Hand-preference and population schizotypy: a meta-analysis

Full text

Page 1

Hand-preference and population schizotypy: A meta-analysis
M. Somers⁎, I.E. Sommer, M.P. Boks, R.S. Kahn
Rudolf Magnus Institute of Neuroscience, Department of Psychiatry, B01.206, University Medical Centre Utrecht, PO Box 85500, 3508 GA Utrecht, The Netherlands
a r t i c l ei n f o a b s t r a c t
Article history:
Received 10 July 2008
Received in revised form 27 October 2008
Accepted 8 November 2008
Available online xxxx
Language functions in schizophrenia patients are represented more bilateral, i.e. less lateralized
than in healthy subjects. This decreased lateralization is also observed in individuals at
increased risk for schizophrenia. Language lateralization is related to handedness; in that left-
and mixed-handed individuals more frequently have decreased lateralization in comparison to
right-handed subjects. Population schizotypy can be considered part of the schizophrenia
spectrum disorders. In line with this, population schizotypy has repeatedly, though
inconsistently, been associated with left-handedness. In order to define the exact association
between handedness and schizotypy, we performed meta-analyses on the available literature.
We found that non-right-handed subjects, but not strong left-handers, had higher scores on
schizotypy questionnaires than right-handed subjects. Mixed-handers showed a trend towards
higher schizotypy in comparison to strong left-handers. It is argued that the higher schizotypy
in non-right-handed individuals reflects the higher incidence of bilateral language
lateralization in this group. Bilateral language organisation may underlie loosening of
association, possibly leading to higher schizotypy scores.
© 2008 Elsevier B.V. All rights reserved.
Keywords:
Schizotypy
Handedness
Laterality
Meta-analysis
1. Introduction
Non-right-handedness has been associated with schizo-
phrenia at least since the 1950s (Straaten, 1955). Indeed,
meta-analysis shows that the frequency of non-right-hand-
edness is almost twice as high in schizophrenia patients in
comparison to healthy subjects (Sommer et al., 2001a).
Handedness is related to cerebral lateralization: approxi-
mately 95% of right-handed subjects have left-cerebral
dominance for language function (Szaflarski et al., 2006)
and 5% have bilateral or right cerebral dominance. In non-
right-handed subjects 70% have left cerebral dominance and
30% show a right dominant or bilateral pattern of language
dominance (Pujol et al., 1999). Given the ease with which
handedness can be assessed, and the weak but consistent
association with cerebral lateralization, hand preference
seems a useful indirect reflection of cerebral dominance in
larger groups. The increased frequency of non-right-handed-
ness in schizophrenia patients has been attributed to a higher
incidence of bilateral language lateralization as compared to
healthy subjects. Functional imaging studies showed that this
was indeed the case (Sommer et al., 2001b, 2003, 2004; Li et
al., 2007a). The lower degree of language lateralization is
caused by increased language activity in frontal and temporal
areas of the right hemisphere (Sommer et al., 2001b, 2003).
Increased language activity of the right hemisphere was also
present in unaffected monozygotic co-twins of schizophrenia
patients (Sommer et al., 2004; Spaniel et al., 2007), indicating
that decreased language lateralization is a genetic predis-
position for schizophrenia. As such, decreased language
lateralization and possibly non-right-handedness may be
useful endophenotypes for psychosis. The association
between decreased cerebral dominance and psychosis has
also been demonstrated in patients with unipolar and bipolar
affective psychosis (Sommer et al., 2007) and subjects at high
genetic risk for psychosis (Li et al., 2007b). It is hypothesised
that more bilateral language representation facilitates magi-
cal and delusional ideas by means of the more diffuse
semantic activation in the right-hemisphere as compared to
the left (Leonhard and Brugger, 1998). Population schizotypy
(i.e. non-clinical subjects with a schizotypal tendency) can be
Schizophrenia Research xxx (2008) xxx–xxx
⁎ Corresponding author. University Medical Centre Utrecht, Department
of Psychiatry, B01.206, PO Box 85500, 3508 GA Utrecht, The Netherlands.
Tel.: +31 88 7556370; fax: +31 88 7555509.
E-mail address: M.Somers@umcutrecht.nl (M. Somers).
SCHRES-03752; No of Pages 8
0920-9964/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.schres.2008.11.010
Contents lists available at ScienceDirect
Schizophrenia Research
journal homepage: www.elsevier.com/locate/schres
ARTICLE IN PRESS
Please cite this article as: Somers, M., et al., Hand-preference and population schizotypy: A meta-analysis, Schizophr. Res.
(2008), doi:10.1016/j.schres.2008.11.010

