Offspring Sex and Degree of Active Maternal Mental Disturbance near Reproduction among Female Mental Patients Thomas
F. McNeil,
Lennart
Kaij, and lnger
Persson-Blennow
A
NUMBER of studies have investigated the relationship between the sex of offspring born to female schizophrenics and the timing of the development of schizophrenic symptomatology in these mothers near reproduction. The basic goal of the studies was to link one sex among the offspring with the occurrence of the mother’s mental illness, thus providing a clue to the nature of the illness. The research on the topic might be characterized as primarily oriented toward (a) the implied teratogenic effect of the maternal illness on fetuses of one sex or (b) the implied effect of fetal sex on the mother’s mental status during pregnancy and postpartum. MATERNAL
TERATOGENIC
EFFECT
ON
FETUSES
OF ONE
SEX
Shearer et al.,’ Taylor,2 and Taylor and Levine3 found evidence that schizophrenics who developed psychosis within 1 month of conception gave live birth only to female offspring; this finding was interpreted as reflecting a schizophrenia blood factor specifically toxic to male fetuses early in gestation. In contrast, Schorer4 found no difference in frequency of male vs. female offspring resulting from pregnancies during which the mother suffered an onset of schizophrenia during the first trimester (or any other trimester). Using a broader time reference, Lane5 found an unexpectedly high rate of female offspring born to schizophrenic women* (in general). The predominance of female offspring was found especially for reproductions before (rather than after) the mother’s first psychiatric hospitalization, for black (rather than white) schizophrenics, and for reactive (but not for process) schizophrenics. Lane interpreted the predominance of female offspring as due to stress in the parents rather than to a biochemical property of schizophrenia per se. Contrasting data from Mednick et al.6 and from Erlenmeyer-Kimling’ showed no predominance of female offspring born to schizophrenic men or women. Rates of obstetric complications (OCs) associated with male vs. female offspring are relevant to the topic of teratogenic effect, assuming a continuum from complete destruction of the fetus to OCs with surviving fetuses. Mednick et *Lane also found a trend toward more female than male offspring for schizophrenic men.” From lhe Deparrment of Psyhiarry. Universit.v of Lund. MaIma, Sweden, and rhe Lafavetle Clinic, Detroit. Mich. Thomas F. McNeil, Ph.D.: Associare Projtissor. l;niversit,v of Lund, and Researcher. Lafa_vette Clinic. Lennart Kaij, M.D.: Associare Professor. University o/Lund. lnger Persson-Blennow, M.L.: .4manuens. Universirp of Lund. This work was partially supported by NIMH research grant 188.57, bv the Grant Foundation. Inc.. and by granr B73-21X-3793-02Bfrom ihe Swedish Medical Research Council. Reprinr requesrs may be addressed io Lennari Kav. M.D.. Deparfment of Psychiatry+. C’niversit! of Lund, 214 01 Malmd. Sweden. E, 1975 bv Grune & Stratton. Inc. Comprehensive Psychiatry, Vol. 16. No. 1 (January/Februaryl,
1975
69
70
MCNEIL. KAIJ. AND
PERSSON-BLENNOW
a1.6 found that female offspring of schizophrenics were associated with considerably more pregnancy complications than were male offspring, a finding that is contrary to the Taylor hypothesis and to the Lane hypothesis. In contrast, McNeil and Kaij* found no significant differences in pregnancy complication or other complication rates associated with male vs. female offspring born to female psychiatric patients (including schizophrenics). However, the male offspring of the patients (both endogenous psychotics and other patients) were significantly more often preterm than were the female offspring of the patients; such a malefemale-offspring preterm difference was only slightly evident in the control group. In general the above studies have not led to any consensus concerning whether female schizophrenics, especially those becoming schizophrenic early during pregnancy, have a differential teratogenic effect on the male conceptus. IMPLIED
EFFECT OF FETAL SEX ON MOTHER’S
MENTAL
STATUS
NEAR
REPRODUCTION
Evidence from Taylor and Levine3 and Taylor2,g suggested a causal relationship between mother’s onset of schizophrenic symptomatology postpartum and the sex of the offspring. Women developing schizophrenia within the 1st month postpartum bore a preponderance of male offspring. Taylor thus thought that the male fetus (when conceived and retained in the uterus) is hormonally prophylactic for development of the mother’s schizophrenic symptomatology during pregnancy; withdrawal of the male hormonal influence through delivery of the child thus allowed or precipitated the postpartum breakdown in the mother. Contrasting data from Melges’O and Schorer* showed no predominance of male offspring among women suffering onset of schizophrenic symptomatology in the 1st month postpartum. Taylor9 attempted to explain the discrepancy between his own results and those of Melges’O as due to sample differences, the Taylor sample being constituted of process schizophrenics and the Melges sample representing women with acute short-lived psychotic episodes. Melges,‘O however, claimed to have made his sample comparable to Taylor’s. In any case, Taylor’s original position that a relationship between postpartum breakdown and male offspring would be found only for process schizophrenics was not supported by other independent studies: Lane5 found the excess of female offspring only for reactive and not for process schizophrenics, and Schorer* found no significant* difference in offspring sex ratio between process vs. reactive types. The lack of agreement in the literature cited above suggested further empirical study of the topic. Our previous study” of severity of maternal mental disturbance near reproduction in female mental patients allowed an evaluation of the relationship between fetal sex and severity of the mother’s mental disturbance both during pregnancy and postpartum. The current study differs from the Taylor and Shearer studies in that (a) broader time periods near reproduction (10 months predelivery, first 10 months postpartum) are here studied; (b) a number of different diagnostic groups are here represented; (c) women with all levels of *Schorer” did find a nonsignificant among the reactive schizophrenics.
