CSF kynurenic acid and suicide risk in schizophrenia spectrum psychosis
Andreas Carlborga,n, Jussi Jokinena, Erik G. J¨ onssona, Sophie Erhardtb, Peter Nordstr¨ oma
aDepartment of Clinical Neuroscience, Psychiatry Section, Karolinska Institutet, Karolinska University Hospital Solna, SE-171 76 Stockholm, Sweden
bDepartment of Physiology and Pharmacology, Karolinska Insitutet, Stockholm, Sweden
a r t i c l e i n f o
Received 15 May 2011
Received in revised form
19 August 2012
Accepted 23 August 2012
a b s t r a c t
Relationships between concentrations of cerebrospinal fluid (CSF) kynurenic acid (KYNA) and suicidal
behavior were evaluated in 59 patients with psychosis after 22 years of follow-up . Three patients died
from suicide and nine patients had a history of attempted suicide. Patients with attempted suicide had
significantly lower concentrations of CSF KYNA.
& 2012 Elsevier Ireland Ltd. All rights reserved.
Schizophrenia is a devastating disorder with a high suicide risk
(Carlborg et al., 2010). Brain dopamine hyperactivity has been the
predominating hypothesis of schizophrenia but in recent years
also other hypotheses including aberrations in glutamatergic
signaling has received considerable attention. Thus compounds
blocking the glutamatergic N-methyl-D-aspartate (NMDA) recep-
tor, e.g. phencyclidine (PCP), induce both positive and negative
symptoms as well as cognitive deficits in healthy humans and
aggravates psychotic symptoms in patients with schizophrenia
(Jentsch and Roth, 1999).
Kynurenic acid (KYNA), an endogenous tryptophan metabolite,
antagonizes both the NMDA receptor (Kessler et al., 1989) and the
cholinergic alpha7nnicotinic receptor (Hilmas et al., 2001) in the
brain. Elevated concentrations of KYNA have been found in the
cerebrospinal fluid (CSF) (Nilsson et al., 2005) and in the post-
mortem prefrontal cortex (Sathyasaikumar et al., 2011) of
patients with schizophrenia. Interestingly, CSF KYNA has been
correlated positively with CSF 5-hydroxyindoleacetic acid (5-
HIAA) (Nilsson et al., 2007), the major metabolite of serotonin,
which has been associated with suicidal behavior among patients
with major depression (˚Asberg, 1997). A recent study analyzed 20
CSF biomarkers among suicide attempters without finding KYNA
to load on any specific factor in a principal component analysis
(Lindqvist et al., 2011). In another report plasma concentration of
kynurenin, the immediate precursor of KYNA was higher among
suicide attempters with major depressive disorder than non-
attempters (Sublette et al., 2011).
The objective of the present study was to investigate the
relationships between CSF KYNA concentrations and suicidal
behavior in a long-term cohort of patients with schizophrenia
spectrum psychosis. We also investigated whether CSF KYNA
concentrations were correlated with severity of disease as mea-
sured by the proportion of days admitted to a psychiatric
inpatient clinic during time of follow-up .
The Regional Ethical Review Board in Stockholm approved the study (2006/
408-31/4). The samples consisted in part of patients from a cohort earlier
described (Carlborg et al., 2009).
The present sample consisted of 59 patients (35 men, 24 women) admitted
and hospitalized between the year 1979 and 1987 at onset of illness with
symptoms of psychosis, i.e. hallucinations, delusions or disorganized behavior
and from whom CSF was collected. Age (mean7standard deviation (S.D.)) at
index admission was 31.278.1 years, range 18.1–51.4 years without any
significant difference between men (30.577.7 years) and women (32.278.7
years). Mean follow-up time was 22.175.5 years (men 22.375.2 vs. women
CSF was obtained by lumbar puncture in a standardized manner (Sedvall and
Wode-Helgodt, 1980). The determination of CSF KYNA concentrations was
performed using an isocratic reversed-phase high performance liquid chromato-
graphy (HPLC), as previously described (Nilsson et al., 2005).
CSF samples were coded, stored and blindly analyzed with regard to KYNA
Information about discharge diagnosis, length and frequency of hospital
admissions were retrieved from the Swedish Inpatient Register. A lifetime main
clinical diagnosis based on International Classification of Diseases (ICD 8–10 ) was
given to each patient according to a hierarchy earlier described (Vares et al., 2006).