Page 2

considered part of the schizophrenia spectrum. For example,
high scores on schizotypy questionnaires were demonstrated
to predict transformation to psychosis in subjects at high
genetic risk (Johnstone et al., 2005). It can be hypothesized
that these subjects with high schizotypal tendencies alsohave
decreased language lateralization. As of yet, no functional
imaging on language lateralization has been performed in
these subjects, though some support for decreased lateraliza-
tion is provided by dichotic listening studies (Broks et al.,
1984a,b; Rawlings and Borge,1987; Poreh et al.,1994). Hand-
preference in population schizotypy, however, has been
investigatedextensively. Severalstudies have assessedpossible
associations between mixed-, left- and non-right-handedness
and higher scores on various population schizotypy question-
naires (Raine, 1991; Kim et al., 1992; Preti et al., 2007; Annett
and Moran, 2006). Evidence is heterogeneous and at times
contradictory (Dragovic et al., 2005), which precludes any
definiteconclusionsonhandednessinpopulationschizotypy.In
order to elucidate the evidence for a possible association
between hand-preference and schizotypal tendency, we per-
formed meta-analyses on the present literature. Primarily, we
aimed to investigate whether schizotypy is elevated in non-
right-handers in comparison to right-handers. Secondarily, we
investigated whether there is a difference between mixed
handedness in comparison to both strong right- and left-
handedness.
2. Methods
2.1. Search criteria
Searches were performed in PubMed and Embase data-
bases up to December 2007. In addition, reference lists from
journal articles were searched for cross-references. Combina-
tions of the following search terms were used: ‘(left-)
handedness’, ‘hand preference’, ‘magical ideation’, ‘schizotyp
(y)(al)’, and ‘psychosis proneness’. Further, we combined
‘schizotypal personality disorder’ and ‘functional laterality’ as
MESH terms. Only journals in English were searched. By
convention, we required at least three suitable studies to
perform a meta-analysis or sub-analysis.
2.2. Inclusion criteria
Papers were included for meta-analysis if they met the
following criteria.
A) Subjects should be at least 11 years old, since the use of
schizotypy scales has not been validated before that
age (Cyhlarova and Claridge, 2005). Hand preference
canreliablybeinvestigated from thatage(Brydenet al.,
2000). Studies of subjects with a mean age above 65
were excluded, since the effect of aging itself might
blur results.
B) Hand-preference should be measured either by self-
report, observation or by a validated Hand Preference
Questionnaire such as the Edinburgh Handedness
Inventory (EHI) (Oldfield, 1971) or the Annett Hand
Preference Questionnaire (AHPQ) (Annett, 1970).
C) The usage of a schizotypy scale with (the possibility to
reconstruct) mean test scores and standard deviation
per hand-preference group OR available exact F, t or p
values.
D) Investigated individuals were not selected on basis of a
special condition possibly related to language laterali-
zation or handedness (such as dyslexia, schizophrenia
or epilepsy). Control subjects for studies investigating
such subjects were allowed.
2.3. Combination of measurement methods
2.3.1. Handedness
Hand-preference can be dichotomously dissected in right-
handedness (RH) and non-right-handedness (NRH), with
non-right-handedness including left- (LH), mixed- (MIX)
and even weak right-handers (McManus, 1985) or hand-
preference can be considered a trichotomous model compris-
ing right-, mixed- and left-handedness (Peters, 1998). In
addition, further sub-classifications such as Annett's 7 or 8-
group exist (Annett, 2004). In our meta-analysis we investi-
gated associations between schizotypy and handedness using
both a dichotomous and trichotomous model.
2.3.2. A RH/NRH dichotomy
For the analysis of right- vs. non-right-handedness we
required studies to either split up subjects in a RH/NRH group
or showdata that allowed for the formation of such a division.
For the latter procedure we applied the following rules: when
studies used a trichotomous model of hand-preference left-
and mixed-handed subjects were pooled into one NRH group.
This required the author to explicitly define hand-preference
boundaries. When an author did not explicitly define
boundaries and used the AHPQ, Annett's criteria for subdivi-
sion of hand-preference were applied (1=RH, 2–7=MIX,
8=LH). After pooling MIX and LH subjects, a combined test
score mean, weighted for group size was calculated.
2.3.3. A RH/MIX/LH trichotomy
For the analysis of the trichotomous model of hand-
preference we used LH, MIX and RH groups as defined by the
authors of a study. If a study further sub-classified hand-
preference groups subjects were again pooled now to yield
the three groups. If pooling was necessary, a combined test
score mean, weighted for group size was calculated. Apart
from the analysis of LH and RH vs. MIX, we also compared LH
and RH.
2.3.4. Schizotypy scales
Studies were selected that primarily or secondarily
investigated the association of hand-preference and popula-
tion schizotypy. Population schizotypy had to be measured by
validated questionnaires, such as the Perceptual Abberation
Scale (PAS) (Chapman et al.,1978), the Magical Ideation Scale
(MIS) (Eckblad and Chapman, 1983), the Schizotypy Ques-
tionnaire (STA) (Claridge and Broks,1984), the Rust Inventory
of Schizotypal Cognition (RISC) (Rust, 1988), the Peters et al.
Delusion Inventory (PDI) (Peters et al., 1999), the Schizotypal
PersonalityQuestionnaire(SPQ)(Raine,1991) andthe Oxford-
Liverpool Inventory of feelings and experiences (O-Life)
(Mason et al., 1995). For our meta-analyses we aimed to use
scores of positive schizotypy (e.g. magical ideation, paranoia,
perceptualabberations),sincemostinstrumentsarerestricted
2
M. Somers et al. / Schizophrenia Research xxx (2008) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Somers, M., et al., Hand-preference and population schizotypy: A meta-analysis, Schizophr. Res.
(2008), doi:10.1016/j.schres.2008.11.010