trend
toward
more
female
offspring
among
the process
than
OFFSPRING
SEX
AND
MATERNAL
71
DISTURBANCE
active disturbance near reproduction are here represented, that is, the women in the current study were at some time psychiatric patients, but not necessarily in association with every reproduction studied; and (d) stricter criteria are here used in diagnosing schizophrenia in the European tradition (see McNeil et al.“.‘“). The following empirical questions were studied: Do the female patients have unusual oflspring sex ratios? Lane’s data would suggest an overweight of female offspring among patients with psychoses corresponding to American “reactive schizophrenia.” Is there a predominance of one sex among offspring born before vs. after mother's first psychiatric hospitalization? Lane’s data would suggest a predominance of female offspring born to schizophrenics before the mother’s first hospitalization. fs there a predominance of male or female oj.spring rorresponding with different levels of active maternal mental disturbance during pregnancy and/or postpartum? Lane’s theory of stress would suggest that regardless of diagnostic
group mothers actively disturbed during pregnancy should have a preponderance of female offspring, while those not disturbed should show a typical proportion between male and female offspring. Is there an imbalance in oflspring sex ratio among mothers whose mental disturbance developed or became more severe postpartum, as contrasted with mothers whose mental status remained the same or improved postpartum? Taylor’s hypothesis would suggest that those mothers becoming more severely disturbed postpartum should have considerably more male than female offspring; data by Shorer and Melges would suggest no difference between these maternal groups. METHODOLOGY
AND SAMPLES
Subjects were selected as (a) all females born after 1912, (h) admitted to &tra Psychiatric Hospital, Malmo, Sweden, during 1958 - 1960, and (c)delivering a live child by fall of 1970 according to the records of the Swedish population register system. Among the 214 women fitting criteria (a) and(h), 213 (99.5%) were successfully traced through the population register, showing 123 (57.7%) of the women to have produced a total of 246 live-born children (including two sets of twins). For the current analyses, one twin-pair discordant for sex was excluded from the sample, and one pair concordant for sex was counted as one reproduction. The resulting sample consisted of 243 reproductions for 123 mothers. Year ofdelivery ranged from 1934 to 1970, with a median delivery year of 1951. The women were diagnostically categorized from psychiatric records by the second author (L.K.) as either endogenous psychotics (process schizophrenia, manic-depressive psychosis, schizophreniclike psychoses [SLP]*) or Other Patients (psychopaths and diverse diagnoses), as described in McNeil and Kaij.” Maternal and offspring sample sizes for each of the diagnostic categories are shown in Table 2. Severity of active maternal mental disturbance was evaluated by the third author (I.P.-B.) for the 10 months prior to delivery and also for the first 10 months postpartum. Severity was defined as the most disturbed status during each of these IO-month periods. Severity ratings were based on information in all known psychiatric records for the patients: additional information regarding official sick leaves based on psychiatric disturbance was obtained through the country-wide health insurance system for I9 patients reproducing during the 1960s and not known to be subsequently psychiatrically rehospitalized. Severity was rated on a seven-point general psychopathology scale (Table I ), which was shown to have high interrater (r = + 0.98) and rate rerate (r = + 0.96) reliabilities,
*Schizophrenic-like psychoses, as defined in McNeil and Kaij.‘Y apparently represent of those patients whose disturbance would be termed reactive schizophrenia in America.
at least some
MCNEIL. KAIJ, AND
72
Table 1. Not disturbed. 2.
