Psychiatry and Clinical Neurosciences (1998), 52, 303±309

Regular Article
The in¯uence of obsessive-compulsive neurosis patients'
premorbid personality on obsessive-compulsive symptoms
and ef®cacy of medication
RYOJI NISHIMURA, MD,1 TATSUYA HOSOBA, PhD,2 MASAAKI IDE, PhD,2 AND HIDETOSHI SEIWA, PhD3
1
Division of Occupational Therapy, Department of Health Sciences, School of Medicine, 2Graduate School of Biosphere Science and 3Department of
Behavioral Sciences, Faculty of Integrated Arts and Sciences, Hiroshima University, Higashi-Hiroshima, Japan

Abstract We conducted a survey with the Lyn®eld obsessive-compulsive symptom questionnaire (revised version) on 48
obsessive-compulsive neurotic patients as the survey subjects. In the factor analysis ®ve factors of obsessions, were
identi®ed: (i) the desire for perfection; (ii) compulsive checking; (iii) washing; (iv) feelings of uncleanliness; and
(v) anthropophobia. High correlations were noted between these factors. We also investigated the premorbid
personalities of obsessive-compulsive neurotic patients with a multidimensional personality scale and obtained an
extroversion dimension and neuroticism dimension. The in¯uence of these premorbid personality dimensions on
obsessive-compulsive symptoms became clear; (i) neuroticism is related to the levels of obsession after onset, but
not related to compulsive behaviors; and (ii) No differences in premorbid personality dimensions were noted
between compulsive checking and compulsive washing behaviors. We also studied whether it was possible to
predict the ef®cacy of pharmacotherapy upon obsessive-compulsive symptoms. It was elucidated that the
obsessions of those whose premorbid personalities are emotionally stable and extroversive are susceptible to
antidepressants. Based on these results, we discussed the usefulness of premorbid personalities in predicting
diversity of obsessive-compulsive symptoms, as well as in prediction the ef®cacy of medication.

Key words ef®cacy of medication, obsessive-compulsive neurosis, premorbid personality.