Page 3

to positive traits. In addition, the O-life and SPQ also measure
negative schizotypy items, such as social anxiety and flat
affect. The O-life consists of four scales measuring Unusual
experiences (UnEx) and Cognitive disorganization (CogDis) as
well as Introvertive Anhedonia (IntAn) and Impulsive Non-
conformity (ImpCon), but these questionnaires are not
designed to sum up scales to produce a single measure
(Mason and Claridge, 2006). The UnEx score seems the best
measure of positive schizotypy. The SPQ measures all 9 traits
of Schizotypal Personality Disorder as described in DSM-IIIR
(Spitzer et al., 1987) and can be dissected in a three factor
model consisting of cognitive–perceptual, interpersonal and
disorganized features (Raine et al., 1994), although some
studies use a two- (Kim et al., 1992) or four-factor model
(Stefanis et al., 2004). A total SPQ score can also be calculated.
When papers showed separate scores from a three- or four
factor model we used total SPQ scores, since summing up
separate positive factors to produce one positive schizotypy
SPQ score has not been validated.
2.4. Meta-analytic techniques
2.4.1. Schizotypy and right-handedness vs. non-right-handedness
For the analysis of schizotypy and right- vs. non-right-
handedness we used the program Comprehensive Meta-
analysis (www.meta-analysis.com). Primarily, the magni-
tude and direction of effect for schizotypal cognition and
hand-preference was calculated for each individual study.
When means and standard deviations were not available,
effect sizes were computed from exact p-values, t-values or
F-values (cf. (Lipsey and Wilson, 2001). Effect sizes for
schizotypy scores in right-handers vs. non-right-handers of
all studies were combined and weighted for sample size to
obtain a mean weighted effect size (“Hedges' g”, Hedges L.V.
and Olkin I., 1985) and p-values using the random effects
module of Comprehensive Meta-analysis software. Further,
the homogeneity statistic (I2) was calculated (Higgins et al.,
2003) to assess heterogeneity of results across studies. I2
describes the percentage of total variation across studies
due to heterogeneity rather than chance (Higgins et al.,
2003). Because the effect size can be overestimated as a
consequence of publication bias, the fail-safe number (N) of
studies was computed (Rosenthal, 1979; Orwin, 1983). This
fail-safe N is an estimation of the number of missing studies
that is needed to change the results of the meta-analysis to
non-significant.
2.4.2. Schizotypy and mixed-handedness vs. left- and right-
handedness
For the meta-analysis of schizotypyand MIX vs. RH and LH
the program Comprehensive Meta-analysis was again used.
The direction of effect for schizotypal cognition in MIX vs. RH
and MIX vs. LH was calculated as above, after which effect
sizes for schizotypy scores of all studies were combined and
weighted for sample size of the studies to obtain a mean
weighted effect size (“Hedges' g”)(Hedges L.V. and Olkin I.,
1985) and p-values using random effects. In addition we
performed a meta-analysis of strong LH vs. RH. As described
earlier the homogeneity statistic (I2) was calculated for both
comparisons to assess heterogeneity of results across studies
(Higgins et al., 2003). Further, the fail-safe N was calculated.
2.4.3. Sub-analyses of right vs. non-right-handedness and
mixed-handedness vs. left- and right-handedness
Since heterogeneity is a major issue in meta-analyses, we
aimed to perform additional meta-analyses of more homo-
geneoussub-groupsbyselectingstudiesofpositiveschizotypy
only.Inaddition,weaimedtoperformseparateanalysesbased
on ethnicity. However, ethnicity is difficult to assess since the
included studies comprised many different countries. In order
to create some form of uniformity, we subdivided studies
based on the language spoken. Further, we aimed to perform
separate meta-analyses of individual schizotypy scales.
3. Results
3.1. Inclusion
Eighty-six journal articles were identified and screened of
which twelve met all criteria for inclusion. Three studies
applied the MIS (Nalcaci et al., 2000; Barnett and Corballis,
2002). Of these, one study (showing a p-value and direction of
effect) could only be used in the analysis of extreme hand-
preference (Nicholls et al., 2005). One study used the PDI
(Preti et al., 2007) and two the STA (Claridge et al., 1998;
Gregory et al., 2003). One study investigated two samples of
subjects that were included for meta-analysis separately
(Annett and Moran, 2006). In the first sample the STA as well
as the UnEx scale from the O-life was used, but detailed data
was only shown for the STA. For the second sample the RISC
was used. One study used only the UnEx subscale from the O-
life (Shaw et al., 2001). Another study used the SPQ as well as
the PAS on a sample of adolescents and a sample of adults
(Chen and Su, 2006). Only SPQ results were used for meta-
analysis. The remaining studies used the SPQ only (Stefanis
et al., 2006; Dragovic et al., 2005). Of these, one study
investigated comparable subgroups, which allowed for aver-
aging of data (Dragovic et al., 2005). From one study that used
a two-factor SPQ model the F-value for cognitive–perceptual
dysfunction in the analysis of RH vs. MIX could be used (Kim
et al., 1992). Thus, the results from all but three studies that
showed SPQ total scores were scores of positive schizotypy.
Since two studies described two different samples (Annett
and Moran, 2006; Chen and Su, 2006), a total of fourteen
samples from twelve studies could be included for meta-
analysis with a total of 10058 subjects. Ten studies (twelve
samples) allowed a dichotomous division in RH and NRH,
yielding 6168 right-handed and 3174 non-right-handed
subjects. Using a trichotomous division yielded 5379 right-
handed, 2153 mixed-handed and 379 left-handed subjects.
Nine studies (ten samples) allowed for the comparison of RH
and MIX. Eight studies (nine samples) allowed for the
comparison of LH and MIX. Nine studies (ten samples) could
be used for the analysis of extreme hand-preference, with
5956 RH and 448 LH subjects. For a complete overview of all
(sub)analyses, see Table 1.
3.2. Schizotypy and hand-preference
3.2.1. Schizotypy and right-handedness vs. non-right-handedness
The overall meta-analysis investigating the association
between schizotypy and right-handedness vs. non-right-
handedness showed a significant effect for non-right-
3
M. Somers et al. / Schizophrenia Research xxx (2008) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Somers, M., et al., Hand-preference and population schizotypy: A meta-analysis, Schizophr. Res.
(2008), doi:10.1016/j.schres.2008.11.010