Borderline
1. Seven-Point
(“Healthy.”
disturbed.
(Practically
stressed shows occasional 3. Mildly
disturbed.
ache, fatigue,
General Psychopathology
Has no known symptoms. “healthy.”
Functions
isolated signs of disturbance,
(Mild symptoms
mild depression,
also appear
mild phobia,
Moderately
disturbed.
(In
viewed as troublesome
non-stressing
by the patient
sidered to be deviant
Scale well in all respects.)
well and without
in non-stressing
but when
situations,
e.g., tenseness, head-
of psychosis or in con-
and can be on medication.)
situations
symptoms
or by those around
or sick by those around
symptoms,
e.g., tenseness, nervousness.)
slight defects as aftermath
nection with aging, etc. Can have seen a physician 4.
Functions
PERSSON-BLENNOW
him/her.
toms are of the same type as in 3., but are found
appear
him/her.
which
The patient
Can be working
can sometimes is sometimes
be con-
or on sick leave. Symp-
to a greater degree, and several may exist at
the same time.) 5. Markedly
ill. (Clear mental
the patient’s The
life.
patient
Examples:
works
symptoms; functions
either
and blunted
emotional conditions.
ill. (Grave
the present.
with
or not at all. Can be hospitalized delusions
associated
Patients can take care of themselves. disturbances
which
Medication
and
hospitalization
pressions, psychoses in full bloom,
with
has stopped,
defective
often
and blunted
conditions
basis.
symptoms; thought
schizophrenia;
clear defective
Insight for the illness exists.)
completely
very serious neurotic
place in
or on sick-pension.
degree; neuroses with rather pronounced
disturbances,
psychiatric
to occupy a dominant
or thrust aside on a temporary
dominate
Insight for the illness can exist or can be missing. Cannot
himself/herself. tivity
these are serious enough
are interfered
sporadically
depressions of moderate
disturbances, 6. Severely
Normal
seem necessary. conditions
the patient’s
adequately Examples:
life for
take care of serious de-
where almost all normal ac-
in institutionalized
persons who therefore
cannot take care of themselves.) 7. Most severely Insight
ill. (Very seriously ill persons; hospitalized,
for the illness is lacking.
fused state. Examples:
blooming
The patient
most often without
is out of contact,
patient’s
consent.
e.g., in a stupor or a very con-
psychoses, grave defects which are impossible to rehabilitate.)
RESULTS
The results are presented in terms of the four questions raised above. Do the female patients have unusual ofsspring sex ratios? As shown in Table 2, the percentage of male offspring among the total 243 offspring of all patients was 56.4%, which is slightly higher than the approximately 51.5% male live births expected among the Swedish general population (calculated for years 1957-1967, per reference 13, p. 76). While the number of offspring in each of the specific diagnostic groups is small (34 to 57 per group), all groups had at least a mild excess of male offspring (52.1% to 6 1.8% male). Is there a predominance of one sex among o&spring born before vs. after mother’s$rst psychiatric hospitalization? As shown in Table 2, the offspring sex ratios for all patients combined are similar before and after mother’s first hospitalization. Patients in the endogenous psychoses group showed little difference in the sex ratio of offspring born before vs. after first hospitalization. Among the other patients group, those with diverse diagnoses had an overweight of female offspring (62.1 ‘%J)before first hospitalization and an overweight of male offspring (73.7%) after first hospitalization (x2 = 5.88, 1 df, p < 0.02).* This pattern was reversed for psychopaths, who had more males (74.1%) before first hospitalization and more females (57.7%) after hospitalization (x2 = 5.51, 1 df, p < 0.02).* *More
than one reproduction
to the same subject were included together
in chi-square
analysis
since the sex of consecutive offspring for the same woman was not more consistent or “dependent” than would be expected by chance.
