| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Departments of Psychiatry (S.S.G., K.C.L., J.L., C.A.P., A.J.P.) and Psychology (S.S.G., K.S.T., K.C.L.), University of North Carolina at Chapel Hill, Chapel Hill, NC.
Address correspondence and reprint requests to Susan S. Girdler, PhD, University of North Carolina at Chapel Hill, CB #7175, Medical Research Bldg. A, Chapel Hill, NC 27599-7175. E-mail: susan_girdler{at}med.unc.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
METHODS: Ten sexually abused women with PMDD were compared with 18 nonabused women with PMDD and 22 nonabused women without PMDD for hypothalamic-pituitary-thyroid axis hormone concentrations during the follicular and luteal phases of confirmed ovulatory cycles.
RESULTS: Compared with the women without PMDD, only the group of women with PMDD with sexual abuse showed greater variance in both cycle phases in thyroid-stimulating hormone concentrations and greater luteal phase variance in free and total thyroxine (T4) and reverse tri-iodothyronine (T3). In the group of nonabused women with PMDD, there was greater variance in follicular phase thyroxine-binding globulin concentrations compared with the group without PMDD. Women with PMDD with abuse had greater mean concentrations of total T3 and thyroxine-binding globulin, greater total T3/free T4 and free T3/free T4 ratios, and lower ratios of free T3/total T3 and free T4/total T4 than either of the other 2 nonabused groups. Greater total T3 concentrations and histories of major depression independently predicted premenstrual symptoms in all women with PMDD, together accounting for 31% to 38% of the variance in anxiety, anger, and depression ratings.
CONCLUSIONS: These results suggest increased conversion of T4 to T3 and increased binding of thyroid hormones in women with PMDD with previous sexual abuse. Abnormal total T3 concentrations may have pathophysiological significance in PMDD.
Key Words: premenstrual dysphoric disorder, • thyroid axis hormones, • sexual abuse, • menstrual cycle.
Abbreviations: PMDD = premenstrual dysphoric disorder;; HPT = hypothalamic-pituitary-thyroid;; TSH = thyroid-stimulating hormone;; T3 = tri-iodothyronine;; T4 = thyroxine;; SA = sexual abuse;; PTSD = posttraumatic stress disorder;; TBG = thyroxine-binding globulin;; PRISM = Prospective Record of the Impact and Severity of Menstrual Symptoms;; SCID = Structured Clinical Interview for DSM-III-R Axis I disorders;; MDD = major depressive disorder;; PS = posttraumatic stress disorder scale of the Minnesota Multiphasic Personality Inventory-2;; MMPI-2 = Minnesota Multiphasic Personality Inventory-2;; rT3 = reverse tri-iodothyronine;; RIA = radioimmunoassay;; SHBG = sex hormone-binding globulin.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
Most studies on HPT axis function in PMDD find no evidence that PMDD is associated with alterations in mean thyroid hormone concentrations (3,8–12). However, a number of studies did find that women with PMDD have greater variability in thyroid-stimulating hormone (TSH, thyrotropin) concentrations or TSH responsiveness to a thyrotropin-releasing hormone challenge compared with women without PMDD (10,11). In our own previous study (12), not only did we find greater variance in TSH concentrations in women with PMDD versus women without PMDD, but we also documented greater variance in tri-iodothyronine (T3) resin uptake (a measure reflecting the percent of unsaturated binding sites on proteins), total thyroxine (T4), and free T4 index. One possible explanation for such findings is that in a subgroup of women with PMDD, dysregulation in the HPT axis does play an etiological role.
Evidence that the HPT axis is stress-responsive, especially to traumatic stress (13), combined with evidence that women with PMDD report more sexual abuse (SA) than women without PMDD (14–16), suggests that PMDD women with histories of trauma may represent a subgroup with respect to HPT axis function. This concept stems, in part, from the work by Mason et al. (17,18) and Wang and Mason (19) showing an abnormal HPT axis hormone profile in male combat veterans with posttraumatic stress disorder (PTSD). The most striking observation common to the studies of Mason et al. (17,18) and Wang and Mason (19) is the remarkable elevation in concentrations of total T3 found in men with combat-related PTSD. These investigators have concluded that there is an increase in the conversion of T4 to T3 in their patients (18). Because the HPT axis profile documented in these patients is different from that for hyperthyroidism, or T3 thyrotoxicosis, or primary T4-binding globulin (TBG) elevation, Wang and Mason (19) have suggested that the thyroid hormone pattern observed may have special clinical relevance for combat-related PTSD.