Page 4

handedness with Hedges' g=0.22 and p=0.01, implicating
higher total schizotypy scores in the non-right-handers as
compared to the right-handers. These findings are plotted in
Fig.1.HeterogeneitywashighwithI2=91%. Thefail-safeNwas
148. After removing an outlier of more than two standard
deviations (Barnett and Corballis, 2002), the effect of handed-
ness on schizotypy scores remained significant, though small
(Hedges' g=0.11, p=0.015). Heterogeneity decreased, with
I2=66%. The fail-safe N was 43. Removing the only study that
investigated adolescents slightly lowered the effect size as
well as heterogeneity (Hedges' g=0.09, pb0.05, I2=57%).
3.2.2. Schizotypy and mixed-handedness vs. strong left- and
right-handedness
The analysis investigating the association between schi-
zotypy and mixed handedness vs. extreme left- or right-
handedness yielded a significant effect for mixed-handedness
in comparison to right-handedness (Hedges' g=0.51,
p=0.0003, Fig. 2). Heterogeneity was high with I2=95%. The
fail-safe N was 223. The difference in schizotypy score
between left-handedness and mixed handedness was also
significant (Hedges' g=0.33, p=0.026, Fig. 3). Heterogeneity
was again high, with I2=76%. The fail-safe N was 31. Here too,
removal of an outlier (Barnett and Corballis, 2002) still
resulted in mixed-handers scoring significantly higher then
right-handers (Hedges' g=0.16, pb0.0001) with heterogene-
ity decreasing to I2=35%. The fail-safe N was 62. In the
comparison of mixed- and left-handers there was a compar-
able effect, with significance only at trend level (Hedges'
g=0.15, p=0.053) and a low I2of 28%. The fail-safe N was only
5. Meta-analysis of schizotypy in strong right- and left-
handed subjects from these studies showed no difference
(Hedges' g=0.06, p=0.51), with I2=61%.
3.2.3. Sub-analyses of schizotypy in right- vs. non-right-
handedness and mixed-handedness vs. strong left- and right-
handedness
3.2.3.1. Sub-analyses of positive schizotypy.
sub-analyses ofpositiveschizotypy byexcluding thosestudies
that showed SPQ total scores. In the comparison of right- vs.
non-right-handedness there was significantly higher schizo-
typy in non-right-handers (Hedges' g=0.31, p=0.015). Hetero-
geneity was high, with I2=93%. The fail-safe N was 93.
Removal of an outlier (Barnett and Corballis, 2002) still
resulted in non-right-handers scoring significantly higher
than right-handers (Hedges' g=0.12, p=0.005). Heterogeneity
was moderately high with I2=40%. The fail-safe N was 16. In
the comparison of mixed- vs. right-handedness there was
significantly higher schizotypy in mixed-handers (Hedges'
g=0.66, p=0.0002). Heterogeneity was high, with I2=96%.
The fail-safe N was 202. Removal of an outlier (Barnett and
We performed
Table 1
Overview meta-analyses of population schizotypy and hand-preference
Comparison Hedges' gp-value
N-samples
N-RH
N-LH
N-NRH
N-MIX
I2
Fail-safe N
RH vs. NRH
RH vs. MIX
LH vs. MIX
LH vs. RH
RH vs. NRH, without outliera
RH vs. MIX, without outliera
LH vs. MIX, without outliera
LH vs. RH, without outliera
RH vs. NRH, positive schizotypy onlyb
RH vs. MIX, positive schizotypy onlyb
LH vs. MIX, positive schizotypy onlyb
LH vs. RH, positive schizotypy onlyb
RH vs. NRH, without outlier, positive schizotypy onlyc
RH vs. MIX, without outlier, positive schizotypy onlyc
LH vs. MIX, without outlier, positive schizotypy onlyc
LH vs. RH, without outlier, positive schizotypy onlyc
RH vs. NRH, STA onlyd
RH vs. MIX, STA onlyd
LH vs. MIX, STA onlyd
LH vs. RH, STA onlyd
RH vs. NRH, English onlye
RH vs. MIX, English onlye
LH vs. MIX, English onlye
LH vs. RH, English onlye
RH vs. NRH, without outlier, English onlyf
RH vs. MIX, without outlier, English onlyf
LH vs. MIX, without outlier, English onlyf
LH vs. RH, without outlier, English onlyf
0.22
0.51
0.33
0.06
0.11
0.16
0.15
n.a.
0.31
0.66
0.45
0.06
0.12
0.18
0.20
n.a
0.16
0.17
0.08
0.13
0.40
0.70
0.38
0.08
0.12
0.15
0.15
n.a
0.01
0.00
0.03
0.51
0.01
0.00
0.053
n.a
0.02
0.0002
0.02
0.55
0.005
0.0001
0.04
n.a
0.003
0.001
0.50
0.29
0.009
0.0005
0.05
0.49
0.008
0.001
0.02
n.a
12
10
9
10
11
9
8
n.a
8
8
7
8
7
7
6
n.a
3
3
3
3
6
7
6
7
5
6
5
n.a
6168
5379
–
–
397
448
–
–
376
n.a
–
–
308
359
–
–
287
n.a
–
–
79
–
–
–
307
358
–
–
286
n.a
3174
–
–
–
3141
–
–
n.a
2304
–
–
–
2271
–
–
n.a
758
–
–
–
2127
–
–
–
2094
–
–
n.a
–
2153
2133
–
–
2141
2121
n.a
–
1958
1938
–
–
1946
1926
n.a
–
679
679
79
–
1838
1818
–
–
1826
1806
n.a
91
95
76
61
66
36
28
n.a
93
96
80
60
40
41
30
n.a
25
15
0
1
95
97
81
64
42
30
0
n.a
148
223
31
0
43
62
5
n.a
93
202
29
0
16
51
5
n.a
6
6
0
0
85
153
18
0
13
28
0
n.a
5956
5951
5162
–
n.a
4063
4050
–
4627
3846
3833
–
n.a
1065
1065
–
1065
3613
3642
4219
3396
3425
n.a
aStudy with outlier was excluded.
bStudies showing SPQ-total scales were excluded.
cStudy with outlier and studies showing SPQ total scales were excluded.
dOnly studies using the STA were included.
eOnly studies in English-speaking populations were included.
fStudy with outlier was excluded and only studies in English-speaking populations were included.
4
M. Somers et al. / Schizophrenia Research xxx (2008) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Somers, M., et al., Hand-preference and population schizotypy: A meta-analysis, Schizophr. Res.
(2008), doi:10.1016/j.schres.2008.11.010