INTRODUCTION relationship between premorbid personalities and obsessive-

An obsessive-compulsive neurosis patient presents with diverse
symptoms that are dif®cult to treat. Conventionally, such traits as
meticulousness, obstinacy and nervousness have been studied as
common to the personalities of obsessive-compulsive neurotic
patients.1 Recently, a number of traits such as `depression and
sense of responsibility',2 `neuroticism, depression and sense of
responsibility',3 and `sense of responsibility, feelings of guilt,
perfectionism and risk avoidance',4 have been considered to
contribute to the formation of obsessive-compulsive symptoms.
Among studies of premorbid personalities of obsessive-com-
pulsive neurotic patients, the study of its possible in¯uence on the
therapeutic process is attracting clinical attention. Nishizono
et al.5 studied whether the premorbid personalities of obsessive-
compulsive neurotic patients in¯uence the ef®cacy of psycho-
pharmaceuticals, and reported that affective tension and
excitement, which form the basis of the obsessive-compulsive
symptoms, were alleviated by using different psychopharma-
ceuticals, depending on the symptom characteristics and
premorbid personalities. In this study, we investigated the
Correspondence address: Ryoji Nishimura, MD, 1-2-3, Kasumi, Minamiku, Hiro-
shima 734, Japan.
Received 20 August 1998; revised 13 November 1998; accepted 18 November
1998.
compulsive symptoms in obsessive-compulsive neurotic patients
focusing on the following three points: (i) personality dimensions
constituting the premorbid personalities of obsessive-compul-
sive neurotic patients; (ii) in¯uence of premorbid personality
dimensions on obsessive-compulsive symptoms; and (iii) in¯u-
ence of premorbid personality dimensions on the alleviation of
obsessive-compulsive symptoms through medication.
SUBJECTS AND METHODS
Subjects
Forty-eight obsessive-compulsive neurotic patients who had
consulted either the Department of Psychiatry of Fukuoka
University Hospital between 1977 and 1989 or the Neuro-
psychiatry Department of Hiroshima University Hospital
between 1991 and 1995 were chosen as subjects. Patients with
Tourette's syndrome and obsessions in schizophrenia were
excluded from the study. The group consisted of 29 males and
19 females. Their mean age was 25.5 years (SD ˆ 11.26).
Distinctions between inpatients and outpatients were not
made. Their main treatment was pharmacotherapy with the
use of an antianxiety agent (bromazepam), antidepressant
(clomipramine), and antipsychotic (haloperidol). The
304
medication was administered according to the judgement and
discretion of the respective doctors. Supportive psychotherapy
was conducted with pharmacotherapy in each case.
Questionnaire and response method
Obsessive-compulsive symptom evaluation scale
In order to objectively and quantitatively investigate obsessive-
compulsive symptoms, the Lyn®eld obsessive-compulsive
symptom questionnaire6 was used with one category,
anthropophobia, added to meet the actualities of obsessive-
compulsive neurosis in Japan. A total of 21 items were rated by
a 5-grade method, where `utterly inapplicable' was scored as 0
and `completely applicable' was scored as 4. The Lyn®eld
obsessive-compulsive symptom questionnaire does not include
the scale of primary obsessive slowness.7
Personality traits measuring scale
We employed a scale that had originally been developed by
Cattel for the measurement of 16 traits and was modi®ed by
Nishizono et al.8 for measurement of 15 traits for use in Japan.
In this pro®le, two traits were placed in opposite poles for each
of 15 trait dimensions. A calculation was made by scoring 1 to
the more frequently selected of the two polar traits in each
dimension, with 0 given to the other. In this way, we obtained
data for 30 traits (15 dimensions ´ 2 traits) per individual
subject and used them in analysis.
Survey method
The responses to the respective scale questionnaires were given
by the doctors in charge of the subjects. Initially, they entered
their responses in both the personality scale and Lyn®eld
obsessive-compulsive symptom questionnaire. The Lyn®eld
obsessive-compulsive symptom questionnaire was responded to
in every therapeutic session thereafter. The medicines admin-
istered in each therapeutic session were also noted by the
doctors.
RESULTS
Review of obsessive-compulsive symptoms
by factor analysis
Table 1 shows the factor analysis results based on the Lyn®eld
scale scores of the 48 obsessive-compulsive neurotic patients at
the beginning of the therapy. The items listed in Table 1 are
limited to those that presented a high loading with one of the
factors. Because of the speculated correlations between differ-
ent obsessive-compulsive symptoms, an oblique rotation via
the promax technique was adopted in the factor analysis. Five
factors were extracted based on the variances in each value,
which explained 68.45% of the entire variance.
With Factor 1, the items related to washing and feelings of
uncleanliness including No. 14: `Do you feel uncomfortable
R. Nishimura et al.
unless you always keep your hands clean?' presented a high
loading. We therefore named Factor 1 the `washing and feel-
ings of uncleanliness' factor. Factor 2 was considered the
`checking behavior' factor because of high loading presented
by checking-behavior-related items such as No. 16: `Do you
become anxious about whether you shut off the gas valve,
water valve and other switches, and return to check them, even
if you know you actually did it?' For Factor 3, items that
related to `obsessions' had a high loading, which included such
items as No. 19: `Do odious words repeatedly come back to
your mind?' These three are commonly reported factors in
research based on factor theory using other obsessive-
compulsive symptom questionnaires; namely, Maudsley's
obsessive-compulsive neurosis questionnaire9 or Padova's
questionnaire,10 and are deemed, in their contents, to be highly
valid as factors forming the obsessive-compulsive symptoms.
For Factor 4, items with a high loading were mostly related
to `anthropophobia', such as No. 21: `Do you sometimes ex-
perience inconvenience due to ereuthophobia or anthropo-
phobia?' The items having high loading for Factor 5 were those
related to the `desire for perfection', including No. 6: `Do you
become worried that you may have made a mistake or have
done something incompletely, even though you had been
quite careful?' Frost et al.11 cited perfectionism, a propensity for
being displeased with anything that is not perfect, as the central
trait of obsessive-compulsive neurotics, the desire for perfec-
tion is considered to be a factor closely related with obsessive-
compulsive neurosis.
Each factor was examined for internal consistency. Four
factors other than the desire for perfection presented 0.700 or
higher a coef®cient, which was satisfactory to prove internal
consistency among the items consisting these factor groups.
The a coef®cient for the `desire for perfection' factor was
slightly low (0.667).
Examination of premorbid personality structure
by correspondence analysis, and classi®cation
of obsessive-compulsive neurotic patients
The 30 premorbid personality trait dimensions were formed
into a structure by using correspondence analysis to examine
central personality dimensions in the obsessive neurotic
patients' premorbid personalities. Premorbid personalities were
surveyed for 36 out of 48 subjects.
Figure 1 shows the results of correspondence analysis con-
ducted based on the data for 30 categorical traits obtained from
the responses to the personality scale questionnaire. The 2-
dimensional resolution presented 41.70%, which explained
nearly half of the data variance. Dimension 1 is mainly related
to traits of extroversion and self-reliance, such as sociability,
emotional stability, self-reliance-bashfulness, emotional insta-
bility and dependence. We called this the extroversion
dimension. Dimension 2 is mainly related to traits of
neuroticism, such as healthy superego, low tension-immature
superego, high tension, anxiety. We called this the neuroticism
dimension. The correlation between these two dimensions
was examined. It revealed a low correlation coef®cient of
Obsessions and premorbid personality 305

