Antipsychotic medication and
longitudinal changes in brain
volume in schizophrenia
- meta-analysis and results from the
Northern Finland 1966 Birth Cohort Study
Jouko Miettunen
jouko.miettunen@oulu.fi
Department of Psychiatry, University of Oulu, Finland
Conflicts of interest:
None
SYSTEMATIC REVIEW AND META-ANALYSIS
 Relatively little is known on factors associating with
longitudinal changes in brain morphometry in
schizophrenia after the illness onset.
 Our aim was to systematically review longitudinal MRI
studies with at least two-year scan-interval on the relation
between the dose or type of antipsychotic medication and
brain morphometric changes in schizophrenia and related
psychoses.
Previous reviews
 decreases in grey matter and global brain volume and
increments in cerebrospinal fluid (CSF) or ventricular
volume during drug treatment
 contradictory findings when comparing typical and atypical
medications
 basal ganglia volume increment after treatment with
especially typical antipsychotics
 have included also short follow-ups!
Moncrieff & Leo (Psychol Med 2010); Navari & Dazzan (Psychol Med 2009);
Smieskova et al. (Curr Pharm Des 2009); Shepherd et al. (Neurosci Biobehav Rev
2012); Fusar-Poli et al. (Neurosci Biobehav Rev 2013)
meta-analysis
 Studies were systematically collected using the
databases of PubMed, Scopus, Web of Knowledge,
and PsycINFO (1223 hits).
 We calculated correlations between antipsychotic
dose and brain volume changes.
 2-year scan-interval was required
 Results reported only as “non-significant” were also
included (correlation = zero)
meta-analysis
brain regions (12)
 total brain, cerebrum, frontal lobe, temporal lobe,
parietal lobe, occipital lobe, cerebellum, limbic area,
basal ganglia, pituitary gland, paracentral lobule, and
CSF and ventricles.
brain tissue types (4)
 grey matter (GM), white matter (WM), CSF, and
volumes
Sample (13 samples)
followup
brain areas
Iowa Longitudinal Study, USA (FE: n=23-211)
2-7 years
several areas
Westmoreland Corson et al. 1999; Ho et al. 2003; McCormick et al. 2005; Ho et al. 2007; Ho et al. 2011
Utrecht Schizophrenia Project, Netherlands (PT: n=90-105)
4.8 years
several areas
2.5 years
caudate nucleus L/R
Boston, MA, USA (PT: n=23)
Davidson et al. 2012
2.9 years
septum pellucidum, cavum septi pellucidum
Chicago, IL, USA (PT: n=20)
Cobia et al. 2012
2 years
several areas
Philadelphia, PA, USA (FE: n=20, PT: n=20)
Gur et al. 1998
2.6 years
whole brain, CSF, frontal lobe, temporal lobe
Melbourne, Australia (FE: n=23, PT: n=11)
Takahashi et al. 2009
Chiba, Japan (0.2 Tesla) (PT: n=15)
Saijo et al. 2001
2 years
whole brain, long and short insular cortex L/R
10 years
lateral ventricles
Toyama, Japan (FE: n=18-20)
Takahashi et al. 2010, 2011a, 2011b, 2013
2.7 years
several areas
Toyama, Japan (PT: n=17)
Takahashi et al. 2012
2.2 years
pituitary gland
Sibling-pair sample, Utrecht, Netherlands (PT: n=10)
Brans et al. 2008
5 years
several areas
van Haren et al. 2007, 2008, 2011; Collin et al. 2012
Prospective Longitudinal Study of Schizophrenia and the Mental
Health Clinical Research Center, Iowa, USA (PT: n=14)
Heitmiller et al. 2004
FE = first episode, PT = previously treated
meta-analysis
 Most (83%) of the reported correlations were
statistically non-significant.
 In the studies with significant findings, use of
antipsychotics more often associated with
decrease than increase of brain volumes,
even in the very few studies taking into
account illness severity.
meta-analysis by tissue type and brain area combinations
studies
n
GM
- cerebrum
WM
- cerebrum
CSF
- CSF and
ventricles
volume
- total
- frontal lobe
- temporal lobe
- cerebellum
random effect meta-analysis
correlation (95% CI)
z test
sign.