34
psychoses
psychoses
23 schizophrenic-like
72 total endogenous
101
51 total other
123 total patients
243
53
27 psychopaths
patients
48
24 diverse diagnoses
142
57
51
psychosis
137
56
31
25
81
31
21
29
Total
N males
Total
Before
56 157
56.4%
27
29
101
40
25
36
N births
89
31
20
11
58
24
15
19
After
45 86
56.7%
26
19
41
17
9
15
N births
55.4%
74.1%
37.9%
57.4%
60.0%
60.0%
52.8%
% males
55.6% 55.8%
48
42.3%
73.7%
56.1%
43.2%
66.7%
66.7%
25
11
14
23
7
6
10
% males
Hospitalization
Hoylitalitation
N males
First
Delivery
First Psychiatric
of Offspring
Hospitalization
N males
First
Timing
before vs. after Mother’s
55.4%
58.5%
52.1%
57.0%
54.4%
61.8%
56.9%
% males
Total
for Reproductions
N births
19 manic-depressive
Group
per
30 process schizophrenia
Diagnosis
of Mothers
2. Sample Sizes and Sex of Offspring
Number
Table
74
MCNEIL. KAIJ. AND
PERSSON-BLENNOW
Is there a predominance of male or female oflspring corresponding with dlxerent levels of active maternal mental disturbance during pregnancy and/or postpartum?* During pregnancy: No difference was found in offspring sex ratio
for all patients not disturbed (categories 1-2, Table 1) vs. those in a disturbed category (categories 3-7) during pregnancy (x2 = 0.06, 1 df, n.s.t). Analysis of the maternal-disturbance-sex-ratio relationship for the specific diagnostic groups was restricted by the low number of women disturbed during pregnancy, but for most of the diagnostic groups (and definitely for the endogenous psychoses) offspring sex ratio was very similar for disturbed subjects and for those not disturbed. Postpartum: Porportionately more women were disturbed, especially seriously disturbed (categories 5-7), postpartum as contrasted with during pregnancy. Little difference in offspring sex ratio was found between all patients disturbed (categories 3-7) and those not disturbed (x” = 1.24, 1 dJ n.s.). Within the specific diagnostic groups, little difference was found in offspring sex ratio between those disturbed and those not disturbed. Is there an imbalance in ofspring sex ratio among mothers whose mental disturbance developed or became more severe postpartum, as contrasted with mothers whose mental status remained the same or improved postpartum? A total of 55 (22.7%) of the 2421 reproductions were followed by increased disturbance (44 of the 55 had been rated “not disturbed” during pregnancy), 166 (68.6%) reproductions were followed by no change postpartum, and 21 (8.7%) reproductions were followed by less severe disturbance postpartum (17 of the 21 were rated “not disturbed” postpartum). For all patients, 28 males and 27 females were born to women developing a more severe mental status postpartum. The combined group of process schizophrenics, schizophrenic-like psychotics, and manic-depressives who became more severely disturbed postpartum showed a nonsignificant tendency (x2 = 2.38, 1 dL n.s.) toward more female offspring (especially for schizophrenics and SLPs), as contrasted with the proportion of females born to schizophrenics, SLPs, and manic-depressives showing no change or showing improvement postpartum. The patients in the diverse diagnoses group showed a predominance of male offspring, both among those becoming more severe postpartum and among those becoming less severe postpartum. Psychopaths showed a predominance of male offspring, both among those becoming more severe and among those showing no change postpartum. DISCUSSION
The current process
study found no predominance
schizophrenics
or
schizophrenic-like
of female offspring for female psychotics
or
any
other
patient
*To conserve journal space, tabular information relevant to the third and fourth questions is not presented here, but may be obtained from the authors upon request. t More than one reproduction to the same subject were included together in chi-square analysis since the sex of consecutive than would
be expected
$Due to a partially for the third and fourth
offspring
for the same woman
was not more consistent
or “dependent”
by chance. missing
hospital
questions.
