| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Department of Obstetrics and Gynecology, University Hospital of Norrland, S-901 85 Umeå, Sweden.
Address all correspondence and requests for reprints to: Torbjörn Bäckström, M.D., Ph.D., Professor, Department of Obstetrics and Gynecology, University Hospital of Norrland, S-901 85 Umeå, Sweden. E-mail: torbjorn.backstrom{at}obstgyn.umu.se
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
-aminobutyric acid-A
receptors), as well as on protein synthesis via genomic action (7, 8).
Consequently, it is reasonable to suspect that E2 and P may
influence mood and mental symptoms. The aim of the present study was to
investigate whether a hormone-symptom relationship exists between
individuals, as was noted between cycles within individuals in an
earlier study (5). As indicators of feedback activity and to elucidate
the driving source for ovarian steroids, gonadotrophins also were
investigated. |
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Thirty patients, 37.8 yr old (range, 27–44 yr), were recruited among women seeking help for PMS at the gynecological outpatient department. All of them showed significant cyclical mood changes and fulfilled the criteria for premenstrual dysphoric disorder in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, DSM-IV (1, 9). They underwent 2 consecutive months of clinical evaluation by prospective daily rating of their symptoms, using a previously validated visual analogue scale (VAS) (10). All patients provided blood samples for hormone assays in, at least, one cycle on cycle days 1–4, and from cycle day 10 throughout the rest of the cycle. In patients who provided two cycles with blood samples, the first cycle was used in the present study. Twelve of the patients participated in the previous study comparing symptoms within individuals between cycles (5). None of the patients received any medication, and they were otherwise healthy. They completed a questionnaire concerning their previous mental and physical health, education, marital status, PMS history, and an Eysenck personality inventory (EPI) (11). All cycles studied were ovulatory, as defined by plasma P values above 15 nmol/L. The day of ovulation was determined as the day after LH surge. The study was approved by the Umeå University Ethics Committee, and informed consent was given by each participant.
Daily rating scale
Every evening, the women completed a previously validated VAS. In total, five negative mood symptoms (depression, anxiety, tension, fatigue, and irritability) and three somatic signs (breast tenderness, swelling, and headache) were rated, as well as the severity of menstrual bleeding (9). For each item, the patient marked, on a 10-cm-long scale (graded 0–10) the severity of a suffered symptom. Zero was defined as the complete absence of the symptom, and 10 as the maximal severity of symptom, as experienced by the patient in an ordinary menstrual cycle. The women were asked also to describe events during the day that might have influenced their mood and to record any medication taken.
Hormone analyses
Analyses of E2, P, FSH, and LH were performed on the
serum samples by RIA (12). E2 was measured using an
antiserum against an E2 17 ß-3-oxime-BSA derivate
(purchased from Miles-Yeda, Rehovot, Israel). The intra- and interassay
coefficients of variation were 10% and 12%, respectively. For P
assays, an antiserum against P 11
-succinyl BSA (Endocrine Science,
Tarzana, CA) was used. The intra- and interassay coefficients of
variation of P assay were 8% and 11%, respectively. FSH and LH were
analyzed by double-antibody RIA (Farmos Diagnostica, Oulu, Finland).
The standard for serum LH RIA was human pituitary LH, WHO 68/40; and
for FSH, WHO 69/104. The intra- and interassay coefficients of
variations were 8.7% and 8.7% for LH and 8.5% and 10.5% for FSH.
All samples were analyzed in duplicate.
Data analysis
The mean values of E2, P, FSH, and LH during the last five days in the premenstrual phase were calculated per patient. Patients were ranked according to the mean serum concentration of the hormone in question. Thereafter, for each hormone, patients were divided into groups with high or low hormone levels, by using the median split method, and were identified in the following text as high- or low-hormone group, respectively. In one patient, FSH and LH were not analyzed, and she was excluded from the statistical evaluation.