Table 1. Factor analysis of a modi®ed version of Lyn®eld Obsessive-Compulsive Symptom Questionnaire (Promax Rotation)

Factor 1 Factor 2 Factor 3 Factor 4 Factor 5 Communality

Washing/feeling of uncleanlines (a = 0.841)

14. Do you feel uncomfortable unless you always keep your hands clean? 0.751 )0.128 )0.034 0.151 0.370 0.792
13. Do you feel uneasy unless your clothes are neat and clean, whatever 0.749 0.010 0.040 )0.073 )0.030 0.561
you are doing?
2. Do you have to pay particular attention to cleanliness in your 0.726 )0.147 )0.108 )0.093 0.239 0.633
everyday life?
11. Do you feel uncomfortable staying in a untidy or dirty room even 0.705 0.210 0.108 )0.091 )0.269 0.603
for a short time?
12. Do you feel uncomfortable unless inside your house is kept tidy and 0.665 0.209 )0.010 0.302 )0.045 0.665
clean all the time?
10. Do you feel disturbed if a job is not done in a pre-determined time 0.478 0.208 0.162 0.263 )0.125 0.559
and order?
Checking behaviour (a = 0.742)
16. Do you become anxious about whether you shut off the gas valve, )0.080 0.902 0.007 )0.079 0.152 0.798
water valve or switches, and return to check them, even if you know
you actually did it?
15. Do you become worried and go back to close windows and doors, 0.122 0.852 )0.200 0.019 0.134 0.743
even if you know you already did it?
9. Do you repeat an action many times until you are fully satis®ed, 0.056 0.633 0.226 )0.193 0.354 0.672
even though you knew it was already done correctly?
Obsessions (a = 0.708)
19. Do odious words repeatedly come back to your mind? )0.017 0.069 0.875 0.018 )0.244 0.758
20. Even though you know it is unnecessary, do you ®nd you can't help 0.049 )0.051 0.813 )0.348 0.308 0.838
but think certain thoughts?
5. Do you terribly regret it when you cannot do as you had planned in 0.171 )0.304 0.599 0.243 0.080 0.537
your mind?
18. Do you sometimes become worried that an accident or some )0.016 0.329 0.527 0.156 )0.023 0.553
misfortune may have occurred to your children or other family members?
Anthropophobia (a = 0.751)
21. Do you sometimes experience in convenience due to ereuthophobia 0.052 )0.116 )0.121 0.935 0.102 0.801
or anthropophobia?
4. Do you become worried after talking to somebody that you may )0.259 0.329 0.135 0.817 0.165 0.795
have said something wrong?
Desire for perfection (a = 0.667)
8. Do you sometimes feel unsatis®ed unless you repeat the same thing )0.027 0.299 )0.242 0.006 0.735 0.564
many times behind others at work or school?
6. Do you become worried that you may have a mistake or done 0.128 )0.065 0.190 0.384 0.656 0.730
something incompletely, although you had been quite careful?
7. Do you check with yourself, thinking that you may have done )0.085 0.269 0.223 0.237 0.644 0.718
something against your conscience?
Explanatory variance with other factors' effects excluded 2.881 2.270 1.989 1.947 1.850
Explanatory variance with other factors' effects ignored 3.260 2.939 3.103 2.496 2.355