The purpose of the present study was 2-fold. In an investigation of 28 women with PMDD compared with 28 women without PMDD, we found that significantly more women with PMDD had histories of SA relative to women without PMDD (20). Thus, we sought to test the hypothesis that the subgroup of women with PMDD with previous SA would show greater variability in thyroid hormone concentrations relative to both nonabused women with PMDD and women without PMDD with no history of SA. Second, we sought to examine whether the thyroid hormone profile of sexually abused women with PMDD resembled that of men with combat-related PTSD reported by Mason et al. (17,18) and Wang and Mason (19).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
Procedure
Premenstrual Dysphoric Disorder Criteria
The diagnosis of PMDD was based on DSM-III-R (21) criteria, which include a) rating of premenstrual symptoms as moderate or severe, b) moderate to severe symptoms during at least 3 of 6 premenstrual days, c) 5 or more symptoms premenstrually, d) at least 1 moderate to severe emotional symptom, e) evidence that symptoms affected function, and f) criteria a to e met on 2 consecutive menstrual cycles. Functional impairment in PMDD women was confirmed by visual inspection of the daily symptom calendar to ensure that luteal phase symptoms were associated with lifestyle impact.
Women without PMDD met the following criteria: a) not completely asymptomatic during the premenstrual week (to exclude women biased toward nonreporting), b) only mild emotional symptoms premenstrually, c) moderate physical symptoms on fewer than 3 days premenstrually with no severe physical symptoms, and d) criteria a to c met during 2 consecutive menstrual cycles. Inspection of the daily calendar confirmed that the premenstrual symptoms in the women without PMDD were not associated with lifestyle impact.
Assessment of Premenstrual Dysphoric Disorder
The Prospective Record of the Impact and Severity of Menstrual Symptoms (PRISM) calendar (22) was used to assess daily the intensity of emotional and physical symptoms. For each symptom, participants assigned 0 if absent, 1 if mild, 2 if moderate, or 3 if severe. In addition, the PRISM calendar incorporates measures of lifestyle impact together with information on life events and the use of medications that may modify symptomatology. The PRISM calendar was completed daily for 2 or 3 menstrual cycles. To discourage retrospective reporting, participants mailed their calendars weekly.
Structured Clinical Interview
Psychiatric Histories
After assigning subjects to the groups with or without PMDD, but before testing, we administered a Structured Clinical Interview for DSM-III-R Axis I disorders (SCID). A psychiatric research nurse conducted all SCID interviews on days 2 to 13 of the menstrual cycle, and all diagnoses were based on a consensus diagnostic session with a psychiatrist (C.P.). Because of the evidence for an association of PMDD with histories of major depressive disorder (MDD) (4), women with previous histories of psychiatric disorders were not excluded. However, at least 3 months since remission of the last episode of MDD was required before study enrollment (6 months before testing), and for all other Axis I disorders, 3 years was required to have elapsed since full remission. The 28 women with PMDD and 25 women without PMDD did not differ significantly in the number with previous MDD (8 vs. 7, respectively), anxiety disorder (1 vs. 1), substance abuse or dependence (6 vs. 3), or eating disorder (1 vs. 0).
Sexual Abuse Histories
At the end of the SCID interview, we asked subjects about SA histories using a modified version of the structured interview developed by Leserman et al. (23). To meet criteria for SA incidents as an adult, there had to be a clear threat of harm or force (pressure for sexual activity was not sufficient). To meet criteria as a child, the threat of force did not have to be as clearly established if it was implied by the age differential between perpetrator and victim. SA was defined as either of 2 types of sexual experiences: forced sexual touching (including oral sex and vaginal penetration with objects) or intercourse. Subjects were also asked about physical abuse experiences during this interview. We have previously reported on the impact of abuse histories (sexual or physical) on sympathetic and HPA axis measures in this cohort of women with PMDD versus without PMDD (20). In the present study, preliminary analyses revealed no association between physical abuse and HPT axis measures for either the women with PMDD or without PMDD. Thus, physical abuse was not considered further in this report.
