Reduced platelet tritium-labeled imipramine binding sites in women with premenstrual sy~drome John F. Steege, MD: Anna L. Stout, PhD,.· b. David L. Knight, BS,c and Charles B. Nemeroff, MD, PhDc, d C
Durham, North Carolina OBJECTIVE: We studied the possible role of serotonergic systems in the cause of premenstrual affective symptoms. STUDY DESIGN: The binding of tritium-labeled imipramine to platelets is thought to parallel central nervous system binding and to indicate serotonergic activity. We measured platelet tritium-labeled imipramine binding sites in the fOllicular and luteal phases in 12 controls and in 9 women with well-documented late luteal phase dysphoric disorder. In statistical analyses we used repeated measures analysis of variance, with Student-Newman-Keuls and Duncan's one-tailed t tests, and Pearson's r. RESULTS: The values of subjects with late luteal phase dysphoric disorder were lower than those of controls (F [1,39J = 5.13, P = 0.03). Both follicular and luteal phase levels were lower in subjects with late luteal phase dysphoric disorder but reached statistical significance only in the follicular phase. CONCLUSION: Lower platelet tritium-labeled imipramine binding in women with late luteal phase dysphoric disorder supports the hypothesis that alteration of central serotonergic systems may contribute to premenstrual dysphoric symptoms. (AM J OBSTET GVNECOL 1992;167:168-72.)
Key words: Late luteal phase dysphoric disorder, premenstrual syndrome, depression, tritium-labeled imiprimine binding The premenstrual occurrence of a variety of physical and emotional symptoms has been termed premenstrual syndrome (PMS) or, when affective symptoms are emphasized, late luteal phase dysphoric disorder.! Although epidemiologic and clinical evidence demonstrate significant psychiatric comorbidity in PMS and late luteal phase dysphoric disorder,2 studies of putative biochemical markers of depression such as the dexamethasone suppression test 3 and the thyrotropin-releasing hormone stimulation test4 have not revealed consistent deviations from normal values in women with PMS. Platelet tritium-labeled imiprimine binding sites, which are believed to represent receptor sites that label the presynaptic serotonin transporter on the presynaptic nerve terminal, are lower in number in drug-free depressed patients during acute depressive episodes than they are in controls.' There is some evidence that
From the Department of Obstetrics and Gynecology," the Division of ,Hedical Psychology,! the Department of Psychiatry,' and the DepartlIlent of Pharmacology,' Duke University Medical Cenler. Supported in part by the John D. and Catherine T. MacArthur Foundation Health Research Network 4: Determinants and Consequences £if Health-Promoting and Health-Damaging Behavior and by the National Institute for Mental Health grant MH-40159. Received for publication March 26, 1991; revised January 10, 1992; accepted January 23,1992. Reprint requests: John F. Steege, MD, Box 3263, Duke University Medical Center, Durham, NC 27710.
6/1/36686 168
the number of these sites returns to normal several months after clinical remission of depression, response to psychotropic medication,6 or electroconvulsive therapy.7 This platelet binding site has, therefore, been described as a state marker." Because the affective state changes in a predictable fashion in late luteal phase dysphoric disorder, this disorder provides a model for the investigation of platelet [3H]imipramine binding sites and their association with depression. This study investigated the number and affinity of platelet [3H]imipramine binding sites across the menstrual cycle in women without clinical affective disorder who fulfilled criteria for late luteal phase dysphoric disorder.
