affect the P300 components.
We generated a P300 amplitude and frequency normative database at Fz, Cz, and Pz for the Indian population. There was a trend towards a decreasing P300 amplitude across age groups, whereas no such trend was found for P300 latency.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3]