Page 5

Corballis, 2002) still resulted in mixed-handers scoring
significantly higher than right-handers (Hedges' g=0.18,
p=0.0001). Heterogeneity decreased with I2=41%. The fail-
safe N was 51. In the comparison of mixed- vs left-handedness
there was significantly higher schizotypy in left-handers
(Hedges' g=0.45, p=0.022). Heterogeneity was high with
I2=80%. The fail-safe N was 29. Removal of an outlier (Barnett
and Corballis, 2002) resulted in mixed-handers still showing
significantly higher schizotypy (Hedges' g=0.20, p=0.045).
Heterogeneity decreased to I2=29%. There was a low fail-safe
N of 5. The comparison of strong left- and right-handers
showed no differences between the groups (Hedges' g=0.06,
p=0.55, I2=59%).
3.2.3.2. Sub-analyses of individual schizotypy scales.
analyses of individual schizotypy scales for all comparisons
was only possible for those studies using the STA. There was a
significantly higher effect for non-right-handers in compar-
ison to right-handers (Hedges' g=0.16, p=0.003, I2=25%, fail-
safe N=6) and for mixed-handers in comparison to right-
handers (Hedges' g=0.17, p=0.001, I2=15%, fail-safe N=6),
but not for mixed-handers in comparison to left-handers
Sub-
Fig. 1. Meta-analysis of schizotypy scores in right-handers and non-right-handers.
Fig. 2. Meta-analysis of schizotypy scores in right-handers and mixed-handers.
5
M. Somers et al. / Schizophrenia Research xxx (2008) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Somers, M., et al., Hand-preference and population schizotypy: A meta-analysis, Schizophr. Res.
(2008), doi:10.1016/j.schres.2008.11.010

Page 6

(Hedges' g=0.08, p=0.50, I2=0%) and strong left- vs. right-
handers (Hedges' g=0.13, p=0.29, I2=1%). For the other scales
no separate analysis was possible, because the number of
studies was too low.
3.2.3.3. Sub-analyses of language as a reflection of ethnicity.
Sub-analyses of language as a reflection of ethnicity wereonly
possible in those studies that were performed in English
speaking subjects. The remaining studies consisted of Italian
(Preti et al., 2007), Greek (Stefanis et al., 2006), Turkish
(Nalcaci et al., 2000), Japanese (Gregory et al., 2003) and
Taiwanese speaking subjects (Chen and Su, 2006) and as such
did not allow for a separate analysis. In English speaking
subjects, non-right-handers scored significantly higher than
right-handers (Hedges' g=0.40, p=0.009, I2=95%, fail-safe
N=85), also after removal of an outlier (Barnett and Corballis,
2002) (Hedges' g=0.12, p=0.008, I2=42%, fail-safe N=13). In
the analysis of mixed- vs. right-handers, mixed-handers score
significantly higher (Hedges' g=0.70, p=0.0004, I2=97%), also
after removal of an outlier (Barnett and Corballis, 2002)
(Hedges' g=0.15, p=0.001, I2=30%, fail-safe N=28). In the
analysis of mixed- vs. left-handers, mixed-handers scored
significantly higher (Hedges' g=0.38, p=0.05, I2=81%, fail-
safe N=18). After removal of an outlier (Barnett and Corballis,
2002), mixed-handers still scored significantly higher than
left-handers (Hedges' g=0.15, p=0.022, I2=0%), but the fail-
safe N decreased to 0. In the analysis of strong left- vs. right-
handers there was no difference in schizotypy (Hedges's
g=0.08, p=0.49, I2=64).
4. Discussion
This study aimed to examine the association between
schizotypy and handedness by performing a meta-analysis on
the literature from 1970 until 2007 concerning population
schizotypy and hand-preference. We could include twelve
studies,providingschizotypyscoresof10058subjectsintotal.
In the main analysis we found a subtle but significant effect
that remained significant after excluding a positive outlier
(Hedges' g=0.11), indicating higher schizotypy in non-right-
handers as compared to right-handers. In the subsequent
analysis we found that mixed-handers show a subtle increase
in schizotypy as compared to right-handers (Hedges' g=0.16),
a similar trend was observed for higher schizotypy in mixed-
handersascomparedtoleft-handers(Hedges'g=0.15).Wedid
not find any difference in schizotypy scores between strong
left- and strong right-handedness. This indicates that non-
right-handedness, and especially mixed-handedness is asso-
ciatedwithschizotypy, althoughtheeffect isonlysmall. These
findings parallel observations in schizophrenia patients
(Sommer et al., 2001a; Dragovic and Hammond, 2005) and
corroborate Crow et al.'s finding that ambidextrous children
more frequently develop schizophrenia in adult life than
right-handed peers (Crow et al., 1996). It is also in line with
findings in ambidextrous subjects showing relative dysfunc-
tion on various cognitive tasks in comparison to strong left-
and right-handers (Crow et al.,1998; Peters et al., 2006).
In our sub-analyses of positive schizotypy, we found
significantly higher schizotypy in non-right-handers in
comparison to right-handers and in mixed-handers in
comparison to right- as well as left-handers. The same effect
was seen in our analysis of studies investigating English
speaking subjects. There was no significant effect in any of the
analyses of strong right- vs. left-handedness. These findings
point towards the same direction as those in our main
analyses. In the sub-analyses of studies using the STA there
was an effect for non-right-handedness in comparison to
right-handedness and for mixed-handedness in comparison
to right-handedness, but not for mixed-handedness in
comparison to left-handedness. However, this analysis com-
prised only three studies and as such a low number of
subjects. Apart from the meta-analysis of the three studies
using the STA, heterogeneity was high in all significant
analyses and can as such not be explained by differences in
Fig. 3. Meta-analysis of schizotypy scores in left-handers and mixed-handers.
6
M. Somers et al. / Schizophrenia Research xxx (2008) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Somers, M., et al., Hand-preference and population schizotypy: A meta-analysis, Schizophr. Res.
(2008), doi:10.1016/j.schres.2008.11.010