Three women without PMDD reported SA. Data from only the 22 women without PMDD with no SA history were included in the statistical analyses. However, the mean HPT axis hormone concentrations for the 3 women without PMDD with previous SA are provided. Thus, for statistical analyses, we grouped women into the following 3 categories: women with PMDD with previous SA (N = 10), women with PMDD with no history of SA (N = 18), and women without PMDD with no previous SA (N = 22). These 3 groups did not differ significantly in age (33.1, 34.4, and 32.3 years, respectively), body mass index (24.8, 26.2, and 23.3), alcohol consumed (0.15, 0.14, and 0.13 drinks/day), or caffeine intake (1.8, 1.6, and 1.4 drinks/day), or number of in cigarette smokers (4, 2, and 3). Among the 10 women with SA and PMDD, 4 had experienced multiple SA incidents, whereas 6 had experienced a single episode. On average, 17 years (range, 3 to 32 years) had elapsed since the last abuse incident. Three women had experienced SA as a child (age, 4 to 12 years), 3 had experienced abuse as an adolescent (age, 14 to 16 years), and 4 had experienced SA as a young adult (age, 18 to 29 years).
Posttraumatic Stress Disorder Symptoms
Because the assessment of PTSD was not included as a standard module in the SCID manual, we assessed PTSD-like symptoms using the supplementary PTSD (PS) scale of the Minnesota Multiphasic Personality Inventory-2 (MMPI-2; 24). The PS scale was administered during the follicular phase of the menstrual cycle. The PS scale contains 60 items pertaining to a broad range of somatic and mental symptoms. The internal consistency and test-retest reliability of this scale in women are high (0.87 and 0.89, respectively). In the current cohort of women, the abused PMDD group (mean score, 51.6 ± 2.6; range, 39 to 71) and nonabused PMDD group (mean score, 55.7 ± 2.1; range, 43 to 73) scored significantly higher (p < .001) than the women without PMDD (mean score, 42.0 ± 1.8; range, 37 to 50), although the 2 PMDD groups did not differ from each other. Only 6 PMDD women total (2 with SA) scored greater than 65 on this scale, indicating clinical levels of distress (24).
Experimental Procedures
We tested each participant twice, once in the follicular phase (days 4 to 9) and once during the luteal phase, 7 to 11 days after home urine testing (Clearplan Easy) had revealed the luteinizing hormone surge that precedes ovulation. Order of testing was counterbalanced within groups. Confirmation of cycle phase was subsequently determined by serum progesterone concentrations. Time of day for testing was held constant for each participant, although it was allowed to vary between subjects. This tactic is consistent with guidelines published by the National Institutes of Health (25) on the successful recruitment and retention of women into clinical research studies. Groups did not differ in the proportion of women tested as a function of time of day (p > .10).1
Approximately an hour after arriving at the laboratory, participants were comfortably seated in a sound-attenuated testing chamber. An intravenous line was established in an arm vein, and a curtain was drawn to prevent the participant from viewing the intravenous apparatus. Subjects rested quietly for 25 minutes before blood was drawn. Then subjects were then exposed to a battery of mental stressors (results reported elsewhere; see 20,26).
Hypothalamic-Pituitary-Thyroid Axis Variables and Assays
Serum concentrations of TSH, total and free T4, total and free T3, reverse T3 (rT3), and TBG were measured in both the follicular and luteal phases of the menstrual cycle. Activation of the HPT axis involves the pituitary release of TSH, which then acts to increase thyroid gland release and production of thyroid hormones. T4 is the major secretory product of the thyroid gland and is the precursor of T3, which is approximately 4 times more metabolically active than T4. Thyroid hormone concentrations are measured as both free (that which is not bound to proteins) and total (bound plus free) fractions. The vast majority of T3 arises from peripheral deiodination of free T4; rT3 represents the metabolically inactive metabolite of free T4 conversion. TBG accounts for approximately 80% of the bound thyroid hormone.
