Acta Psychiatr Scand 2002: 105: 317±319 Copyright ã Munksgaard 2002

Printed in UK. All rights reserved ACTA PSYCHIATRICA

SCANDINAVICA
ISSN 0001-690X

Short communication
Age-at-onset and schizophrenia:
reversed gender e€ect
B. N. Gangadhar, C. Panner Selvan, D. K. Subbakrishna, B. N. Gangadhar1, C. Panner
N. Janakiramaiah. Age-at-onset and schizophrenia: Selvan1, D. K. Subbakrishna2,
reversed gender e€ect N. Janakiramaiah1
Acta Psychiatr Scand 2002: 105: 317±319. ã Munksgaard 2002. 1
Department of Psychiatry and 2Department of
Biostatistics, NIMHANS, Bangalore 560 029, India
Objective: This study seeks an explanation for reversed gender e€ect
on age-at-onset (AAO) in schizophrenia. The hypothesis is older AAO
in males would be detected in a sample where higher infant mortality
(IMR) prevailed.
Method: Case records of International Classi®cation of Diseases-10
(ICD-10) schizophrenia patients from two states (n ˆ 70 each) with an
IMR of 13 and 67 per thousand were reviewed and AAO was obtained
by using the recorded age and duration of illness.
Results: In the sample from the state with lower IMR, AAO did not
di€er between the two sexes. However, men had older AAO than
women in the state with ®vefold higher IMR.
Conclusion: Gender di€erences in AAO may be a function of perinatal B. N. Gangadhar, Additional Professor, Department of
complications. In places where infants with perinatal complications are Psychiatry, NIMHANS, Bangalore 560 029, India.
less likely to survive, hence high IMR, a small group of potentially E-mail: bng@nimhans.kar.nic.in
youngest AAO schizophrenic males may be eliminated thus changing
the gender e€ect on AAO. Accepted for publication 12 November, 2001

subpopulation at risk for very young AAO schi-

Introduction
The tenth edition of International Classi®cation of
Diseases (1) notes that schizophrenia has an earlier
onset in males. Several studies, mainly from the
West, support this (2). Nonetheless, some studies
failed to replicate earlier age-at-onset (AAO) in
males. Indeed, one Indian group has reported
earlier onset in females (3), a ®nding con®rmed in
another sample recently (4). Some of the explana-
tions for earlier onset of schizophrenia in males
are, lack of protective e€ects of estrogens (2),
perinatal brain trauma (5, 6) and familial loading
(2, 7). Familial loading and perinatal insult can
antagonize the protection o€ered by estrogen (2).
Perinatal complication and hence, mortality occurs
more often in male than female infants (8). Preti
et al. (9) found nearly threefold higher frequency
of severe obstetric/perinatal complications in males
than female schizophrenics. It is, therefore, a likely
explanation for earlier AAO in males. Murthy
et al. (3) argued that in some countries with less
than adequate perinatal care facilities (perhaps so
about two decades ago), infants with such trauma
may not survive, thus reducing the size of this
zophrenia. We hence hypothesized that older AAO
in males would be associated with higher infant
mortality rates (IMR) (assuming that infants with
perinatal trauma would succumb).
Material and methods
The study was record-based. NIMHANS although
located in Karnataka, being a premier national
institute, attracts some patients from neighboring
states. Kerala is one of such states and has lowest
IMR in the country ± 13 of 1000, as against
Karnataka that has IMR of 67 of 1000 (8). We
looked for the case records under the diagnosis of
schizophrenia (1). Seventy patients from Kerala
had registered during one calendar year. For each
of these patients the next consecutive patient
from Karnataka formed the control (n ˆ 70).
Two psychiatrists (BNG, CPS) jointly reviewed
the ®les, veri®ed the illness duration and then
arrived at AAO thereof. Yet another measure
retrieved was age at ®rst consultation. Independent
sample t-test was used for comparing AAO
between the sexes.

