Abstract
Background
Subjects with a diagnosis of schizophrenia (Scz) overweight unexpected evidence in probabilistic inference. A neurobiological explanation for this effect is that cortical disinhibition (e.g. hypofunction of interneuron N-methyl-D-aspartate receptors) in Scz makes neural network states (or ‘attractors’) unstable and easily altered: in predictive coding, this means reduced precision of prior beliefs. Indeed, previous work modelling EEG data in the mismatch negativity (an archetypal predictive coding paradigm) showed prefrontal disinhibition in both Scz and their relatives.
Methods
Hierarchical Bayesian belief updating models were tested in two independent datasets (n=80 – published previously – and n=167) comprising subjects with schizophrenia, and both clinical and non-clinical controls (some tested when unwell and on recovery) performing the ‘probability estimates’ version of the beads task (a probabilistic inference task). Models with a standard learning rate or including a parameter encoding the effects of attractor instability were formally compared.
Results
The model simulating unstable attractors had most evidence in all groups in both datasets. Two of its four parameters differed between Scz and non-clinical controls in each dataset: attractor instability (p=0.01 and p=0.00004 in datasets 1 and 2) and response stochasticity (p=0.002, p=0.0007). These parameters correlated in both datasets (ρ=-0.38, p=0.0004; ρ=-0.35, p=0.000001). The clinical controls showed similar parameter distributions to Scz when unwell, but were no different to controls once recovered.
Conclusions
These findings support the hypothesis that attractor network instability - or in predictive coding terms, greater prior variance - contributes to belief updating abnormalities in Scz, and raise questions about potential glutamatergic or other neuromodulatory mechanisms.
Subjects with a diagnosis of schizophrenia (Scz) overweight unexpected evidence in probabilistic inference. A neurobiological explanation for this effect is that cortical disinhibition (e.g. hypofunction of interneuron N-methyl-D-aspartate receptors) in Scz makes neural network states (or ‘attractors’) unstable and easily altered: in predictive coding, this means reduced precision of prior beliefs. Indeed, previous work modelling EEG data in the mismatch negativity (an archetypal predictive coding paradigm) showed prefrontal disinhibition in both Scz and their relatives.
Methods
Hierarchical Bayesian belief updating models were tested in two independent datasets (n=80 – published previously – and n=167) comprising subjects with schizophrenia, and both clinical and non-clinical controls (some tested when unwell and on recovery) performing the ‘probability estimates’ version of the beads task (a probabilistic inference task). Models with a standard learning rate or including a parameter encoding the effects of attractor instability were formally compared.
Results
The model simulating unstable attractors had most evidence in all groups in both datasets. Two of its four parameters differed between Scz and non-clinical controls in each dataset: attractor instability (p=0.01 and p=0.00004 in datasets 1 and 2) and response stochasticity (p=0.002, p=0.0007). These parameters correlated in both datasets (ρ=-0.38, p=0.0004; ρ=-0.35, p=0.000001). The clinical controls showed similar parameter distributions to Scz when unwell, but were no different to controls once recovered.
Conclusions
These findings support the hypothesis that attractor network instability - or in predictive coding terms, greater prior variance - contributes to belief updating abnormalities in Scz, and raise questions about potential glutamatergic or other neuromodulatory mechanisms.
| Original language | English |
|---|---|
| Title of host publication | Biological Psychiatry |
| DOIs | |
| Publication status | Published - 1 May 2018 |
| Event | Society for Biological Psychiatry - New York, United States Duration: 1 May 2018 → … https://sobp.org/meetings/2018-annual-meeting/ |
Conference
| Conference | Society for Biological Psychiatry |
|---|---|
| Country/Territory | United States |
| City | New York |
| Period | 1/05/18 → … |
| Internet address |