Serum for total T4, free T4, total T3, free T3, and TBG was collected into four 7-ml serum separator tubes at the end of the 25-minute baseline rest period. Assays were run in duplicate using radioimmunoassay (RIA) procedures via commercial kits purchased from Becton-Dickinson. The intra-assay and interassay coefficients of variation were, respectively, 6% and 8% for total T4, 5% and 8% for free T4, 5% and 7% for total T3, 4% and 7% for free T3, and 7% and 9% for TBG. rT3 levels were measured using RIA commercial kits purchased from Serono with intra-assay and interassay coefficients of variation of 3% and 7%, respectively. For each HPT axis measure, all samples (ie, from all groups and both phases) were randomly assigned positions within a single assay, thereby removing any systematic influence that interassay variability might have exerted on results.
Because both estrogen and thyroid hormones (27) increase the plasma concentration of sex hormone-binding globulin (SHBG), serum SHBG was assessed in both the follicular and luteal phases with the use of a commercial enzyme-linked immunosorbent assay kit (Alpco Diagnostics). The specificity of this kit allows no cross-reactivity with thyroid binding globulin. The sensitivity of the direct enzyme-linked immunosorbent assay SHBG is 0.2 nmol/l.
Serum estradiol and progesterone levels were determined via RIA commercial kits purchased from ICN Biomedicals. The specificity of the antiserum for estradiol is high, showing only 0.01% to 1.45% cross-reactivity with other steroid hormones with the exception of estrone, for which there is 6% cross-reactivity. The specificity of the antiserum for progesterone is similarly high, showing only 0.01% to 2.5% cross-reactivity with other steroid compounds. Progesterone concentrations greater than 3.0 ng/mL during luteal phase testing were used to confirm that all women were tested during an ovulatory cycle. Because previous detection of the midcycle luteinizing hormone surge accompanying ovulation was a prerequisite for scheduling luteal testing, no participant was tested during an anovulatory cycle.
Data Analysis
First we examined whether histories of SA in women with PMDD were associated with a greater degree of variability in the distribution of each HPT axis measure relative to nonabused women without PMDD. Thus, for each thyroid measure and in each cycle phase, we created an F statistic based on the ratio of the variance in that measure for the women with PMDD and SA to the variance in that same measure for the women without PMDD. An F statistic was similarly calculated using the ratio of women with PMDD with no SA to women without PMDD. We tested whether the variance in each PMDD group was different from the variance in women without PMDD using a 2-tailed F test (10,12).
Next we investigated whether women with PMDD with histories of SA represent a distinct subgroup of women with PMDD with respect to HPT axis variables. Thus, for each HPT axis measure, a 2 (cycle phase) x 3 (group) repeated measures analysis of variance was conducted, with cycle phase as the repeated factor.
Finally, based on the evidence for significant relationships between total T3 concentrations and mental symptoms in male veterans with combat-related PTSD (19), we also examined whether total T3 correlated with premenstrual symptom severity in the total group of women with PMDD. To minimize Type I errors, only the premenstrual symptom ratings for depression, anxiety, irritability, anger, restlessness, and insomnia were selected because they are most similar to the symptoms reported by Wang and Mason (19) to relate to total T3 levels. Moreover, to examine the degree to which total T3 served as an independent predictor of premenstrual symptoms, for each premenstrual symptom (based on the mean prospective symptom severity score from the PRISM calendars), stepwise multiple regression techniques were used. For the models predicting each of the premenstrual symptoms listed, luteal phase concentration of total T3, history of MDD, history of SA, and the PTSD symptom score from the MMPI were always included as potential predictor variables.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
|
|
Thyroid-Stimulating Hormone
Although in both cycle phases, mean TSH concentrations were at least 75% higher in the sexually abused women with PMDD compared with both nonabused groups (Table 1), this difference did not reach conventional levels of statistical significance, probably because of the greater variance in TSH concentrations in the abused women with PMDD, as noted.
Thyroxine-Binding Globulin
Regardless of menstrual phase, sexually abused women with PMDD exhibited greater TBG concentrations relative to both nonabused groups of women (F[2,47] = 3.3; p = .04; Table 1).