record,
one reproduction
had to be excluded
from
the analyses
OFFSPRING
SEX AN0
MATERNAL
DISTURBANCE
75
group (contrary to Lanes); no predominance of female offspring for process or schizophrenic-like psychotics before schizophrenics, manic-depressives, (rather than after) first hospitalization (contrary to Lane5); no predominance of female offspring for patients actively disturbed during pregnancy (contrary to Shearer et al.’ and to Taylor and Levine3); and no predominance of male offspring for patients disturbed or becoming more disturbed postpartum (contrary to Taylor and Levine3). The current study thus supports the results of Schorer4 and Melges,‘O notably Melges’ finding (p. 1027) of a slight excess of female offspring for schizophrenics becoming disturbed postpartum. It should be remembered in comparing the current study with the reviewed literature that the current study differed from other studies in several respects. The currently used time periods of 10 months prepartum and 10 months postpartum were considerably less time-specific than were the l- to 2-month periods used by Shearer et al. and Taylor and Levine. * Furthermore, these authors restricted cases to those “developing schizophrenic symptomatology,” while the current disturbances also included already existing disturbance and lesser degrees of disturbance (per Table 1) among patients at some time defined as schizophrenic or schizophrenic-like psychotic. Perhaps as a result of broader limits on both time and disturbance, the current study found active mental disturbance among 14% (15/107) of the reproductions for schizophrenics and SLPs (combined), as contrasted with, for example, only 4.1% (14/343) of known reproductions among the sample of Shearer et al. Furthermore, the current study of offspring sex in all known reproductions for the women, rather than only reproductions associated with the development of symptomatology, allowed not only study of more general deviations in offspring sex ratio (e.g., Lanes), but also the possibility of using the psychotic groups during nondisturbed reproductions as their own control. Such control guards against a predominance of offspring of a given sex, independent of maternal mental status near reproduction. For example, the predominance of male offspring among psychopaths becoming more severely disturbed postpartum (a noteworthy finding in terms of the Taylor hypothesis) was paralleled by the predominance of male offspring among psychopaths showing “no change” in mental status postpartum. Inclusion of diagnostic groups in addition to schizophrenics not only potentially provides the opportunity to test the specificity of maternal-disturbance-offspring-sex relationships for schizophrenics (as Melges noted, p. 1027). but such additional patient groups allow the study of more general theoretical positions, such as Lane’s5 theory of parental stress as the factor underlying the predominance of female offspring of schizophrenic parents. If the theory of stress were correct, other types of patients under stress should also show a predominance of female offspring. Unfortunately for testing the specificity-toschizophrenia hypothesis, the only reviewed studies including patients in addition to schizophrenics (Melges’ and the current study) found no significant relationship between maternal disturbance and fetal sex among any diagnostic group, including schizophrenics. ‘AS Melges’” postpartum
pointed out (p. 1027). there is nothing hormonally
period studied by Taylor
and Levine..’
specific about the more restricted
MCNEIL, KAIJ. AND
76
PERSSON-BLENNOW
SUMMARY
The relationship between sex of offspring born to female mental patients and the degree of active maternal mental disturbance near reproduction was studied for 243 reproductions by 123 patients. Severity of active mental disturbance was judged on a seven-point general psychopathology scale, based primarily on information from all known psychiatric records. All diagnostic groups showed a slight excess of male over female offspring; offspring sex was unrelated to degree of active maternal mental disturbance both during pregnancy and postpartum; and a slight overweight of female offspring was found among endogenous psychotics developing a more severe disturbance postpartum. REFERENCES 1. Shearer ML, Davidson RT, Finch SM: The sex ratio of offspring born to state hospitalized schizophrenic women. J Psychiatr Res 5:349350, 1967 2. Taylor MA: Sex ratios of newborns: Associated with prepartum and postpartum schizophrenia. Science 164:723-724, 1969 3. Taylor MA, Levine R: Puerperal schizophrenia: A physiological interaction between mother and fetus. Biol Psychiatry I:977101, 1969 4. Schorer CE: Gestational schizophrenia. Can Psychiatr Assoc J 17:SS-259-263, 1972 5. Lane EA: The sex ratio of children born to schizophrenics and a theory of stress. Psycho1 Ret 19:579-584, 1969 6. Mednick SA, Mura E, Schulsinger F, et al: Perinatal conditions and infant development in children with schizophrenic parents. Sot Biol 18:103-113, 1971 7. Erlenmeyer-Kimhng L: Mortality rates in the offspring of schizophrenic parents and a physiological advantage hypothesis. Nature 220:798-800, 1968
8. McNeil TF, Kaij L: Reproduction among female mental patients: Obstetric complications and physical size of offspring. Acta Psychiatr Stand 50:3-15, 1974 9. Taylor MA: Sex ratios of newborns and schizophrenia. Science 168:15I-152, 1970 10. Melges FT: Postpartum psychiatric reactions: Time of onset and sex ratio of newborns. Science 166:1026-1027, 1970 1I. McNeil TF, Persson-Blennow I, Kaij L: Reproduction in female psychiatric patients: Severity of mental disturbance near reproduction and rates of obstetric complications. Acta Psychiatr Stand 50:23332, 1974 12. McNeil TF, Kaij L: Obstetric complications and physical size of offspring of schizophrenic, schizophrenic-like, and control mothers. Br J Psychiatry 123:341-348, 1973 1970 (Statistical 13. Statistisk Arsbok Abstract of Sweden). Stockholm, Norstedt & Saner, 1970