Page 7

study population language or schizotypy scales. However,
removal of one study that was an outlier of more than 2
standard deviations in all comparisons (Barnett and Corballis,
2002) resulted in lowering heterogeneity to moderate or even
low in some analyses. This study is an outlier because the
mixed-handed group (n=12) that was investigated scored
extraordinarily high on the Magical Ideation Scale.
The fail-safe N was calculated forall analysesto investigate
publication bias. As can be seen in Table 1, publication bias is
unlikely in the main analyses and at least moderately unlikely
in most sub-analyses except those of left- vs. mixed handed-
ness without outlier, left- vs. mixed-handedness without
outlier in studies of positive schizotypy and the analyses of
studies using the STA.
In schizophrenia as well as in population schizotypy de-
creased cerebral lateralization is considered the underlying
basis for the increased frequency of non-right-handedness
(Crow,1997; Shaw et al., 2001). Crow proposed that decreased
cerebrallateralization,inparticularofphonologicalsequencesof
languagemaypredisposeforpsychoticsymptoms,asthiswould
cause a disturbance of the mechanism allowing a person to
distinguish his thoughts from the speech he produces and the
speech input that is received from others (Crow,1997).
Although no functional imaging has been performed on
language lateralization inpopulation schizotypy, several dicho-
tic listening studies found schizotypal subjects to exhibit a
relatively greater left ear preference for unilaterally presented
verbal stimuli than normal controls (Broks et al., 1984a,b;
Rawlings and Borge, 1987; Poreh et al., 1994), which suggests
increased language dominance of the right hemisphere.
Findings of a negative correlation between schizotypy scores
and performance asymmetry on a verbal divided visual field
task in male subjects point towards the same direction (Broks,
1984). Further, relatives of schizophrenia patients, who are
knowntohaveincreasedschizotypyscores(Appelsetal.,2004)
also show decreased language lateralization (Sommer et al.,
2004; Liet al.,2007b).As such, we hypothesizethatourfinding
of higher schizotypy in non-right-handers originates from a
higher prevalence of bilateral language representation. The
prevalence of bilateral language representation might even be
higher in mixed-handers than in strong left-handers. However,
this remains speculative, since the accurate prevalence of
bilateral language lateralization in mixed-handedness as
compared to strong left-handers is presently not known.
In contrast to bilateral language representation, strong
right-hemispheric lateralization has not been associated with
psychosis (e.g. (Sommer et al., 2001b). The prevalence of
right-hemisphere language dominance shows a linear
increase with the strength of left-handedness (Isaacs et al.,
2006; Knecht et al., 2000). This may be the reason why our
meta-analysis found no increase in schizotypy in strong left-
handers as compared to right-handers.
Exactly how decreased cerebral lateralization predisposes
to schizophrenia spectrum tendencies remains elusive.
Possibly, right hemisphere activity might underlie loosening
of associations. In tasks of semantic priming a more diffuse
spreading of activation has been shown in the right hemi-
sphere in contrast to the more focused spreading character-
istic for the left hemisphere (Chiarello et al., 1990). This
diffuse spreading could facilitate the occurrence of distant
associations and thus irrational thinking. Indeed, semantic
network activations have been shown to spread further
within a shorter period of time in thought-disordered
schizophrenia patients. Indirect semantic priming measured
by event-related potentials was increased in these patients
relative to non-thought-disordered patients and healthy
controls (Kreher et al., 2007). Further, fMRI studies have
shown activation of the right-hemisphere homologue of
Wernicke's area during speech production in (right-handed)
thought disordered schizophrenia patients (Kircher et al.,
2002), while controls performing the same task showed left-
lateralised activity. In addition to an association with thought
disorder, a recent fMRI study in 24 schizophrenia patients has
shown the right hemisphere homologue of Broca's area to be
activated during auditory verbal hallucinations (Sommer
et al., 2008). This implicates that decreased lateralization
might underlie different aspects of psychotic experiences.
In conclusion, non-right-handed subjects had higher
schizotypy levels than right-handed subjects, while there
was no difference between strong right- and strong left-
handed subjects. This supports the hypothesis of decreased
lateralization being a vulnerability model for schizophrenia-
spectrum features. Further neuroimaging studies on cerebral
lateralization of high schizotypy subjects are required to
elucidate the origins of the relation between non-right-
handedness and schizotypy.
Role of funding source
Funding has played no role in the realization of this report.
Contributors
M. Somers wrote all draft versions and prepared the manuscript for
publication. The initial proposal for the meta-analysis came from I. Sommer,
who also revised all draft versions. M. Boks advised on methodological
procedures. End revision was done by R. Kahn.
Conflicts of interest
There are no conflicts of interest.
References
Annett, M., 1970. A classification of hand preference by association analysis.
Br. J. Psychol. 61, 303–321.
Annett, M., 2004. Hand preference observed in large healthy samples:
classification, norms and interpretations of increased non-right-handed-
ness by the right shift theory. Br. J. Psychol. 95, 339–353.
Annett, M., Moran, P., 2006. Schizotypy is increased in mixed-handers,
especially right-handed writers who use the left hand for primary
actions. Schizophr. Res. 81, 239–246.
Appels, M.C., Sitskoorn, M.M., Vollema, M.G., Kahn, R.S., 2004. Elevated levels
of schizotypal features in parents of patients with a family history of
schizophrenia spectrum disorders. Schizophr. Bull. 30, 781–790.
Barnett, K.J., Corballis, M.C., 2002. Ambidexterity and magical ideation.
Laterality 7, 75–84.
Broks, P., 1984. Schizotypy and hemisphere function-II. Performance asym-
metry on a Verbal Divided Visual-Field task. Pers. Individ. Differ. 5 (6),
649–656 Ref Type: Journal (Full).
Broks, P., Claridge, G., Matheson, J., Hargreaves, J., 1984a. Schizotypy and
hemisphere function-IV. Story comprehension under binaural and
monaural listening conditions. Person. Indiv. Diff. 5 (6), 665–670.
Broks, P., Claridge, G., Matheson, J., Hargreaves, J., 1984b. Schizotypy and
hemisphere function-IV. Story comprehension under binaural and
monaural listening conditions. Person. Indiv. Diff. 5 (6), 665–670.
Bryden, P.J., Pryde, K.M., Roy, E.A., 2000. A developmental analysis of the rela-
tionshipbetweenhandpreferenceandperformance:II.Aperformance-based
method of measuring hand preference in children. Brain Cogn. 43, 60–64.
Chapman, L.J., Chapman, J.P., Raulin, M.L., 1978. Body-image aberration in
schizophrenia. J. Abnorm. Psychology 87, 399–407.
7
M. Somers et al. / Schizophrenia Research xxx (2008) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Somers, M., et al., Hand-preference and population schizotypy: A meta-analysis, Schizophr. Res.
(2008), doi:10.1016/j.schres.2008.11.010

