Glutamate levels in the anterior cingulate cortex in un-medicated first episode psychosis: a proton magnetic resonance spectroscopy study

Converging lines of evidence suggest that glutamatergic dysfunction may contribute to the pathophysiology of first episode psychosis. We investigated whether first episode psychosis patients free from all pharmacological treatments and illicit substances show cortical glutamatergic alterations. One-hundred and eleven volunteers including 65 healthy volunteers and 46 first episode psychosis patients free from all pharmacological treatments (28 drug naïve) underwent a proton magnetic resonance spectroscopy scan measuring glutamate levels in the bilateral anterior cingulate cortex. Symptom severity was measured using the Positive and Negative Syndrome Scale (PANSS) and cognition was measured using the Wechsler Adult Intelligence Scale (WAIS) digit symbol test. There were no differences in glutamate levels between patients and controls. These findings remained unchanged when adjusting for the effects of age, sex and ethnicity or when restricting the analyses to patients who were both medication naïve to all pharmacological treatments and illicit substances. Whilst these findings do not preclude glutamatergic alterations in psychosis, methodological advances are needed for us to investigate whether patients show alterations in other aspects of glutamate function, such as pre-synaptic glutamate or release.


Metabolite levels in healthy volunteers and first episode psychosis patients Glutamine
Sixty glutamine datasets exceeding the Cramér-Rao lower bounds ratio was excluded. There was no main effect of group on glutamine levels when adjusting for scanner (F(1,50)=0.27, p=0.61), and this remained the case after adjusting for the effects of scanner, age, sex and ethnicity (F(1,33)=0.56, p=0.46). There was also no main effect of group on glutamine levels when adjusting for scanner and restricting the analysis to medication naïve patients who were free from illicit substances (F(1,26)=0.26, p=0.62) or when restricting the analysis to medication naïve patients, free from all illicit substances who also met the diagnostic criteria for schizophrenia (F(1,22)=0.33, p=0.57).

GLX
No GLX datasets exceeded the Cramér-Rao lower bounds ratio. There was no main effect of group on GLX levels (F(1,110)=1.92, p=0.17) and this remained the case after adjusting for the effects of scanner, age, sex and ethnicity (F(1, 84)=1.53, p=0.22). There was also no main effect of group on GLX levels when adjusting for the effects of scanner and restricting the analysis to medication naïve patients who were also free from illicit substances (F(1,71)=8.56, p=0.14) or when restricting the analysis to medication naïve patients, free from all illicit substances who also met the diagnostic criteria for schizophrenia (F(1,67)=1.45, p=0.23).

NAA
No NAA datasets exceeded the Cramér-Rao lower bounds ratio. There was no main effect of group on NAA levels (F(1,106)=0.95, p=0.33), and this remained the case after adjusting for the effects of scanner, age, sex and ethnicity (F(1, 84)=0.78, p=0.38). There was no main effect of group on NAA levels when adjusting for the effects of scanner and restricting the analysis to medication naïve patients who were also free from illicit substances (F(1,71)=1.16, p=0.29) or when restricting the analysis to medication naïve patients, free from all illicit substances who also met the diagnostic criteria for schizophrenia (F(1,67)=2.74, p=0.10).

Relationship between metabolite levels and symptom severity Total symptom severity scores
There was no relationship between glutamate levels and PANSS total symptom severity scores when adjusting for scanner (ß=-1.24, SE=1.85, p=0.51, R 2 =0.24) and this remained the case when restricting the analysis to patients who were both medication naïve and not taking illicit substances (ß=-0.98, SE=2.69, p=0.72, R 2 =0.11).
There was also no relationship between glutamate levels and PANSS total symptom severity scores when adjusting for scanner differences, age, sex and ethnicity (ß=-1.02, SE=2.12, p=0.63, R 2 =0.22).

Negative symptom severity scores
There was no relationship between glutamate levels and PANSS negative symptom severity scores (ß=-0.01, SE=0.63, p=0.99, R 2 =0.28) and this remained the case when restricting the analysis to patients who were both medication naïve and not taking illicit substances (ß=-0.52, SE=1.04, p=0.62, R 2 =0.08). There was also no relationship between glutamate levels and PANSS negative symptom severity scores when adjusting for the effects of scanner, age, sex and ethnicity (ß=-0.17, SE=0.66, p=0.80, R 2 =0.26).

General symptom severity scores
There was no relationship between glutamate levels and general symptom severity scores (ß=0.15, SE=0.93, p=0.87, R 2 =0.15), and this remained the case when restricting the analysis to patients who were both medication naïve and not taking illicit substances (ß=-0.01, SE=1.36, p=0.99, R 2 =0.03). There was also no relationship between glutamate levels and PANSS general symptom severity scores when adjusting for scanner differences, age, sex and ethnicity (ß=0.45, SE=1.06, p=0.67, R 2 =0.14).
Patients also showed no association between glutamate levels and cognitive function, as determined by the digit symbol coding test (ß=-0.16, SE=0.29, p=0.60, R 2 =0.01), and this remained the case when controlling for scanner differences, age, sex and ethnicity (ß=-.13, SE=0.32, p=0.69, R 2 =0.06). There was no association between glutamate levels and the digit symbol coding test when restricting the analysis to patients who were both medication naïve and not currently using illicit substances for scanner 1 (ß=0.41, SE=0.51, p=0.44, R 2 =0.07) and scanner 2 (ß=0.41, SE=0.51, p=0.44, R 2 =0.07). There was no association between glutamate levels and the digit symbol coding test when restricting the analysis to patients who were both medication naïve and not currently using illicit substances for scanner 1 (ß=-0.07, SE=0.38, p=0.85, R 2 =0.001) and scanner