Material and methods Twenty-three women aged 24 to 42 (mean 32.2 ± 6.0 SD) were recruited by advertisement and word of mouth to participate in a study of cardiovascular and neuroendocrine reactivity across the menstrual cycle. Subjects with late luteal phase dysphoric disorder and controls were of equivalent parity (mean ± SD 0.63 ± 0.98 for all subjects, range 0 to 2), and all were within ± 15 %of ideal body weight. Women with late luteal phase dysphoric disorder reported symptoms for 3 to 11 years (mean ± SD 5.7 ± 2.8) and participated in the study primarily to obtain free evaluation for their premenstrual symptoms. Subjects gave informed consent under a protocol approved by the
Volume 167 Number I
institutional review board and were paid for their participation. All subjects completed the Premenstrual Assessment Form 9 as a screening instrument and then completed prospective symptom charting over two consecutive menstrual cycles. Twelve women were designated asymptomatic controls, and nine met the DSM-IlIR criteria l for late luteal phase dysphoric disorder. A sym ptom was scored as present and cyclic if it was rated at s2 on a six-point ordinal scale throughout the postmenstrual follicular phase of the menstrual cycle and was present at moderate or severe levels (a score of 2:4) during at least 4 days of the premenstrual week. During baseline cycles, ovulation was detected by basal body temperature recordings. During cycles in which laboratory sessions took place, ovulation was documented by the Q-test (Becton Dickinson, Franklin Lakes, N.J.). A total of 142 subjects were screened. All subjects met the following additional inclusion criteria: a normal Minnesota Multiphasic Personality Inventory profile, which was completed during the follicular phase; absence of Axis I diagnosis on the Structured Clinical Interview for the DSM-III-RlO; a normal follicular phase Beck Depression Inventory score ll ; the absence of serious medical illness such as thyroid dysfunction; ingestion of no psychotropic, hormonal, or other medications; and a menstrual cycle length of 24 to 32 days. All subjects had normal complete physical and gynecologic examinations and normal laboratory studies, including complete blood counts, urinalysis, and thyroid-stimulating hormone and free thyroxine evaluations. Subjects completed two laboratory sessions in the follicular and luteal phases of the menstrual cycle in counterbalanced order. Relative to the day of ovulation, cycle days of the laboratory studies were similar for control subjects and those with late luteal phase dysphoric disorder (follicular: control - 5.1 ± 1.8 vs late luteal phase dysphoric disorder -6.2 ± 2.2, NS; Luteal: control 9.6 :±: 2.4 vs late luteal phase dysphoric disorder 8.7 ± 1.5, NS). All sessions began between 1:30 and 3 PM. Each session lasted approximately 75 minutes, during which subjects were exposed to various stressors. At the beginning and end of each session, 35 ml of blood was drawn into evacuated glass tubes containing 2 ml of acid citrate-dextrose solution, which served as an anticoagulant. Platelet-rich plasma was obtained by centrifugation at 100g for 2 minutes; centrifugation was repeated three times. The platelet-rich plasma was then centrifuged (16,000g for 10 minutes at 4° C), and the resulting platelet-rich pellet was suspended in 50 mmol of Tris buffer (pH 7.5 at 4° C) containing 150 mmol of sodium chloride and 20 mmol of ethylenediaminetetraacetic acid by means of a Polytron homogenization medium probe (Polytron, Brinkmann In-
[3Hjlmipramine binding sites in PMS
169
struments, Westbury, N.Y.) at setting of 4 for 10 to 15 seconds. After centrifugation at 39,000g for 10 minutes, membranes were prepared by hypotonic lysis in 5 mmol Tris (pH 7.5), containing 5 mmol ethylenediam'inetetraacetic acid and were then homogenized and centrifuged at 39,000g for 10 minutes. The supernatant was discarded, and the pellet was washed by resuspension in 70 mmol Tris (pH 7.5) and was centrifuged again at 39,000g for 10 minutes. The washed pellet was finally resuspended in the assay buffer, which consisted of 50 mmol Tris (pH 7.5 at 4° C) containing 120 mmol of sodium chloride and 5 mmol of potassium chloride. Aliquots were taken for protein deteJ;mination. Platelets were then frozen at 70° C until they were used for the binding assay. Platelet [3H]imipramine binding was conducted with a modification of the procedure of Nemeroff et a!.l2 In each sample, total binding was assayed at seven different concentrations (0.25 nmol to 8 nmol) of pH]imipramine (specific activity 45.4 Ciimmol); the dissociation constant (Kd) is approximately 1 nmo!. Each assay tube (all concentrations were measured in triplicate) contained approximately 100 f.Lg of platelet protein, and the final assay volume was 250 mm'. Tubes were incubated with the radiolabel for 60 minutes on ice. The reaction was terminated by addition of 5 ml of ice-cold assay buffer, and bound radioactivity was collected by filtration under vacuum over glass fiber filters presoaked for 3 hours in buffer. The filters were then washed three times with 5 ml of ice-cold assay buffer, and bound radioactivity was collected by filtration under vacuum over glass fiber filters presoaked for 3 hours in buffer. The filters were then washed three times with 5 ml of ice-cold assay buffer. Nonspecific binding was defined as binding persisting in the presence of 100 f.Lmol of desipramine. Specific binding represented 80% of total binding, which was less than 5% of the total counts added. Scatchard analysis was used to obtain the maximum binding and Kd for ['H]imipramine. Linear regression of Scatchard plots was used to determine the correlation coefficient r. Assays with r < 0.90 were excluded. The interassay coefficients of variation were 10% for maximum binding and 5% for Kd. Comparison of the first and second assays of imipramine binding sites in each laboratory session was performed by paired Student t test. The Student-Newman-Keuls test was used to compare pooled [3H]imipramine level results from controls with those from subjects with late luteal phase dysphoric disorder. Repeated measures analysis of variance was used to measure group (control vs late luteal phase dysphoric disorder) x time (follicular vs luteal) interactions. Analyses of variance with Duncan's one-tailed t tests were used to compare results from both groups within one