Page 8

Chen, W.J., Su, C.H., 2006. Handedness and schizotypy in non-clinical
populations: influence of handedness measures and age on the relation-
ship. Laterality 11, 331–349.
Chiarello, C., Burgess, C., Richards, L., Pollock, A., 1990. Semantic and
associative priming in the cerebral hemispheres: some words do, some
words don't … sometimes, some places. Brain Lang. 38, 75–104.
Claridge, G., Broks, P., 1984. Schizotypy and hemisphere function: I. Theoretical
considerationsandthemeasurementofschizotypy.Pers.Individ.Differ.5(6),
633–648 Great Britain, Pergamon Press. Ref Type: Generic.
Claridge, G., Clark, K., Davis, C., Mason, O., 1998. Schizophrenia risk and
handedness: a mixed picture. Laterality 3, 209–220.
Crow, T.J., 1997. Is schizophrenia the price that Homo sapiens pays for
language? Schizophr. Res. 28, 127–141.
Crow, T.J., Done, D.J., Sacker, A., 1996. Cerebral lateralization is delayed in
children who later develop schizophrenia. Schizophr. Res. 22, 181–185.
Crow, T.J., Crow, L.R., Done, D.J., Leask, S., 1998. Relative hand skill predicts
academic ability: global deficits at the point of hemispheric indecision.
Neuropsychologia 36, 1275–1282.
Cyhlarova, E., Claridge, G., 2005. Development of a version of the Schizotypy
Traits Questionnaire (STA) for screening children. Schizophr. Res. 80,
253–261.
Dragovic, M., Hammond, G., 2005. Handedness in schizophrenia: a
quantitative review of evidence. Acta Psychiatr. Scand. 111, 410–419.
Dragovic, M., Hammond, G., Jablensky, A., 2005. Schizotypy and mixed-
handedness revisited. Psychiatry Res. 136, 143–152.
Eckblad, M., Chapman, L.J., 1983. Magical ideation as an indicator of
schizotypy. J. Consult. Clin. Psychol. 51, 215–225.
Gregory, A.M., Claridge, G., Clark, K., Taylor, P.D., 2003. Handedness and
schizotypy in a Japanese sample: an association masked by cultural
effects on hand usage. Schizophr. Res. 65, 139–145.
Hedges, L.V., Olkin, I., 1985. Statistical Methods for Meta-analysis. Orlando,
Florida. Ref Type: Generic.
Higgins, J.P., Thompson, S.G., Deeks, J.J., Altman, D.G., 2003. Measuring
inconsistency in meta-analyses. BMJ 327, 557–560.
Isaacs, K.L., Barr, W.B., Nelson, P.K., Devinsky, O., 2006. Degree of handedness
and cerebral dominance. Neurology 66, 1855–1858.
Johnstone, E.C., Ebmeier, K.P., Miller, P., Owens, D.G., Lawrie, S.M., 2005.
Predicting schizophrenia: findings from the Edinburgh High-risk Study.
Br. J. Psychiatry 186, 18–25.
Kim, D., Raine, A., Triphon, N., Green, M.F., 1992. Mixed handedness and
features of schizotypal personality in a nonclinical sample. J. Nerv. Ment.
Dis. 180, 133–135.
Kircher, T.T., Liddle, P.F., Brammer, M.J., Williams, S.C., Murray, R.M., McGuire,
P.K., 2002. Reversed lateralization of temporal activation during speech
production in thought disordered patients with schizophrenia. Psychol.
Med. 32, 439–449.
Knecht,S.,Drager,B.,Deppe,M.,Bobe,L.,Lohmann,H.,Floel,A.,Ringelstein,E.B.,
Henningsen, H., 2000. Handedness and hemispheric language dominance
in healthy humans. Brain 123 (Pt 12), 2512–2518.
Kreher, D.A., Holcomb, P.J., Goff, D., Kuperberg, G.R., 2007. Neural evidence for
faster and further automatic spreading activation in schizophrenic
thought disorder. Schizophr. Bull.
Leonhard, D., Brugger, P.,1998. Creative, paranormal, and delusional thought:
a consequence of right hemisphere semantic activation? Neuropsychia-
try Neuropsychol. Behav. Neurol. 11, 177–183.
Li, X., Branch, C.A., Ardekani, B.A., Bertisch, H., Hicks, C., DeLisi, L.E., 2007a. fMRI
study of language activation in schizophrenia, schizoaffective disorder and
in individuals genetically at high risk. Schizophr. Res. 96,14–24.
Li, X., Branch, C.A., Bertisch, H.C., Brown, K., Szulc, K.U., Ardekani, B.A., DeLisi,
L.E., 2007b. An fMRI study of language processing in people at high
genetic risk for schizophrenia. Schizophr. Res. 91, 62–72.
Lipsey, M.W., Wilson, D.B., 2001. The way in which intervention studies have
“personality”andwhyitisimportanttometa-analysis.Eval.HealthProf.24,
236–254.
Mason, O., Claridge, G., 2006. The Oxford-Liverpool Inventory of Feelings and
Experiences (O-LIFE): further description and extended norms. Schi-
zophr. Res. 82, 203–211.
Mason, O., Claridge, G., Jackson, M., 1995. New scales for the assessment of
schizotypy. Pers. Individ. Differ.18 (1), 7–13 Great Britain, Pergamon. Ref
Type: Generic.
McManus, I.C., 1985. Handedness, language dominance and aphasia: a
genetic model. Psychol. Med. Monogr. Suppl. 8, 1–40.
Nalcaci, E., Kalaycioglu, C., Cicek, M., Budanur, O.E., 2000. Magical ideation
and right-sided hemispatial inattention on a spatial working memory
task: influences of sex and handedness. Percept. Mot. Skills 91, 883–892.
Nicholls, M.E., Orr, C.A., Lindell, A.K., 2005. Magical ideation and its relation to
lateral preference. Laterality 10, 503–515.