The number of symptoms expressed per day and summarized symptom scores for negative mental and physical symptoms were calculated. The number of expressed symptoms per day can be compared between individuals, because the absence of a symptom is considered to be equal for each subject. It is problematic to compare VAS symptom scores between subjects, because the scores are subjectively defined, and therefore, the actual scores do not necessarily indicate the same severity between individuals. VAS scores, however, are frequently used as an absolute measure, even between individuals (13, 14, 15). Therefore, we give the summarized symptom score for the patient groups, as well. For each patient and day, the expressed symptoms were counted. The number of expressed mental symptoms per day can vary from 0–5, and expressed physical symptoms can vary from 0–3. A summarized symptom score was calculated by adding the symptom VAS scores for each patient and day. The summarized score for negative mental symptoms and physical symptoms can range from 0–50 and 0–30 per day, respectively. Differences in symptoms per day between high- and low-hormone groups were tested, using two-way ANOVA, over the last 10 premenstrual days, with the factors’ premenstrual days x hormone high or low group. Differences in mean hormone levels for the last 5 premenstrual days were tested using ANOVA. Preovulatory peak values of FSH and LH, as well as the results from the EPI, were tested using the Mann-Whitney U-test, because normal distribution was not assumed. A P-value < 0.05 was considered significant.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Both the number of expressed symptoms per day and summarized
symptom scores per day showed a similar pattern of change through the
menstrual cycle. The symptom parameters showed decreased severity
during the first 3 days of menstrual cycle and increased severity after
ovulation, with a maximum during the last 5 premenstrual days (Fig. 1
).
|
The patients with high luteal phase-E2 concentration
(high-E2 group) were compared with patients with lower
luteal E2 (low-E2 group). More mental symptoms
were expressed, and higher summarized mental symptom scores were found
in the high-E2 group, compared with the low-E2
group (Fig. 2, Table 1). The three summarized physical
symptoms were also higher in the high-E2 group (Table 1).
Of the individual symptoms, anxiety, depression, tension, and fatigue
were more often expressed in the high-E2 group, compared
with the low-E2 group [F (1, 296–297) = 38.3–20.1,
P < 0.001; data not shown]. The scores of all
individual mental symptoms were higher in the high-E2 group
[F (1, 296–297) = 42.5–8.8; P < 0.001–0.005; data
not shown].
|
|
In the high-P group, the summarized mental symptom scores per day were
lower, compared with the low-P group (Table 1
). However, the summarized
somatic scores did not differ between the two groups. The individual
and total number of expressed symptoms (mental or somatic) did not
differ between the two groups. Of the individual mental symptoms,
tension, anxious, and fatigue were lower in the high-P group [F (1,
296–297) = 6.6–7.18; P < 0.01–0.05; data not
shown].
In the high-LH group, mental symptoms were more often expressed, and
the summarized mental symptom scores were higher, compared with the
low-LH group (Table 2). Among individual
summarized symptoms, depression and anxiety were more pronounced in the
high-LH group [F (1, 276–277) = 34.5–5.36; P <
0.001–0.05]. The low- and high-FSH groups did not differ in mean
number of expressed (mental or somatic) symptoms. Summarized mental and
physical symptom scores were not different between the groups (Table 2).
|
and 2). No significant differences were seen, between
high and low groups, for any of the hormones. The ages of the hormone
groups were compared, and the mean age in the low-E2 group
was 35 ± 1.5 (SEM) yr vs. 37 ± 1.6
yr in the high group. The difference was not significant for any
hormone. Hormones
The high-E2 group had a higher level of LH and FSH
during the late luteal phase, compared with the low-E2
group (Table 1). The low-P group had a higher preovulatory LH peak
(Table 1). The levels of E2 and FSH were higher in the
high-LH group (Table 2). The high-FSH group showed higher values for
E2 and LH, but lower P, compared with the low-FSH group
(Table 2).
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
More mental symptoms were seen in the high-LH group than in the low-LH group. This may be caused by the fact that the individuals are distributed similarly between the groups. The relation between high LH and symptoms, therefore, may be a confounder to the E2-symptom relation. No relation was, however, found between FSH levels and mental or physical symptoms. In line with an earlier report (5), the relation between serum P and symptoms was not as evident as for E2. Because symptoms are only present in cycles where a corpus luteum is developed (4, 20), a factor from the corpus luteum must be involved in provoking the symptoms. Moreover, cyclical mood changes develop when the progestogen is added to a sequential postmenopausal treatment (16, 18). On the other hand, P by it self, without the presence of estrogen, has not been shown to have a negative effect on the mood (21).