Reverse Tri-iodothyronine
There were no significant effects involving rT3 concentrations.
Evaluation of Tri-iodothyronine Conversion
A strategy to determine alterations in conversion of T4 to T3 is to compute ratios of T3 to T4 (18).
Free Tri-iodothyronine/Free Thyroxine Ratio
As depicted in Figure 2, a significant phase by group interaction was found (F[2,48] = 3.2; p = .05), because only in the luteal phase did sexually abused women with PMDD show a greater ratio than both of the nonabused groups.
|
|
Free Tri-iodothyronine/Total Tri-iodothyronine and Free Thyroxine/Total Thyroxine Ratios
Regardless of cycle phase, sexually abused women with PMDD had lower free T3/total T3 ratios relative to nonabused women with PMDD and women without PMDD (2.3 vs. 2.6 and 2.7, respectively; F[2,47] = 3.8; p = .03) and lower free T4/total T4 ratios relative to nonabused women with PMDD and women without PMDD (1.8 vs. 2.0 and 2.1, respectively; F[2,47] = 5.3; p = .008).
Mean Thyroid Hormone Concentrations in Women Without Premenstrual Dysphoric Disorder With Sexual Abuse
The 3 women without PMDD with previous SA did not differ appreciably from either the women without PMDD with no abuse or the women with PMDD with no abuse in follicular or luteal phase concentrations of total T3 (97.1 ± 9.3 and 100.5 ± 8.0 ng/dl, respectively), free T3 (0.27 and 0.24 ng/dl, respectively), total T4 (7.4 ± 1.8 and 6.1 ± 0.7 µg/dl, respectively), free T4 (1.4 ± 0.01 and 1.2 ± 0.2 ng/dl, respectively), or TBG (23.5 ± 3.0 and 26.4 ± 6.5 µg/ml, respectively). In contrast with the higher TSH concentrations noted in the sexually abused women with PMDD, the 3 sexually abused women without PMDD had follicular and luteal concentrations that were 23% to 42% lower than the other women without PMMD (1.3 ± 0.6 and 1.2 ± 0.6 uU/ml, respectively).
Relationship of Total Tri-iodothyronine Concentration to Premenstrual Symptoms in Women With Premenstrual Dysphoric Disorder
For women with PMDD as a whole, luteal phase concentrations of total T3 were positively correlated with premenstrual depression ratings (r = 0.43; p < .05), anger ratings (r = 0.39; p < .05), and anxiety ratings (r = 0.39; p < .05), but not with ratings of irritability (r = 0.33), restlessness (r = 0.10), or insomnia (r = 0.09).
Table 2 summarizes the multiple stepwise regression results relating selected predictor variables to premenstrual symptoms. Although a history of MDD most consistently predicted premenstrual symptom severity, luteal phase T3 concentrations also independently predicted greater premenstrual depression, anxiety, and anger ratings in women with PMDD as a whole. Neither a history of SA nor the PS score accounted for significant variance in any premenstrual symptom.
|
Sex Hormone Concentrations
As expected, luteal phase concentrations of both estradiol and progesterone were significantly higher than follicular phase concentrations for all women (mean estradiol, 55.0 ± 6.4 vs. 36.8 ± 6.2 pg/ml; mean progesterone, 16.2 ± 2.5 vs. 0.50 ± 0.12 ng/ml; F values [1,47] = 12.4 and 116.6, respectively; p values < .001). There were no significant main effects or interactions involving group for either estradiol or progesterone.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
The most parsimonious interpretation of the endocrine data is that women with PMDD with SA histories experience increased peripheral conversion of T4 to T3. This distortion of peripheral events may or may not reflect parallel changes in central events. However, it appears unlikely that central events would remain entirely unchanged while the brain receives a distinctly altered proportion of T3 and T4. It must be noted, however, that not all endocrine findings are supportive of the notion of increased peripheral formation of T3. If increased conversion occurs, one might expect lower free T4 concentrations combined with elevations in both free T3 and total T3 concentrations. We did not find the changes in free T4 or free T3, and their absence remains unexplained. We did, however, find that the relative proportion of both free T3 and total T3 concentrations to free T4 concentrations was higher in the sexually abused women with PMDD, a pattern consistent with increased conversion of T4 to T3 (18). Thus, increased conversion of T4 to T3 in the periphery, and perhaps in the brain, remains the most likely formulation of the chief findings. Whatever the mechanism, there appears to be something systematically different in the sexually abused women with PMDD shifting from a T4 economy toward a T3 economy. Despite the finding of elevated T3 concentrations in this subset of women with PMDD, they gave no other evidence of a hyperthyroid state. SHBG, a sensitive measure of thyroid state (27), was not increased in the women with PMDD with SA. Thus, it appears that increased binding by TBG offsets the enhanced formation of T3, thereby serving as a compensatory mechanism to maintain a euthyroid state (18).