170 Steege et al.
July 1992 Am J Obstet Gynecol
c:
1600
'(j)
.- 1400 0
.... a..
-
1200
E "0 E
800
0
' 0)
00
•
1000
600 >< ro E 400
III
o
o
200
o
o
Follicular
Luteal PMS (N=9)
Follicular Luteal Controls (N= 12)
Fig. 1. ['H]Imipramine binding sites in controls and women with premenstrual syndrome by menstrual cycle phase. Bars, means; shaded areas, ± I SEM. Only paired data points were used for statistical analysis.
cycle phase. Pearson's r was used to measure the correlation between luteal platelet binding site density and Beck Depression Inventory scores from baseline cycles. The p < 0.05 level of significance was used. Results
Control subjects had, as expected, normal Beck Depression Inventory scores in the follicular (1.0 ± 1.2, mean ± SD) and luteal (2.4 ± 3.2, mean:±: SD) phases of a baseline charting cycle. PMS subjects were normal in the follicular phase but were significantly depressed in the luteal phase; this was demonstrated by increased Beck Depression Inventory scores (2.1 ± 2.90 vs 17.1 ± 10.0, P< 0.001). Four of the nine PMS subjects had luteal phase Beck scores in the clinically depressed range (> 16). The number of [3H]imipramine binding sites in samples obtained before and after the laboratory sessions did not differ significantly. This attests to the stability of platelet binding site characteristics over short time intervals, i.e., test-retest reliability. The mean value for each day was therefore used for subsequent analyses. For the analyses of variance, one subject was dropped from each group as a result of our inability to obtain samples in both menstrual cycle phases. When follicular and luteal phase values within each subject group were combined, platelet [3H]imipramine binding site density for PMS women was significantly lower than that of controls (800.2 ± 237.8 vs 981.4 ± 258.5 fmol/gm protein [F(l,39) = 5.13, P = 0.03]). Repeated measures analysis of variance revealed a time effect at the trend level (F[l,17] = 2.57, P = 0.12), but no time X group effect. When platelet binding site number by cycle phase across subject groups was considered, (Fig. 1) women
with PMS had lower maximum binding values than did controls in the follicular phase (744.8 ± 247.6 vs 963.0 ± 240.9 fmoll gm protein) by analysis of variance (F[I,20] = 4.12, P = 0.057) and Duncan's one-tailed t test (p < 0.05), but the two groups were not different in the luteal phase (862.4 ± 225.4 vs 1006.3 ± 284.2 fmol/gm protein; F[l,18], p = 0.27, and p > 0.05, respectively). The increase in the number of [3H]imipramine binding sites in platelets in PMS women from the follicular to the luteal phase was statistically significant (750.5 ± 264.1 vs 862.4 ± 225.4 fmol/gm protein, p < 0.05). Luteal phase Beck Depression Inventory scores were not correlated with either follicular or luteal platelet [3H]imipramine binding site density in PMS subjects. Comment
This study documents lower numbers of platelet [3H]imipramine binding sites in women with well-documented PMS who, in the follicular phase of the menstrual cycle, were clearly free of depression by stringent clinical and psychometric criteria. In results consistent with those of previous reports,'" 14 asymptomatic women showed no influence of menstrual cycle phase on platelet [3H]imipramine binding. Subjects in the present study with late luteal phase dysphoric disorder had lower values in the follicular as opposed to the luteal phase. Rojansky et al. 15 also noted differences in [3H]imipramine binding in controls as opposed to women with "premenstrual changes," but they noted lower maximum binding values in early versus late luteal phase samples from symptomatic women. However, they observed a significant cycle phase effect on maximum binding in controls, a change that was not observed in this study or in previous reports. 13.14 Their
[3Hjlmipramine binding sites in PMS
Volume 167 Number 1
subject population included some women with a lifetime diagnosis of depression, but this study excluded those women. The large degree of variability in the assay results and the limited number of times of sampling may further account for these apparent discrepanCIes. The density of [3H]imipramine binding sites has been reported to be reduced in depressed patients in the majority of studies 8 and in scattered reports of a few possibly related conditions such as anorexia,16 agoraphobia,17 and obsessive-compulsive disorder. 