Introduction

Schizophrenia represents a chronic disorder affecting nearly 1% of the world’s population. Schizophrenia appearing in less than 18 years of age is referred early onset schizophrenia (EOS). EOS have more severe premorbid neurodevelopmental abnormalities, worse long-term outcome, more cytogenetic anomalies, and potentially greater loading of family histories for schizophrenia and associated spectrum disorders than their adult counterparts^1,2. Previous network-oriented analysis study has been reported reduced functional connectivity in sensory regions, default mode, frontoparietal and cingulate-opercular network^3,4 in EOS. Altered dALFF in the right middle temporal gyrus (MTG)³ has also been reported. We used whole-brain effective connectivity (EC) model (comprising whole brain integration, segregation, metastability and intrinsic ignition) to interpret the brain dynamics in 16 patients of EOS to understand time-dependent communicability and propagation of neuronal activity within the brain state.

Materials and Methods

Sixteen EOS patients (mean age 17.68±2.15 years; 6F/10M) who met ICD-10 (code F20) criteria with positive (hallucinations, delusions, altered thoughts) and negative (social withdrawal, loss of speech) symptoms were recruited from the psychiatric clinics. Ten healthy subjects (age=20.9±0.99, 4F/6M) with no history of neurological and psychiatric disorder were recruited from local community.
Data was acquired on a 3T MRI (Ingenia, Philips) using 32-channel head coil after obtaining informed consent. Functional MRI volumes were acquired using echo-planar T2* -weighted sequence with 35 slices of 4 mm thickness, 210 dynamics, TE/TR:30/2000 ms, FOV = 250 mm, flip angle =90. Tl weighted image consisting of 350 slices was also acquired.
Resting data was analyzed with MELODIC and Structural Connectivity Matrix computed with FMRIB’s Diffusion Toolbox (www.fmrib.ox.ac.uk/fsl). Functional resting state atlas was used for parcellation to obtain the BOLD time series of the 214 cortical and subcortical brain areas in each individual's native EPI space, their time series extracted⁵ and a group structural connectivity (SC) mask was obtained by averaging the all-control subjects' SC matrix and applying a threshold of 80% to maintain the top 20% of strongest connections to binarize the SC⁶ for the computation of EC.

Result

Altered causal interactions of whole-brain effective connectivity was observed in EOS patients. We observed decreased whole brain integration (p=.021), increased whole brain segregation (p=.016) (Figure 1a, b), decreased whole brain metastability (p=.027) (Figure 1c) and low whole brain intrinsic ignition (p=.027) (Figure 1d) in EOS as compared to controls.
In-silico perturbation applied to individual ROIs, spread along the network and the effect of regional perturbations on the rest of the network were estimated. The finding showed an altered global integration of local endogenous events in early onset schizophrenia as compared to controls. Global brain responses (Figure 2b) undergo a transient peak (immediately after the initial perturbations) and then decay as the effects of the stimuli dilute with time and the system relaxes back to its stationary state. This relaxation is also observed by the homogenization of the response matrices during the longer latencies (Figure 3a). The global response curves exhibited different pattern in controls and EOS.

Discussion

Neural propagation of endogenous and exogenous perturbations in the brain was studied using model-free and model-based analysis methods, to elucidate the mechanisms behind early onset schizophrenia. Reduced functional connectivity in sensory regions, default mode network (DMN), frontoparietal network and the cingulate-opercular network has been reported^4,7. Altered dALFF was observed in bilateral precuneus, right middle temporal gyrus (MTG)³ in EOS. But the understanding of the mechanisms behind EOS is still missing.
We studied the propagation of endogenous and in-silico exogenous perturbations in patients with EOS, based upon directed and causal interactions estimated from resting-state fMRI. This provides time-dependent communicability and propagation of neuronal activity within the brain state. The method employed to investigate the directional causal interactions between brain regions, thus elucidating alterations in the broadcasting and the integrating capacities of individual areas or pathways, between normal control and EOS patients. These results show that patients with EOS have a decreased capacity for neural propagation and responsiveness to brain states.

Conclusion

The cerebral capacity of propagation of neuronal activity and integration of global and local (naturally occurring) events is affected in EOS, with respect to that in healthy controls.

References

1. Keshavan, M.S et al. Journal of psychiatric research. 1994;28 (3): 239-265.

2. Cannon, T.D. et al. Schizophrenia bulletin. 2003; 29 (4): 653-669.

3. Li, Q. et al. Frontiers in neuroscience. 2020; 14: 901

4. Wang, X. et al. Schizophrenia research. 2019; 208: 308-316

5. Finn, E.S. et al. Nature neuroscience. 2015; 18(11): 1664-1671.

6. Gilson, M. et al. NeuroImage. 2019; 201: 116007.

7. Jiang, S. et al. Journal of Central South University. Medical sciences 2010; 35(9): 947-951