The case for the present HPT axis findings having behavioral significance is bolstered by the remarkable similarity of our findings in women with PMDD with SA histories and those reported in men with combat-related PTSD (17–19). Specifically, like the patients with combat-related PTSD, the abused women with PMDD displayed elevated concentrations of total T3 and TBG, but normal concentrations of free T4 and rT3. Also, although the results are not statistically significant, both groups displayed TSH concentrations 50% to 75% higher than their control counterparts. Sexually abused women with PMDD and men with combat-related PTSD were also similar in that they exhibited elevated free T3/free T4 and total T3/free T4 ratios, and lower free T3/total T3 and free T4/total T4 ratios. Moreover, for both sexually abused women with PMDD and men with combat-related PTSD, total T3 concentrations were positively related to symptom severity. The 2 groups differed only in that greater total T4 concentrations were not observed in the women with PMDD, a result found in some (17) but not all patients with combat-related PTSD (19).
It appears remarkable that sexually abused women with PMDD and men with combat-related PTSD show highly similar and unusual changes in the HPT axis. Our data suggest that SA alone is not associated with these changes; the study of Wang and Mason (19) demonstrates that combat experience alone is insufficient. It appears that in women, PMDD must be present for SA to precipitate these HPT axis changes; in men, PTSD must be present for the combat experience to produce the changes. Although we did not have the statistical power to address this issue in the present study, we did examine mean concentrations of thyroid hormones in the 3 women without PMDD with previous SA, and, with the exception of TSH concentrations that were 23% to 42% lower than those of other women without PMDD, all other HPT axis variables were remarkably similar to those of nonabused women. This preliminary observation is consistent with the results of a study by De Bellis et al. (28) in which no differences were found between adolescent girls with severe or repeated SA and age-matched controls in measures of T3, T4, free T4, or TSH. These results are also consistent with the findings from a large-scale community-based epidemiologic study of older men and women that found no association between SA histories and thyroid hormone concentrations in women (29).
It cannot be proven that the HPT axis changes are causal of symptoms in a subset of women with PMDD or in a subset of men with PTSD, but it appears likely. The symptoms in both conditions resemble a state of hyperarousal, which might be expected from chronic activation of a system known in part for its ergotropic properties. Although the sexually abused women with PMDD are not hyperthyroid, they do show evidence of hyperarousal. For example, we observed that in both cycle phases, abused women with PMDD had TSH levels at least 75% higher than those of all other women. Although this difference was not statistically significant, these differences in TSH may be biologically important and indicate that the thyroid gland is being driven excessively (30). In a post hoc fashion, we also examined other indices of arousal in this cohort of women, finding that, relative to women without PMDD, women with PMDD with SA exhibited significantly greater resting heart rates in both the follicular (78 vs. 69 bpm, respectively) and luteal phases (76 vs. 70 bpm), whereas nonabused women with PMDD fell midway between (73 to 75 bpm). Although women with PMDD with SA also had higher resting norepinephrine concentrations than women without PMDD in both the follicular (208 vs. 186 pg/ml, respectively) and luteal phases (234 vs. 192 pg/ml), these differences were not statistically significant.