16 The latter two disorders are among those overrepresented in an epidemiologic study of psychiatric disease and premenstrual symptoms (Duke Epidemiologic Catchment Area Study. Unpublished data). The number of platelet [3H]imipramine binding sites was about 18% lower in women with PMS than in controls; this was a smaller difference than the average 35% reduction seen in studies of depressed patients, which suggests that PMS may be a variant of depression modulated by the menstrual cycle. The observation of a strong trend toward reduced binding sites to the serotonin transporter in women with PMS during their asymptomatic times is of interest. Platelets have a mean life-span of 21 days. Because [3H]imipramine binding sites are cell membrane characteristics of the anucleate platelet, they may be seen as evidence of synthetic events that occurred weeks earlier in the megakaryocyte. Low-follicular-phase density of the binding sites and the emotional state of the preceding luteal phase may both be the results of luteal phase biosynthetic events. Although previous studies have shown that platelet [3H]imipramine binding sites return to normal after the clinical resolution of depression, most researchers have measured this change 1 to 2 months after clinical improvement has occurred. 7 . 18. 19 PMS may indeed provide a viable model for the study of affective state change that may resemble an acute depressive episode. Sex hormones are well known to affect serotonergic systems. Pharmacologic doses of estrogen in rats· o and physiologic replacement doses (10 mg of estradiol valerate monthly) in surgically menopausal women 21 increase the number of these binding sites approximately 20%. The time course of these changes is uncertain. In the latter study,21 women given estrogen showed improved mood com pared with that of their preoperative state, but controls failed to show any decline in mood while deprived of estrogen replacement. The precise relationship of estrogen levels to mood thus remains enigmatic with respect to both premenstrual and postmenopausal states. Previous studies of serotonergic mechanisms in PMS have yielded contradictory findings. Whole blood serotonin levels··' 23 and 5-hydroxytryptamine 24 uptake are reported to be lower in premenstrual women with
171
premenstrual syndrome, and the maximum volume of platelet serotonin uptake is reportedly decreased premenstrually as well. 22 Malmgren et al.,25 in a study limited by the absence of daily symptom ratings, compared women who had PMS with control subjects and found no effect of subject group or cycle phase on platelet serotonin uptake. The lack of correlation of a state measure of depression (Beck Depression Inventory score) with the density of binding sites is consistent with reports of depressed patients in which psychometric measures of mood failed to correlate with this measure of presynaptic serotonergic function. 12 The high level of variance in binding site data in general makes it difficult to show such associations without studying large numbers of subjects and diminishes the potential clinical diagnostic value of this laboratory measure. This study provides impetus for the further investigation of the dynamics of this presynaptic serotonergic binding site as a marker of affective state across the menstrual cycle and the relationship between premenstrual emotional changes and more continuous depressive disorders. REFERENCES
I. Diagnostic and statistical manual of mental disorders. 3rd ed, revised. Washington: American Psychiatric Association,1987. 2. Halbreich U, Endicott J. Relationship of dysphoric premenstrual changes to depressive disorders. Acta Psychiatr Scand 1985;71:331. 3. Haskett R, Steiner M, Carroll BS. Dexamethasone suppression tests and the menstrual cycle. J Affective Disord 1984;6:191. 4. Roy-Byrne PP, Rubinow DR, Hoban MC, et al. TSH and prolactin responses to TRH in patients with premenstrual syndrome. Am J Psychiatry 1987; 144:480. 5. Lewis DA, McChesney C. Tritiated imipramine binding distinguishes among subtypes of depression. Arch Gen Psychiatry 1985;42:485-8. 6. Braddock LE, Cowen PJ, ElliottJM, Fraser S, Stump K. Changes in the binding to platelets of ['HJ-imipramine and fHJ-yohimbine in normal subjects taking amitriptyline. Neuropharmacology 1984;23:285-6. 7. Langer SZ, Sechter D, Loo H, Raisman R, Zarifian E. Electroconvulsive shock therapy and maximum binding of platelet ['Hj-imipramine in depression. Arch Gen Psychiatry 1986;43:949-52. 8. Langer SZ, Galzin AM, Poirier MF, Loo H, Sechter D, Zarifian E. Association of ['HJ-paroxetine binding with the 5HT transporter in brain and platelets: relevance to studies in depression. J Recept Res 1987;7:499-521. 9. Halbreich U, Endicott J, Schacht S, Nee J. The diversity of premenstrual changes as reflected and the Premenstrual Assessment Form. Acta Psychiatr Scand 1982; 65:46-65. 10. Spitzer RL, WilliamsJBW, Gibbon M. Instruction manual for the structured clinical interview for the DSM-IlI-R. April, 1987 revision. New York: Biometrics Research Departments, New York State Psychiatric Institute, 1987. 11. Beck AT, Rush AJ, Shaw BF, et al. Cognitive therapy of depression. New York: Guilford Press, 1979:398-9. 12. NemeroffCB, Knight DL, Krishnan RR, Slotkin TA, Bissette G, Melville ML, Blazer DG. Marked reduction in the number of platelet tritiated imipramine binding sites in
Steege et al.