A high E2 concentration was related to high LH and FSH
levels. High LH and E2 concentrations coincided, suggesting
that LH regulated the E2 level and not the opposite. If the
ovary were primarily driving the E2 production, one would
have expected a low LH and FSH, because of negative feedback. Because
the opposite was found, we suggest that the pituitary is producing more
LH, which is stimulating the corpus luteum to produce more
E2. An interesting observation is that women in the
perimenopause have higher estrogen and LH secretion, compared with
younger women, during the luteal phase. This supports our finding that
it is possible to find situations where both LH and E2 are
increased simultaneously (22). In this study, the mean age was 2 yr
higher in the high-E2 group than in the low group, but the
difference was not significant. Therefore, we believe that the
difference in E2 and LH was not caused only by a difference
in chronological age. Biological age might, however, be different. Age
seems also to be of importance for symptom severity in PMS, because
more sever symptoms are expressed with increasing age (2). The finding
of a high LH, together with a high E2, in the more
symptomatic patients suggests that the latter patients are less
sensitive to E2-negative feedback. Interestingly, the more
symptomatic PMS patients are also less sensitive to benzodiazepine and
pregnanolone in a -aminobutyric acid-A receptor-linked central
nervous system function (23, 24).
The Eysenck personality inventory was compared between the groups. There were no observed differences in the neuroticism score between groups, indicating that differences in personality could not explain their disparity of symptom severity.
To estimate the severity of subjective symptoms, the number of expressed symptoms per day and the summarized symptom score per day were calculated. The validity of symptom scores between individuals may be questioned, because no absolute reference value, except absence of symptoms, can be obtained. Only few attempts have been made in the literature to find methods with which to estimate the severity of PMS. Methods presented as a way to estimate severity have, in fact, often been methods to establish cyclicity (25). However, in one study, all scores for physical and behavioral symptoms were summarized to estimate the severity of PMS (26). This is similar to the way summarized scores are presented in this study, but we distinguish between mental and physical symptoms.
The International Headache Society has suggested a way of estimating
the severity of headache, in which patients are asked to keep a diary
regarding the presence or absence of headache symptoms. The number of
days with headache during a certain time period is then used to
classify the severity of symptoms. The higher number of days with
headache, the more severe is the condition (27). Because PMS is also a
subjective syndrome, this may be a way to measure its severity. To
count the number of days with or without symptoms, during a
premenstrual period, would give a figure that is comparable between
patients, and this is used in the present study. Severino and
co-workers assessed significant cyclicity by using spectral density
analysis. To grade the severity, they counted, for each patient, the
number of symptoms with significant premenstrual spectral density. A
higher number of significantly expressed symptoms was considered to
indicate a more severe PMS (28). A combination of these two methods is
to count the number of symptoms expressed per day during a certain time
period (29), which was done in this study, over a time period of 10
premenstrual days. Both methods gave very similar results, as shown in
Fig. 1.
In conclusion, high levels of E2 and LH in the luteal phase seem to be related to more severe symptoms in patients with PMS. Intervention studies will be needed to show weather there is a causal relationship between high E2 and P, on one hand, and more severe PMS symptoms, on the other. The higher levels of E2 may be caused by increased LH stimulation of the corpus luteum.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received November 26, 1997.
Revised March 4, 1998.
Accepted March 13, 1998.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. Thys-Jacobs, D. McMahon, and J. P. Bilezikian Differences in Free Estradiol and Sex Hormone-Binding Globulin in Women with and without Premenstrual Dysphoric Disorder J. Clin. Endocrinol. Metab., January 1, 2008; 93(1): 96 - 102. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Bjorn, I. Sundstrom-Poromaa, M. Bixo, S. Nyberg, G. Backstrom, and T. Backstrom Increase of Estrogen Dose Deteriorates Mood during Progestin Phase in Sequential Hormonal Therapy J. Clin. Endocrinol. Metab., May 1, 2003; 88(5): 2026 - 2030. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