The elevated total T3 seen in the sexually abused women with PMDD may be part of an arousal response, because the thyroid gland itself can, under emergency situations, meet the body’s metabolic needs by direct secretion of T3 (31). Although normally, the dynamics of the HPT axis are such that negative feedback mechanisms would restore values to baseline levels, the possibility exists that increased total T3 levels could be sustained in the face of chronic or severe stress (30). At the same time, because T3 can easily enter the brain and promote noradrenergic neurotransmission (32), increased T3 may perpetuate further arousal. Thus, the persistent elevation of total T3 seen in the women with PMDD with histories of SA and also in the men with combat-related PTSD may represent an adaptation gone awry if the thyroid system continues to prepare the organism for a severe stressor that no longer exists but somehow remains encoded in the brain (30).
Despite the striking HPT axis similarities seen between sexually abused women with PMDD and men with combat-related PTSD, these similarities do not necessarily imply that women with PMDD with previous SA suffer from PTSD. Although the women with PMDD as a whole scored higher on the PTSD scale of the MMPI than the women without PMDD, the abused women with PMDD did not differ from the nonabused women with PMDD in the PTSD-like symptoms. Additional evidence that the women with PMDD with previous SA did not suffer from current PTSD comes from the 2-month to 3-month prospective symptom records that were used to exclude women with chronic symptoms. Certainly 1 limitation of the present study involves the lack of a structured interview to assess not only previous PTSD but also current PTSD, because the PTSD scale of the MMPI is not specific to PTSD but will be elevated in any form of psychopathology (24). Despite this limitation, however, there is little evidence that current PTSD contributed to abuse-related differences in HPT axis measures in the present cohort. A second limitation to our study concerns our inability to perform formal analyses investigating the impact of SA on HPT axis variables in women without PMDD women. Although the percentage of women without PMDD women in our study who reported SA (12%) is consistent with national statistics from nonpatient samples (33), suggesting that our study did not suffer from a sampling bias, the number of subjects was too small to allow valid analyses.
This study is among the first to assess menstrual cycle influence on thyroid hormone concentrations during confirmed ovulatory cycles. Although the results of previous studies have been mixed (3,12,34,35), perhaps because of small sample sizes, the present findings suggest that in medically healthy, euthyroid women, the menstrual cycle exerts an influence on thyroid hormone concentrations, specifically total T4, free T4, and free T3, with luteal phase concentrations lower than follicular phase concentrations. Future studies investigating the clinical significance, if any, of these subtle changes in thyroid hormone concentrations during the menstrual cycle in healthy women would be of interest.
In conclusion, the results of our study suggest that women with PMDD with histories of SA are distinct from other women with PMDD and from healthy women without PMDD with respect to both variability and mean hormone concentrations of HPT axis variables. Specifically, we obtained evidence for markedly increased conversion of T4 to T3 in sexually abused women with PMDD, which is apparently compensated for by increased peripheral binding of T3. We speculate that the elevated concentrations of total T3 in women with PMDD with previous SA, although not reflecting a hyperthyroid state, may reflect a hyperarousal state. Although the functional significance of increased conversion of T4 to T3 is uncertain, if the process occurs in brain and in periphery, it may well contribute to symptoms. Thus, it may be justified to explore the use of a therapeutic agent such as propranolol in PMDD because it crosses the blood–brain barrier and is effective in reducing T3 concentrations (36). Indeed, a recent placebo-controlled study found that propranolol significantly reduced premenstrual symptoms, including depression, in women with severe premenstrual syndrome (37).
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
The authors are grateful to Robert D. Utiger, MD, for his comments on an earlier version of this manuscript. The authors are also grateful to Unipath Diagnostics for their generous donation of Clearplan Easy ovulation predictor kits, to Catherine Stanwyck and Sara Benjamin for their role as project coordinators, to Astrid Ertola, MA, for her assistance with data analysis, to Cheryl Walker for conducting all hormone assays, and to Dot Faulkner for her assistance with manuscript preparation.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
Received for publication July 3, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESACKNOWLEDGMENTSREFERENCES |
|---|
,5-triiodothyronine levels in cerebrospinal fluid in patients with endogenous depression. Acta Endocrinol (Copenh) 1991; 124: 166–72.This article has been cited by other articles:
 |
|
 |
|
  X. Bai, J. Li, L. Zhou, and X. Li Influence of the menstrual cycle on nonlinear properties of heart rate variability in young women Am J Physiol Heart Circ Physiol, August 1, 2009; 297(2): H765 - H774. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