r ˆ 0.020, meaning that these two dimensions were inde-
pendent of each other.
In order to investigate the in¯uence of personality dimen-
sions on obsessive-compulsive symptoms and on pharmaco-
therapy, we divided the obsessive-compulsive patients into
four types based on the two personality dimensions. The group
of emotionally unstable and introversive patients was noted to
have nervous traits with high anxiety and tension (upper left in
Fig. 1). We therefore named this group the `unstable type'. The
introversive and weak-willed group of obsessive-compulsive
neurotic patients was named the `depressed type', as interper-
sonal avoidance and hypothymia were noted with them (lower
left in Fig. 1). The type with insensitive and extroversion
(upper right in Fig. 1) was considered the `elated type' in view
of their highly cyclic, hyperthymic and sociable traits. The
relaxed and extroversive group (lower right in Fig. 1) was
306
self-reliant and reality af®rmative, with a low level of anxiety
and tension, named the `stable type'. The proportion of the
respective types of patients were 10 unstable types, 11 depressed
types, six elated types and nine stable types.
In¯uence of premorbid personality
on obsessive-compulsive symptoms
Multiple regression analysis
In order to research how the two premorbid personality
dimensions affect obsessive-compulsive symptoms, we con-
ducted a multiple regression analysis with the two dimensions
as the explanatory variables. For criterion variables, the data for
each obsessive-compulsive symptom factor collected in the ®rst
session were used. In the results of analysis it was found that
the extroversion dimension contributed to anthropophobia
(b ˆ )0.634, P < 0.01), and neuroticism contributed to ob-
sessions (b ˆ 0.550, P < 0.01). This result suggests that: (i) the
more introversive a person is, the more easily he or she tends to
develop anthropophobia; and (ii) the more emotionally un-
stable and neurotic a person is, the more easily he or she tends
to develop obsessions. Between compulsive behaviours (com-
pulsive washing and compulsive checking), no difference was
accorded to the premorbid personality differences.
Analysis of variance
For a further detailed investigation of the in¯uence of pre-
morbid personalities on obsessive-compulsive symptoms, the
early levels of each symptom by premorbid personality type
R. Nishimura et al.
Figure 1. Two-dimensional Structure of
Obsessive-Compulsive Patients' Premorbid
Personalities. Thirty (15 ´ 2) personality
traits were laid out in two dimensions ac-
cording to factor loading with correspon-
dence analysis. Each personality trait is
indicated by a dot. The axis of abscissa
represents the Extroversion Dimension, and
the axis of ordinates represents the Neu-
roticism Dimension.
were studied. Figure 2 shows the score for each obsessive-
compulsive symptom and total score of obsessive-compulsive
symptoms, both obtained in the ®rst session and respectively
classi®ed according to the premorbid personality types.
Through a single-factor analysis of variance for each obsessive-
compulsive symptom, obsessions and anthropophobia indicated
a principal effect and tendency difference of premorbid per-
sonalities. (F [3, 32] ˆ 2.68, P < 0.10; F [3, 32] ˆ 4.36,
P < 0.01). Tukey's Honestly Signi®cant Difference (HSD) test
for obsessions and anthropophobia revealed respectively that the
unstable type tends to have a signi®cantly higher level of
symptoms than the stable type.
Incorporating these with the multiple regression analysis
results, the in¯uence of premorbid personality on obsessive-
compulsive symptoms can be summarized into the following
three generalizations: (i) the greater the premorbid neuroti-
cism, the higher obsession level at onset. Those who are extro-
versive with low neuroticism, in particular, are unlikely to
develop obsessions; (ii) No differences in premorbid person-
ality dimensions were noted between compulsive checking and
compulsive washing behaviors; and (iii) those who were more
introversive presented a higher level of anthropophobia after
onset. This tends to be particularly strong if the person rep-
resents a higher level of neuroticism.
In¯uence of premorbid personality and medication
on the alleviation of obsessive-compulsive symptoms
For the evaluation indexes of pharmacotherapy's alleviating
effect on obsessive-compulsive symptoms, level differences
from the initial levels of the ®rst session to the levels of the
Obsessions and premorbid personality 307