All patients were searched in the Central Bureau of Statistics to identify those
who had died. Reported causes of death were obtained from the Causes of Death
Contents lists available at SciVerse ScienceDirect
journal homepage: www.elsevier.com/locate/psychres
0165-1781/$-see front matter & 2012 Elsevier Ireland Ltd. All rights reserved.
nCorresponding author. Tel.: þ46 739660658; fax: þ46 859033857.
E-mail addresses: firstname.lastname@example.org,
email@example.com (A. Carlborg).
Psychiatry Research 205 (2013) 165–167
Register at the National Board of Health and Welfare. Time of follow-up ended in
July 2006 or at time for death. The available medical records were scrutinized for
information about previous suicide attempts.
In initial analyses evaluating skewness and kurtosis with Shapiro–Wilks test, CSF
KYNA concentrations were not normally distributed and therefore transformed into
normal distribution using the natural logarithms. As CSF KYNA concentrations
correlated with age (r¼0.30, P¼0.02), all group comparisons of CSF KYNA values
were adjusted for age using linear regression. Po0.05 was considered significant.
JMP 7, SAS Institute inc., was used for all statistical analyses.
The patients received a main lifetime diagnosis of schizophrenia
(n¼31), psychosis not otherwise specified (n¼14), schizoaffective
disorder (n¼4), schizophreniform disorder (n¼3) and delusional
disorder (n¼1). Six patients received a lifetime main diagnosis
outside the non-organic psychotic spectrum. None of the patients
had a diagnosis of co-morbid mood disorder or substance abuse.
Mean number of days admitted to a psychiatric inpatient clinic was
44.2764.3 day/year for the patients, similar in men (44.6756.3)
and women (43.7775.7 years). Four patients were drug-naı ¨ve,
29 were drug free and 26 were on neuroleptic medication at time
of lumbar puncture. There was no significant difference in mean
CSF KYNA concentrations between these groups (1.8271.36,
1.3470.78, and 1.6271.41 nM, respectively).
Of the 59 patients three (5%), one man and two women died
from suicide during the follow-up by jumping from high places or
drowning. Time from lumbar puncture to suicide was 8.478.2
years, range 3.4–17.9 years. Medical records for 39 of the patients
were possible to retrieve. From analysis of these records nine
patients (23%) had a history of attempted suicide at initial
admission or during follow-up .
Fig. 1 shows CSF KYNA concentrations in suicide attempters
and non-attempters adjusted for age. Mean CSF KYNA concentra-
tions were significantly lower in suicide attempters compared to
patients lacking history of suicide attempt (0.8370.49 and
1.6871.42 nM, respectively; F¼7.5, P¼0.009). CSF KYNA con-
centrations were positively correlated with CSF 5-HIAA concen-
trations (r¼0.64, Po0.001).
The correlation between CSF KYNA concentrations and number of
days hospitalized was not significant (p¼0.8). No significant
difference in mean CSF KYNA concentrations was found between
the diagnostic groups or gender (men 1.3270.71, women 1.7671.4;
In the present study psychosis patients with a history of
attempted suicide had significantly lower CSF KYNA concentra-
tions at the onset of illness. CSF KYNA has been shown to
correlate with CSF 5-HIAA (Nilsson et al., 2007) and this was also
the case in our study. The reason for this correlation is unknown
but might be related to the availability of tryptophan, the
common precursor of both serotonin and KYNA (Nilsson et al.,
2007). Low concentrations of CSF 5-HIAA, a sign of dysfunction of
the serotonin system, have been associated with suicidal behavior
in mood disorders (˚Asberg, 1997) and a combination of CSF
biomarkers has been suggested as indicators of future suicidal
behavior (Lindqvist et al., 2011). However, a large study was
unable to find an association between low concentrations of CSF
5-HIAA and suicidal behavior in patients with schizophrenia
(Carlborg et al., 2009).
Proportion of days admitted to a psychiatric inpatient clinic was
used to correlate CSF KYNA concentrations with the severity of
disease. No such correlation was found, suggesting that this proxy
for disease severity is not influenced by CSF KYNA concentrations.
The present study is limited by the small sample size. Further-
more, we were not able to retrieve the medical records for all
patients. Thus, information about prior suicide attempt may be
lacking. The use of number of days admitted to an inpatient clinic
may not fully account for the complexity of disease severity.