317
Gangadhar et al.
Table 1. Patient characteristics and AAO of the two samples
Samples from
Variables Kerala (n ˆ 70)
Age (years) 29.8  9.0
Male : Female1 37 : 33
AAO (years)2 23.5  8.3
Age (years) at first Consultation3 24.8  7.8
AAO (years)**
Male 23.5  8.5
Female 23.5  8.2
Age (years) at first Consultation***
Male 25.1  7.9
Female 24.4  7.8
1
Figures in cells are mean  SD in all except one cell that has number of subjects.
Kerala±Karnataka difference: 2P ˆ 0.181; 3P ˆ 0.028.
Sex difference: Kerala; **P ˆ 0.99, ***P ˆ 0.68; Karnataka; **P ˆ 0.045,
***P ˆ 0.091.
Results
Samples from Kerala and Karnataka were not
di€erent with respect to age and sex distribution.
The mean AAO and age at ®rst consultation of
Kerala sample were lesser than Karnataka; the
di€erence being signi®cant only for the latter
measure. Sex di€erence was absent in the Kerala
sample, whereas in the Karnataka sample the mean
AAO and age at ®rst consultation were older in
males (Table 1).
Discussion
The mean AAO in this (record-based) sample was
similar to an earlier prospective sample from the
same center (3) or from another Indian sample
(10). This validates partially the AAO information
obtained from the records. Sex and age distribu-
tion in the samples of the two states too was
similar. A ®vefold di€erence in the IMR in the
two states brought out a di€erential sex-e€ect on
measures of AAO. Sex di€erence was absent in
the lower-IMR Kerala state sample. In the sample
from the state with higher IMR, Karnataka,
males had older AAO than females. The age at
®rst consultation too was older (trend) in this
sample.
With higher IMR in Karnataka, more infants
with perinatal complications (more in males) may
have succumbed. This may have reduced the size of
the population at risk contributing to lower AAO.
As perinatal complications being thrice more
common in males (9), this di€erential IMR would
confound the sex-e€ect. Higher IMR may explain
why older AAO was detected in males. Younger
AAO in males as in literature (2) was not detected
318
in the lower IMR state's sample and needs
explanation. The IMR in Germany is only one-
third (3.5/1000) (11), as that of Kerala. This may
Karnataka (n ˆ 70) explain why the younger AAO for males in
29.6  8.9
Germany (2) is not seen in the Kerala sample of
38 : 32 this study. The limitations of record-based data
25.4  8.5 have to be considered in accepting these results.
27.9  8.8 For example, it is dicult to relate AAO from the
27.4  8.6 case ®les to onset of psychosis reliably. The role of
23.2  8.0 familial loading was not examined in this record-
based study. This is yet another limitation consid-
29.5  8.5
ering that females have higher incidence of familial
25.9  9.0
morbidity (12).
Kerala being the most literate state in the
country, the patients may have sought medical
help earlier. While the AAO was only 2 years
younger in Kerala sample, the mean age at ®rst
consultation was 3 years earlier than in Karnataka
sample (Table 1). However, the sex di€erence was
absent in the Kerala sample. The social status of
women in the two states is not di€erent. Social
factors may have a less likely role a€ecting AAO.
In summary, perinatal complications may in¯uence
the sex di€erence in AAO. Future studies on sex
di€erences in AAO of schizophrenia may have to
control for perinatal complications and perhaps
familial loading.
References
1. WORLD HEALTH ORGANIZATION. International classi®ca-
tion of diseases, 10th edn. Geneva: WHO, 1992.
2. KONNECKE R, HAFNER H, MAURER K, LOFFLER W,
HEIDEN W. Main risk factors for schizophrenia. increase
familial loading and pre- and peri-natal complications
antagonize the protective e€ect of estrogen in women.
Schizophr Res 2000;44:81±93.
3. MURTHY GVS, JANAKIRAMAIAH N, GANGADHAR BN,
SUBBAKRISHNA DK. Sex di€erence in age at onset: a dis-
crepant ®nding from India. Acta Psychiatr Scand
1998;97:321±325.
4. SUBBAKRISHNA DK, MURALI N, GANGADHAR BNN,
JANAKIRAMAIAH N. Younger age at onset of schizophrenia
in females: a replicative study. In: SUBBAKRISHNA DK,
KALIAPERUMAL VG, eds. Statistical methods and appli-
cation in biology and medicine. Bangalore, India:
NIMHANS, 2001:253±260.
5. BYRNE M, BROWNE R, MULRYAN N et al. Labor and
delivery complications and schizophrenia, case control
study using contemporaneous labor ward records. Br J
Psychiatry, 2000;176:531±536.
6. KIROV G, JONES PB, HARVEY I et al. Do obstetric com-
plications cause the earlier age at onset in male than female
schizophrenics? Schizophr Res 2000;20:117±124.
7. ALDA M, AHREUS B, LIT W et al. Age of onset in familial
and sporadic schizophrenia. Acta Psychiatr Scand
1996;93:447±450.
8. PARK K. Park's textbook of preventive and social medi-
cine, 15th edn, Jabalpur, India: Banaresidas Bharat Pub-
lishers, 1997:342±391.

9. PRETI A, CARDASCIA L, ZEN T, MARCHETTI M, FAVARETTO
G, MIOTTO P. Risk for obstetric complications and schi-
zophrenia. Psychiatry Res 2000;96:127±139.
10. HAMBRECHT M, MAURER K, HAFNER H, SARTORIUS N.
Transnational stability of gender di€erences in schizophre-
nia? Eur Arch Psychiatry Clin Neurosci 1992;242:6±12.
Age at onset of schizophrenia
11. NOLTE E, BRAND A, KOUPILOVA I, MCKEE M. Neonatal
and postneonatal mortality in Germany since uni®cation.
J Epidemiol Community Health 2000;54:84±90.
12. WOLYNIEK PS, PULVER AE, MCGRATH JA, TAM D,
SCHIZOPHRENIA. Gender and familial risk, J Psych
Research 1992;26:17±27.
319