Figure 2. Initial Level of Each Obsessive-

Compulsive Symptom and Entire Obsessive-
Compulsive Symptoms for Each Premorbid
Personality Type. The left series of bars in the
®gure shows the initial levels of ®ve obsessive-
compulsive symptoms as indicated by a mean
score per symptom item for each premorbid
personality type. The right series shows the
total score of obsessive-compulsive symptoms
for each premorbid personality type. The
length of the line sticking out on top of each
bar graph indicates the standard error of mean.
Obsessive-compulsive symptoms were all as-
sessed by using the Lyn®eld Obsessive-Com-
pulsive Symptom Questionnaire, Revised
Version (Nishizono, Ushijima, and Nishimura,
1978). (j), Unstable type; ( ), depressed type;
( ), elated type; (h), stable type.

eighth-week session (last survey) were used. The indexes were DISCUSSION
examined in proportion to the initial level, where the initial
level is assumed to be 100. A 50% or greater decrease from the Diversity and concurrence
initial level was operationally de®ned in this study as a symp- of obsessive-compulsive symptoms
tom improvement. The administered drugs and doses were:
Diversity and concurrence of obsessive-compulsive symptoms
bromazepam (antianxiety agent) 7.73 ‹ 7.01 mg/day (n ˆ 15),
have already been con®rmed quantitatively in previous Euro-
haloperidol (antipsychotic) 9.33 ‹ 8.86 mg/day (n ˆ 10) and
pean and American research.9 Van Oppen et al.12 has reported
clomipramine (antidepressant) 74.82 ‹ 68.27 mg/day (n ˆ 23).
on the correlations between factors from Hodgson and
Table 2 shows the effect of medication on each obsessive-
Rachman's9 diversity studies and the concurrence of various
compulsive symptom as examined for improvement from the
symptoms. While there had been no such analysis reported on
initial level up to the level at the end of survey. Table 3 shows
obsessive-compulsive neurotic patients in Japan, our study of
the results of Chi-squared test according to premorbid per-
such Japanese patients presented similar results to the previous
sonality types measured against improvement or non-im-
studies. Three factors of obsessions: (i) compulsive checking;
provement of symptoms. Table 4 shows the relationship
(ii) washing; and (iii) feelings of uncleanliness were almost
between premorbid peronality and pharmacotherapeutic effects
identical in content with the previous ®ndings. The two factors
on obsessions.
of the desire for perfection and anthropophobia had not been
included in past ®ndings. The desire for perfection, as stated by
Table 2. Improvement ratio from initial level to post-session level and its
relationship with medications
Table 3. Improvement ratio from initial level to post-session level and its
Antianxiety agent Antidepressant Antipsychotic relationship with premorbid personalities

Obsessions NS v2(1) = 6.812à NS Premorbid personality ´ obsessions v2 (3) = 7.219+

Desire for perfection NS NS NS Premorbid personality ´ desire for perfection v2 (3) = 4.946 NS
Compulsive eating NS v2(1) = 5.711  NS Premorbid personality ´ compulsive checking v2 (3) = 1.450 NS
Wash/feelings of uncleanliness NS v2(1) = 2.820+ NS Premorbid personality ´ wash/feeling of uncleanliness v2 (3) = 3.452 NS
Anthropophobia NS v2(1) = 7.202à NS Premorbid personality ´ anthropophobia v2 (3) = 2.641 NS
Total Score NS v2(1) = 5.310  NS Premorbid personality ´ total score v2 (3) = 2.067 NS

+
P < 0.10;  P < 0.05; àP < 0.01. +
P < 0.10.
308 R. Nishimura et al.