13.
14. 15.
16.
17. 18.
19.
geriatric depression. Arch Gen Psychiatry 1988;45:91923. Poirier MF, Benkelfat C, Galzin AM, Langer SZ. Platelet 3H-imipramine binding and steroid hormones serum concentrations during the menstrual cycle. Psychopharmacology (Berl) 1986;88:86-9. Stowell LI, McIntosh Cj, Cooke R, Ellis PM. Adrenoreceptor and imipramine receptor binding during the menstrual cycle. Acta Psychiatr Scand 1988;78:366-8. Rojansky N, Halbreich U, Zanden K, Barkai A, Goldstein S. Imipramine receptor binding and serotonin uptake in platelets of women with premenstrual changes. Gynecol Obstet Invest 1991 ;31: 146-52. Weizman R, Carmi M, Tyano S, Rehavi M. Reduced 3H_ imipramine binding but unaltered 3H-serotonin uptake in platelets of adolescent enuretics. Psychiatry Res 1986;19:37-42. Lewis DA, Noyes R, Coryell W, Clancy J. Tritiated imipramine binding to platelets is decreased in patients with agoraphobia. Psychiatry Res 1985;16:1-9. Suranyi-Cadotte BE, Wood PL, Nouir NPV, Schwartz G. Normalization of platelets (3H) imipramine binding in depressed patients during remission. Eur J Pharmacol 1982;85:357 -8. Suranyi-Cadotte BE, Quirion R, McQuade P, Nair NPV,
July 1992 Am J ObstetGynecol
Schwartz G, Mosticyan S, Wood PL. Platelet 'H-imipramine binding: a state-dependent marker in depression. Prog Neuropsychopharmacol Bioi Psychiatry 1984;8:73741.
20. Rehavi M, Sepcuti H, Weizman A. Up-regulation of imipramine binding and serotonin uptake by estradiol in female rat brain. Brain Res 1987;410:135-9. 21. Sherwin BB, Suranyi-Cadotte BE. Up-regulatory effect of estrogen on platelet 3H-imipramine binding sites in surgically menopausal women. Bioi Psychiatry 1990;28;33948. 22. Taylor DL, Mathew Rj, Mo BT, Weineman ML. Serotonin levels and platelet uptake during premenstrual tension. Neuropsychobiology 1984; 12: 16-18. 23. Rapkin Aj, Edelmuth E, Chang L, Reading AE, McGuire MT, Su TP. Whole blood serotonin in premenstrual syndrome. Obstet Gynecol 1987;70:533-7. 24. Ashby CR, Carr LA, Cook CL, Steptoe MM, Franks DD. Alteration of platelet serotonergic mechanisms and monoamine oxidase activity in premenstrual syndrome. BioI Psychiatry 1988;24:225-33. 25. Malmgren R, Collins A, Nilsson CG. Platelet serotonin uptake and effects of vitamin B6 treatment in premenstrual tension. Neuropsychobiology 1987;18:83-8.