3
-0.02 (-0.30 - 0.27)
0.12
0.901
3
0.04 (-0.07 - 0.16)
0.72
0.471
4
0.00 (-0.11 - 0.11)
0.00
0.998
3
5
9
0.00 (-0.15 - 0.15)
-0.12 (-0.41 - 0.16)
-0.02 (-0.14 - 0.10)
0.00
0.86
0.34
>0.999
0.389
0.731
3
0.03 (-0.08 - 0.14)
0.47
0.639
Statistically significant findings
Cerebrum
*** atypical dose associated with less progressive decrease of GM (Brans et al. 2008)
* clozapine dose associated with reduction and typical and clozapine dose associated with reduction of GM (Ho et al. 2011)
Frontal lobe
*** total dose associated with reduction in first episode patients (Gur et al. 1998)
** clozapine associated with reduction of GM (Ho et al. 2011) and with increase of right superior and right and left medial
frontal cortex (van Haren et al. 2011), non-clozapine atypical dose associated with increase of right medial frontal cortex
(van Haren et al. 2011), and typical dose associated with reduction of left precentral cortex (van Haren et al. 2011)
* total, typical and nonclozapine atypical dose associated with reduction of GM (Ho et al. 2011), olanzapine and clozapine
increased volume in a cluster in superior frontal gyrys (van Haren et al. 2007)
Temporal lobe
*** total dose associated with reduction in first episode patients (Gur et al. 1998), total and atypical dose associated with less
reduction of right and left caudal superior temporal gyrys, total dose with less reduction of left planum temporale
(Takahashi et al. 2010) and of whole fusiform gyrus (Takahashi et al. 2011b)
** clozapine associated with decrease of GM (Ho et al. 2011)
* total dose associated with decrease of GM (Ho et al. 2011)
Parietal lobe
* total, clozapine and nonclozapine atypical dose associated with GM reduction and nonclozapine atypical dose with WM
increment (Ho et al. 2011)
^ clozapine correlated with reduction of sulcal CSF (Ho et al. 2011)
CSF and ventricles
Limbic system
Basal ganglia
*** typical dose associated with increase and atypical and risperidone with decrease of anterior cingulate cortex
(McCormick et al. 2005)
** clozapine associated with increase in right cingulate cortex (van Haren et al. 2011)
* clozapine associated with decrease of thalamus (Ho et al. 2011)
*** typical dose increased and atypical dose decreased globus pallidus (Westmoreland Corson et al. 1999), clozapine
increased putamen (Ho et al. 2011)
** typical dose increased putamen (Westmoreland Corson et al. 1999), total dose increased putamen (Ho et al. 2011)
* clozapine decreased and nonclozapine atypical increased caudate volume, typical dose and nonclozapine atypical dose
increased putamen (Ho et al. 2011), typical dose increased volume in selected voxels in caudate (van Haren et al. 2007)
***=large (r ≤ 0.5), **=moderate (0.3 ≤ r < 0.5), *=small (r > 0.1), ^=smaller than 0.1 correlation
meta-analysis
 In performed meta-analyses of specific brain areas no
significant associations were found.
 We did not find any clear differences between typical and
atypical exposure and brain volume change.
 Studies focusing on first episode patients were more likely
to find an effect than studies on previously treated patients.
 In the studies with significant findings in grey matter, high
antipsychotic dose was more frequently associated with
decrease (27% of correlations) than with increase (10%).
Northern Finland 1966 Birth Cohort
 All members of the general population-based
Northern Finland Birth Cohort 1966 with any
psychotic disorder were invited for a MRI brain scan
at the age of 34 years.
 A follow-up was conducted on average 9 years later.
 Brain scans at both time points were obtained from
33 subjects with schizophrenia.
 scan-interval antipsychotic cumulative dose associated with
total brain volume loss, even when adjusted with symptoms
(beta =-0.43, p=0.018).
SOFAS = Social and Occupational Functioning Assessment Scale,
PANSS = Positive and Negative Syndrome Scale
Veijola J, et al. Manuscript
conclusions
 Antipsychotic medication may contribute to changes in
brain structure in some areas of the brain.
 The current available data is inconclusive and therefore
more good quality long-term follow-up studies are needed
focusing on the possible association between antipsychotic
medication and brain structures.
 Covariates ?
 severity of illness, substance abuse, smoking, duration of
illness
 Combine data sets ?
Sanna Huhtaniska, BMed; Juha Veijola, prof.; Matti Isohanni, prof.; Erika
Jääskeläinen, MScD; Jouko Miettunen, PhD
 Department of Psychiatry, Institute of Clinical Medicine, University of Oulu,
Oulu, Finland & Department of Psychiatry, Oulu University Hospital, Oulu,
Finland
Noora Hirvonen, MA
 Information Studies, Faculty of Humanities, University of Oulu, Oulu,
Finland
Jukka Remes, MSc
 Department of Diagnostic Radiology, Oulu University Hospital, Oulu,
Finland
Graham K Murray, MScD
 Department of Psychiatry, University of Cambridge, Cambridge, UK
Email: jouko.miettunen@oulu.fi