Oldfield, R.C., 1971. The assessment and analysis of handedness: the
Edinburgh inventory. Neuropsychologia 9, 97–113.
Orwin,R.G.,1983.A fail-safeN foreffectsizeinmeta-analysis.J.Educ.Stat. 8 (2),
157–159 Ref Type: Journal (Full).
Peters, M., 1998. Description and validation of a flexible and broadly usable
handedness questionnaire. Laterality 3, 77–96.
Peters, E.R., Joseph, S.A., Garety, P.A., 1999. Measurement of delusional
ideation in the normal population: introducing the PDI (Peters et al.
Delusions Inventory). Schizophr. Bull. 25, 553–576.
Peters, M., Reimers, S., Manning, J.T., 2006. Hand preference for writing and
associations with selected demographic and behavioral variables in
255,100 subjects: the BBC internet study. Brain Cogn. 62, 177–189.
Poreh, A.M., Whitman, D.R., Ross, T.P., 1994. Creative thinking abilities and
hemisphericasymmetryinschizotypalcollegestudents.Curr.Psychol.12,
344–352.
Preti, A., Sardu, C., Piga, A., 2007. Mixed-handedness is associated with the
reporting of psychotic-like beliefs in a non-clinical Italian sample.
Schizophr. Res. 92, 15–23.
Pujol, J., Deus, J., Losilla, J.M., Capdevila, A., 1999. Cerebral lateralization of
language in normal left-handed people studied by functional MRI.
Neurology 52, 1038–1043.
Raine, A.,1991. The SPQ: a scale for the assessment of schizotypal personality
based on DSM-III-R criteria. Schizophr. Bull. 17, 555–564.
Raine, A., Reynolds, C., Lencz, T., Scerbo, A., Triphon, N., Kim, D., 1994.
Cognitive–perceptual, interpersonal, and disorganized features of schi-
zotypal personality. Schizophr. Bull. 20, 191–201.
Rawlings, D., Borge, A., 1987. Personality and hemisphere function: two
experiments using the dichotic shadowing technique. Pers. Individ.
Differ. 8 (4), 483–488 Ref Type: Journal (Full).
Rosenthal, R., 1979. The file drawer problem and tolerance for null results.
Psychol. Bull. 86 (3), 638–641 Ref Type: Journal (Full).
Rust, J.,1988. The Rust Inventory of Schizotypal Cognitions (RISC). Schizophr.
Bull. 14, 317–322.
Shaw, J., Claridge, G., Clark, K., 2001. Schizotypy and the shift from dextrality:
a study of handedness in a large non-clinical sample. Schizophr. Res. 50,
181–189.
Sommer, I., Ramsey, N., Kahn, R., Aleman, A., Bouma, A., 2001a. Handedness,
language lateralisation and anatomical asymmetry in schizophrenia:
meta-analysis. Br. J. Psychiatry 178, 344–351.
Sommer, I.E., Ramsey, N.F., Kahn, R.S., 2001b. Language lateralization in
schizophrenia, an fMRI study. Schizophr. Res. 52, 57–67.
Sommer, I.E., Ramsey, N.F., Mandl, R.C., Kahn, R.S., 2003. Language
lateralization in female patients with schizophrenia: an fMRI study.
Schizophr. Res. 60, 183–190.
Sommer, I.E., Ramsey, N.F., Mandl, R.C., van Oel, C.J., Kahn, R.S., 2004.
Language activation in monozygotic twins discordant for schizophrenia.
Br. J. Psychiatry 184, 128–135.
Sommer, I.E., Vd Veer, A.J., Wijkstra, J., Boks, M.P., Kahn, R.S., 2007. Comparing
language lateralization in psychotic mania and psychotic depression to
schizophrenia; a functional MRI study. Schizophr. Res. 89, 364–365.
Sommer, I.E., Diederen, K.M., Blom, J.D., Willems, A., Kushan, L., Slotema, K.,
Boks, M.P., Daalman, K., Hoek, H.W., Neggers, S.F., Kahn, R.S., 2008.
Auditory verbal hallucinations predominantly activate the right inferior
frontal area. Brain.
Spaniel, F., Tintera, J., Hajek, T., Horacek, J., Dezortova, M., Hajek, M., Dockery,
C., Kozeny, J., Hoschl, C., 2007. Language lateralization in monozygotic
twins discordant and concordant for schizophrenia. A functional MRI
pilot study. Eur. Psychiatr. 22, 319–322.
Spitzer, R.L., Skodol, A.E., Gibbon, M., Williams, J.B.W., 1987. DSM-III-R:
Diagnostic and Statistical Manual of Mental Disorders, 3rd edn. American
Psychiatric Association, Washington, D.C.
Stefanis,N.C.,Smyrnis,N.,Avramopoulos,D.,Evdokimidis,I.,Ntzoufras,I.,Stefanis,
C.N.,2004.Factorialcompositionofself-ratedschizotypaltraitsamongyoung
males undergoing military training. Schizophr. Bull. 30, 335–350.
Stefanis, N.C., Vitoratou, S., Smyrnis, N., Constantinidis, T., Evdokimidis, I.,
Hatzimanolis, I., Ntzoufras, I., Stefanis, C.N., 2006. Mixed handedness is
associated with the disorganization dimension of schizotypy in a young
male population. Schizophr. Res. 87, 289–296.
Straaten, V.A., 1955. Considerations on the occurrence of manifest left-
handedness in schizophrenia. Ned. Tijdschr. Geneeskd. 99, 706–710.
Szaflarski, J.P., Holland, S.K., Schmithorst, V.J., Byars, A.W., 2006. fMRI study of
language lateralization in children and adults. Hum. Brain Mapp. 27,
202–212.
8
M. Somers et al. / Schizophrenia Research xxx (2008) xxx–xxx
ARTICLE IN PRESS
Please cite this article as: Somers, M., et al., Hand-preference and population schizotypy: A meta-analysis, Schizophr. Res.
(2008), doi:10.1016/j.schres.2008.11.010