Table 4. Breakdown of medications administered to patients with im- Through investigation of obsessive-compulsive symptoms
proved/unimproved obsessions (initial-post-sessions) by using these two dimensions, the following two points have
Medicines become clear: (i) those with higher levels of neuroticism in
Antianxiety + + + +
their premorbid personalities are more likely to develop ob-
Antidepressant + + +
sessions. This tendency cannot be found with other obsessive-
compulsive symptoms; and (ii) no differences in premorbid
Antipsychotic + + +
personality dimensions were noted between compulsive
Premorbid personality
checking and compulsive washing behaviours.
Non-improvement
The extroversion dimension was irrelevant to the levels of
obsessive-compulsive symptoms, which agrees with the pre-
Unstable 2 3 1 6
vious studies.14,15 However, it was noted from the results of
Depressed 1 4 1 1 7
analysis of variance that emotionally stable and extroversive
Elated 1 2 1 4
types have a low level of obsessions.
Stable 1 1
Previous studies have pointed out that neuroticism is highly
Total 1 6 4 4 3 18 related to the levels of obsessive-compulsive symptoms.3 Our
study revealed that its relationship is particularly close with
Improvement obsessions.
Unstable 2 2
Depressed 1 2 3
Elated 1 1 2
Effect of medication on each
obsessive-compulsive symptom
Stable 4 2 6
Table 2 shows the effect of medication on each obsessive-
Total 1 2 5 5 13
compulsive symptom as examined for improvement from the
+ indicates administration of that medicine, and ®gures represent the initial level up to the level at the end of survey. In the results of
number of patients. a Chi-squared test, according to administration or non-ad-
ministration of psychopharmaceuticals measured against im-
provement or non-improvement of symptoms, it was found
Frost et al.11 was highly correlated with other symptom factors that only the antidepressant was signi®cantly ef®cacious in
and is deemed to be closely related to obsessive-compulsive improving obsessive-compulsive symptoms as shown in each
neurosis. This study also elucidated that anthropophobia is a symptom factor, except for the desire for perfection, as well as
symptom recognised among obsessive-compulsive neurotic in anthropophobia and the total score. It has become evident
patients when examined by the factor theory. It would be more that antianxiety medications and antipsychotics do not have
agreeable to both of our analysis results and previous research ef®cacy on any of the symptoms. Previous research has cited
to consider anthropophobia as an anxiety disorder highly co- antidepressants as the most ef®cacious medicine for improve-
variant with obsessive-compulsive neurosis, as pointed out by ment of obsessive-compulsive symptoms, and that antianxiety
Marks.13 medications and antipsychotics were less effective.16 The results
In the results of correspondence analysis of obsessive-com- of our research support this position, revealing that the ef®-
pulsive neurotic patients' premorbid personalities based on the cacy/lack of ef®cacy of these medications did not depend on
categorical data, we obtained two dimensions of extroversion the kinds of obsessive-compulsive symptoms.
and neuroticism as the principal factors constituting the
structure of premorbid personalities of obsessive-compulsive
Premorbid personality type and improvement
neurotic patients.
of symptoms
Previous studies have pointed out that the extroversion
dimension was unrelated to the levels of obsessive-compulsive Table 3 shows the results of a Chi-squared test according to
symptoms. However, our study results suggest that this is a premorbid personality types measured against an improvement
dominant dimension forming the morbid personalities of ob- or non-improvement of symptoms. The symptom alleviating
sessive-compulsive neurotic patients. in¯uence of premorbid personality types was noted, with
Concerning the neuroticism dimension, previous studies tendency differences, only on obsessions. In addition, we in-
have also reported that among the diverse dimensions consti- vestigated whether there was any difference among different
tuting the premorbid personalities of obsessive-compulsive premorbid personality types regarding the pharmacotherapeu-
neurotic patients, this dimension is particularly high in obses- tic effect on obsessions (Table 4). Due to the limited number
sive-compulsive patients. The results of our study have made it of subjects, a statistical test was not conducted. In general, only
clear that this dimension forms the premorbid personality the stable type among the four premorbid personality types
structure of obsessive-compulsive patients. In addition, the indicated a high therapeutic effect from medication. Out of
correlation between these two dimensions was examined. It eight unstable type patients, pharmacotherapy was ef®cacious