The development of ultradian rhythms in the human fetus Mary Pillai, MB, David K. James, MD, and Michael Parker, MB Bristol, Canada OBJECTIVE: Our objective was to determine the normal sequence of neurobehavioral development in the human fetus between 14 weeks' gestation and delivery. STUDY DESIGN: The study was performed by longitudinal ultrasonographic observation of 45 low-risk singleton fetuses. RESULTS: During the first half of the midtrimester there was a high rate of spontaneous movement that appeared randomly distributed. By the end of that trimester an increase in the duration of intervals of quiescence resulted in activity appearing cyclically distributed, with the duration of quiet cycles progressively increasing to term. Fetal mouthing and breathing were linked with cyclic behavior from the time of their emergence. Fetal heart rate pattern A could be recognized from around 32 weeks, due to a reduction in baseline variability in quiet cycles after 30 weeks, whereas pattern B could be recognized several weeks earlier. From the time cyclic behavior and heart rate patterns could be recognized, intervals of coincidence of the fetal behavioral state variables increased progressively with advancing gestation. (AM J OBSTET GVNECOL 1992;167:172-7.)
Key words: Fetal behavior, ultradian rhythms From the University Department of Obstetrics and Gynaecology, Bristol Maternity Hospital. Supported by Action Research for the Crippled Child. Received for publication August 15, 1991; revised Decembel' 20, 1991; accepted December 30, 1991. Reprint requests: Mary Pillai, Women's Hospital, 735 Notre Dame Ave., Winnipeg, Manitoba, Canada R3E OL8. 611 /35916 172
The first half of pregnancy is characterized by rapid emergence of a wide variety of fetal movements of relatively high frequency but little neurologic organization." 2 During the second half of pregnancy these movements become organized. Diurnal rhythms are not demonstrable at 13 weeks but have been reported
Durham, North Carolina OBJECTIVE: We studied the possible role of serotonergic systems in the cause of premenstrual affective symptoms. STUDY DESIGN: The binding of tritium-labeled imipramine to platelets is thought to parallel central nervous system binding and to indicate serotonergic activity. We measured platelet tritium-labeled imipramine binding sites in the fOllicular and luteal phases in 12 controls and in 9 women with well-documented late luteal phase dysphoric disorder. In statistical analyses we used repeated measures analysis of variance, with Student-Newman-Keuls and Duncan's one-tailed t tests, and Pearson's r. RESULTS: The values of subjects with late luteal phase dysphoric disorder were lower than those of controls (F [1,39J = 5.13, P = 0.03). Both follicular and luteal phase levels were lower in subjects with late luteal phase dysphoric disorder but reached statistical significance only in the follicular phase. CONCLUSION: Lower platelet tritium-labeled imipramine binding in women with late luteal phase dysphoric disorder supports the hypothesis that alteration of central serotonergic systems may contribute to premenstrual dysphoric symptoms. (AM J OBSTET GVNECOL 1992;167:168-72.)
Key words: Late luteal phase dysphoric disorder, premenstrual syndrome, depression, tritium-labeled imiprimine binding The premenstrual occurrence of a variety of physical and emotional symptoms has been termed premenstrual syndrome (PMS) or, when affective symptoms are emphasized, late luteal phase dysphoric disorder.! Although epidemiologic and clinical evidence demonstrate significant psychiatric comorbidity in PMS and late luteal phase dysphoric disorder,2 studies of putative biochemical markers of depression such as the dexamethasone suppression test 3 and the thyrotropin-releasing hormone stimulation test4 have not revealed consistent deviations from normal values in women with PMS. Platelet tritium-labeled imiprimine binding sites, which are believed to represent receptor sites that label the presynaptic serotonin transporter on the presynaptic nerve terminal, are lower in number in drug-free depressed patients during acute depressive episodes than they are in controls.' There is some evidence that
From the Department of Obstetrics and Gynecology," the Division of ,Hedical Psychology,! the Department of Psychiatry,' and the DepartlIlent of Pharmacology,' Duke University Medical Cenler. Supported in part by the John D. and Catherine T. MacArthur Foundation Health Research Network 4: Determinants and Consequences £if Health-Promoting and Health-Damaging Behavior and by the National Institute for Mental Health grant MH-40159. Received for publication March 26, 1991; revised January 10, 1992; accepted January 23,1992. Reprint requests: John F. Steege, MD, Box 3263, Duke University Medical Center, Durham, NC 27710.