In conclusion, mean concentration of CSF KYNA at onset of
illness was significantly lower in schizophrenia spectrum psycho-
sis patients with a history of attempted suicide later in life.
The neurochemical correlates of suicidal behavior need further
Financial report was provided through the regional agreement
on medical training and clinical research between Stockholm
County Council and the Karolinska Institutet, the Swedish Research
Council (nos. K2007-62X-15078-04-3, K2008-62P-20597-01-3,
K2009-61P-21304-04-4, and K2009-61X-21305-01-1), and the
Wallenberg Foundation and the HUBIN project.
˚Asberg, M., 1997. Neurotransmitters and suicidal behavior. The evidence from
cerebrospinal fluid studies. Annals of the New York Academy of Sciences 836,
Carlborg, A., Winnerb¨ ack, K., J¨ onsson, E.G., Jokinen, J., Nordstr¨ om, P., 2010. Suicide
in schizophrenia. Expert Review of Neurotherapeutics 10, 1153–1164.
Hilmas, C., Pereira, E.F., Alkondon, M., Rassoulpour, A., Schwarcz, R., Albuquerque,
E.X., 2001. The brain metabolite kynurenic acid inhibits alpha7 nicotinic
receptor activity and increases non-alpha7 nicotinic receptor expression:
physiopathological implications. Journal of Neuroscience 21, 7463–7473.
Jentsch, J.D., Roth, R.H., 1999. The neuropsychopharmacology of phencyclidine:
from NMDA receptor hypofunction to the dopamine hypothesis of schizo-
phrenia. Neuropsychopharmacology 20, 201–225.
Kessler, M., Baudry, M., Lynch, G., 1989. Quinoxaline derivatives are high-affinity
antagonists of the NMDA receptor-associated glycine sites. Brain Research
Lindqvist, D., Janelidze, S., Erhardt, S., Tr¨ askman-Bendz, L., Engstr¨ om, G., Brundin,
L., 2011. CSF biomarkers in suicide attempters—a principal component
analysis. Acta Psychiatrica Scandinavica 124, 52–61.
Nilsson, L.K., Linderholm, K.R., Engberg, G., Paulson, L., Blennow, K., Lindstr¨ om,
L.H., Nordin, C., Karanti, A., Persson, P., Erhardt, S., 2005. Elevated levels of
kynurenic acid in the cerebrospinal fluid of male patients with schizophrenia.
Schizophrenia Research 80, 315–322.
40 4550 55
Log CSF KYNA
Fig. 1. Log CSF KYNA concentrations (nM) adjusted for age in patients with
schizophrenia spectrum psychosis. The thin line and bold lines show mean CSF
KYNA concentrations in non-attempters (’) and attempters (J), respectively.
Asterix (n) indicates suicides. One suicide had a suicide attempt history.
A. Carlborg et al. / Psychiatry Research 205 (2013) 165–167
Nilsson, L.K., Nordin, C., J¨ onsson, E.G., Skogh, E.G., Erhardt, S., 2007. Cerebrospinal
fluid kynurenic acid in male patients with schizophrenia—correlation with
monoamine metabolites. Acta Neuropsychiatrica 19, 45–52.
Sathyasaikumar, K.V., Stachowski, E.K., Wonodi, I., Roberts, R.C., Rassoulpour, A.,
McMahon, R.P., Schwarcz, R., 2011. Impaired kynurenine pathway metabolism
in the prefrontal cortex of individuals with schizophrenia. Schizophrenia
Bulletin 37, 1147–1156.
Sedvall, G.C., Wode-Helgodt, B., 1980. Abberant monoamine metabolite levels in
CSF and family history of schizophrenia. Their relationships in schizophrenic
patients. Archives of General Psychiatry 37, 1113–1116.
Sublette, M.E., Galfalvy, H.C., Fuchs, D., Lapidus, M., Grunebaum, M.F., Oquendo,
M.A., Mann, J.J., Postolache, T.T., 2011. Plasma kynurenine levels are elevated
in suicide attempters with major depressive disorder. Brain, Behavior, and
Immunity 25, 1272–1278.
Vares, M., Ekholm, A., Sedvall, G.C., Hall, H., J¨ onsson, E.G., 2006. Characterisation of
patients with schizophrenia and related psychosis: evaluation of different
diagnostic procedures. Psychopathology 39, 286–295.
A. Carlborg et al. / Psychiatry Research 205 (2013) 165–167