revealed a low correlation coef®cient of r ˆ 0.020, meaning in only two. The two patients with improvements were ad-
that these two dimensions are independent of each other. ministered the antianxiety agent. Although other medicines
Obsessions and premorbid personality
they were inef®cacious. With the depressed type group of 10
patients, the medication therapy was effective with three. For
two out of these three patients, all three varieties of medicines
were used. Six patients belonged to the elated type, two of
whom improved with medication; one underwent antide-
pressant treatment, and another received combined medication
including the antidepressant. With the stable type, six out of
seven patients improved. All six were given the antidepressant
(including two who were under the combined use of three
medicines). The sole patient without improvement was ad-
ministered the antianxiety agent only.
These results can be summarized into the following two
points: (i) regardless of the kind of obsessive-compulsive
symptom, antidepressants provide the greatest therapeutic ef-
fect. However, they do not present any effect on the desire for
perfection; and (ii) Obsessions of the Stable Type are highly
sensitive to pharmacotherapy.
The in¯uence of premorbid personality on medication ef-
®cacy, was particularly strong in obsessions. Many of the pre-
vious studies pointed out that, in general, antidepressants are
ef®cacious.16±18 In our study, the antidepressant's ef®cacy was
conspicuous in obsessions with extroversive and less neurotic
type premorbid personalities. This suggests the importance of
taking the premorbid personality of the obsessive-compulsive
neurotic patient into consideration when considering
pharmacotherapeutic sensitivity. Antianxiety agents and
antipsychotics did not directly contribute to alleviation of
obsessive-compulsive symptoms. However, Nishizono, et al.8
stated that antianxiety agents and antipsychotics are often ad-
ministered to patients who are defensive due to their strong
tension/anxiety in interpersonal situations. They pointed out
that these medicines are useful in alleviating the anxiety and
tension of obsessive-compulsive neurotic patients, even if they
do not have direct ef®cacy.
In the future it will be necessary to study the relationship
between premorbid personality and the ef®cacy of medications
with more controlled research.
REFERENCES
1. Pollak JM. Obsessive-compulsive personality: A review. Psychol. Bull.
1979; 86: 225±241.
309
2. Eysenck HJ, Barrett P, Wilson G, Jackson C. Primary trait mea-
surement of the 21 components of the P-E-N system. Eur. J. Psychol.
Assess. 1992; 8: 109±117.
3. Scarrabelotti MB, Duck JM, Dickerson MM. Individual differences in
obsessive-compulsive behaviour: The role of the Eysenckian dimen-
sions and appraisals of responsibility. Pers. Indiv. Diff. 1995; 18: 413±
421.
4. Frost RO, Steketee G, Cohn L, Griess K. Personality traits in sub-
clinical and non-obsessive-compulsive volunteers and their parents.
Behav. Res. Ther. 1994; 32: 47±56.
5. Nishizono M, Ushizima S, Nishimura R. Structure of obsessive
compulsive neurosis seen from the ef®cacy of psychotropic medica-
tion. Ann. Report Pharmacopsychiatr. Res. Found. 1978; 10: 207±214 (in
Japanese).
6. Allen JJ, Tune GS. Lyn®eld obsessional/compulsive questionnaire.
Scot. Med. J. 1975; 20: 21±31.
7. Veale D. Primary obsessive slowness. Br. J. Psychiatr. 1993; 162: 198±
203.
8. Nishizono M, Nishimura R, Sarada Y, Ushizima S. Multidimensional
structure of obsessive-compulsive neurosis seen from the ef®cacy of
psychotropic medication. Ann. Rep. Pharmacopsychiatr. Res. Found.
1979; 11: 206±213 (in Japanese).
9. Hodgson RJ, Rachman S. Obsessional-compulsive complaints. Behav.
Res. Ther. 1977; 15: 389±395.
10. Sanavio E. Obsessions and compulsions: The Padua Inventory. Behav.
Res. Ther. 1988; 26: 169±177.
11. Frost RO, Marten P, Lahart C, Rosenblate R. The dimensions of
perfectionism. Cog. Ther. Res. 1990; 14: 449±468.
12. van Oppen P, Hoekstra RJ, Emmelkamp PMG. The structure of
obsessive-compulsive symptoms. Behav. Res. Ther. 1994; 33: 15±23.
13. Mark IM. Fear, phobias and rituals. Oxford University Press, New
York, 1987.
14. Clark LA, Watson D, Mineka S. Temperament, personality and the
mood and anxiety disorders. J. Abnorm. Psychol. 1994; 103: 103±116.
15. Pollak JM. Relationship of obsessive-compulsive personality to ob-
sessive-compulsive disorder: A review of the literature. J. Psychol.
1987; 121: 137±148.
16. Nakazima T, Taga T. Pharmacotherapy in obsessive compulsive dis-
orders. Psychiatry 1994; 36: 54±58 (in Japanese).
17. Jenike MA, Baer L, Summergrad P, Weilburg JB, Holland A, Sey-
mour R. Obsessive-compulsive disorder: A double-blind, placebo-
controlled trial of clomipramine in 27 patients. Am. J. Psychiatr. 1989;
146: 1328±1330.
18. Jenike MA, Buttolph L, Baer L, Ricciardi J, Holland A. Open trial of
¯uoxetine in obsessive-compulsive disorder. Am. J. Psychiatr. 1989;
146: 909±911.