This study will investigate whether a single dose of diazepam (5mg) compared to placebo can modulate brain chemistry (GABA/glutamate levels) and function (blood flow, neural response and connectivity during tasks and at-rest) in 24 individuals at clinical high-risk for psychosis.

The pathophysiology of psychosis involves elevated subcortical dopamine function, but the factors driving this are still unclear. Evidence from a neurodevelopmental animal model of psychosis suggests that this arises through a pathway linking psychosocial stress, corticolimbic hyperresponsivity, and GABA/glutamate imbalance. In response to stress/negative emotion, amygdala hyperresponsivity decreases GABA interneuron function in the hippocampus through strong direct projections. Decreased hippocampal GABA function leads to disinhibition of hippocampal pyramidal cells, elevating local activity and glutamate levels. Increased output from the hippocampus to the striatum elevates dopamine release in the striatum, and increases the firing of dopaminergic neurons in the midbrain. These neurobiological effects are associated with cognitive (e.g., working memory) and emotional deficits (e.g., increased anxiety). Moreover, peripubertal (premorbid) administration of benzodiazepines at anxiolytic doses to this animal of psychosis is shown to normalise hippocampal activity, thereby preventing the emergence of striatal hyperdopaminergia and associated behavioural abnormalities in adulthood. Collectively, these findings indicate that GABA dysfunction and emotional hyperresponsivity may play a critical role in the development of psychosis in humans, and suggest that clinical interventions targeting this pathway have the potential to reduce the risk of developing the disorder.

This study will use multimodal neuroimaging (MRS, ASL, rs-fMRI, tb-fMRI) to assess whether the acute administration of a benzodiazepine can modulate the pathway linking corticolimbic response and GABA/glutamate levels in people in the premorbid stage of psychosis (at "clinical high risk", CHR). Using a randomised, double-blind, placebo-controlled, crossover design, 24 CHR-P participants will undergo two MRI sessions, once under an acute oral dose of diazepam (5 mg; generic) and once under oral placebo (50 mg ascorbic acid), with a minimum 3-week washout period between visits.

Inclusion Criteria:

• Age range 18-40 years
• Capacity to consent to participation in the study
• Inclusion into one of three groups as assessed by the CAARMS: i) genetic vulnerability group, ii) attenuated psychosis group, iii) brief intermittent psychosis symptoms group. This instrument has been modified to additionally allow the scoring of the SIPS v.520. The scoring of the SIPS v.5 is included for comparative purposes and does not constitute inclusion criteria.
• Inclusion based on meeting criteria for "basic symptoms" which are assessed using the Schizophrenia Proneness Instrument (SPI-A)21

Exclusion Criteria:

• History of neurological disorders
• Current exposure to any drug with potential GABAergic or glutamatergic effects other than antipsychotics, mood stabilisers, antidepressants. This includes opiates, psychostimulants, benzodiazepines, atomoxetine, memantine, ketamine, cough medication containing dextromethorphan
• Current or past exposure to any antipsychotic medication
• Pregnancy/breastfeeding
• Contra-indication to MRI scanning (e.g., metal in body, such as pacemakers or implants, claustrophobia)
• IQ < 70 as determined with the shortened version of the Wechsler Adult Intelligence Scale III (WAIS-III)22