6/1/36686 168
the number of these sites returns to normal several months after clinical remission of depression, response to psychotropic medication,6 or electroconvulsive therapy.7 This platelet binding site has, therefore, been described as a state marker." Because the affective state changes in a predictable fashion in late luteal phase dysphoric disorder, this disorder provides a model for the investigation of platelet [3H]imipramine binding sites and their association with depression. This study investigated the number and affinity of platelet [3H]imipramine binding sites across the menstrual cycle in women without clinical affective disorder who fulfilled criteria for late luteal phase dysphoric disorder.
Material and methods Twenty-three women aged 24 to 42 (mean 32.2 ± 6.0 SD) were recruited by advertisement and word of mouth to participate in a study of cardiovascular and neuroendocrine reactivity across the menstrual cycle. Subjects with late luteal phase dysphoric disorder and controls were of equivalent parity (mean ± SD 0.63 ± 0.98 for all subjects, range 0 to 2), and all were within ± 15 %of ideal body weight. Women with late luteal phase dysphoric disorder reported symptoms for 3 to 11 years (mean ± SD 5.7 ± 2.8) and participated in the study primarily to obtain free evaluation for their premenstrual symptoms. Subjects gave informed consent under a protocol approved by the
Volume 167 Number I
institutional review board and were paid for their participation. All subjects completed the Premenstrual Assessment Form 9 as a screening instrument and then completed prospective symptom charting over two consecutive menstrual cycles. Twelve women were designated asymptomatic controls, and nine met the DSM-IlIR criteria l for late luteal phase dysphoric disorder. A sym ptom was scored as present and cyclic if it was rated at s2 on a six-point ordinal scale throughout the postmenstrual follicular phase of the menstrual cycle and was present at moderate or severe levels (a score of 2:4) during at least 4 days of the premenstrual week. During baseline cycles, ovulation was detected by basal body temperature recordings. During cycles in which laboratory sessions took place, ovulation was documented by the Q-test (Becton Dickinson, Franklin Lakes, N.J.). A total of 142 subjects were screened. All subjects met the following additional inclusion criteria: a normal Minnesota Multiphasic Personality Inventory profile, which was completed during the follicular phase; absence of Axis I diagnosis on the Structured Clinical Interview for the DSM-III-RlO; a normal follicular phase Beck Depression Inventory score ll ; the absence of serious medical illness such as thyroid dysfunction; ingestion of no psychotropic, hormonal, or other medications; and a menstrual cycle length of 24 to 32 days. All subjects had normal complete physical and gynecologic examinations and normal laboratory studies, including complete blood counts, urinalysis, and thyroid-stimulating hormone and free thyroxine evaluations. Subjects completed two laboratory sessions in the follicular and luteal phases of the menstrual cycle in counterbalanced order. Relative to the day of ovulation, cycle days of the laboratory studies were similar for control subjects and those with late luteal phase dysphoric disorder (follicular: control - 5.1 ± 1.8 vs late luteal phase dysphoric disorder -6.2 ± 2.2, NS; Luteal: control 9.6 :±: 2.4 vs late luteal phase dysphoric disorder 8.7 ± 1.5, NS). All sessions began between 1:30 and 3 PM. Each session lasted approximately 75 minutes, during which subjects were exposed to various stressors. At the beginning and end of each session, 35 ml of blood was drawn into evacuated glass tubes containing 2 ml of acid citrate-dextrose solution, which served as an anticoagulant. Platelet-rich plasma was obtained by centrifugation at 100g for 2 minutes; centrifugation was repeated three times. The platelet-rich plasma was then centrifuged (16,000g for 10 minutes at 4° C), and the resulting platelet-rich pellet was suspended in 50 mmol of Tris buffer (pH 7.5 at 4° C) containing 150 mmol of sodium chloride and 20 mmol of ethylenediaminetetraacetic acid by means of a Polytron homogenization medium probe (Polytron, Brinkmann In-
[3Hjlmipramine binding sites in PMS
169
struments, Westbury, N.Y.) at setting of 4 for 10 to 15 seconds. After centrifugation at 39,000g for 10 minutes, membranes were prepared by hypotonic lysis in 5 mmol Tris (pH 7.5), containing 5 mmol ethylenediam'inetetraacetic acid and were then homogenized and centrifuged at 39,000g for 10 minutes. The supernatant was discarded, and the pellet was washed by resuspension in 70 mmol Tris (pH 7.5) and was centrifuged again at 39,000g for 10 minutes. The washed pellet was finally resuspended in the assay buffer, which consisted of 50 mmol Tris (pH 7.5 at 4° C) containing 120 mmol of sodium chloride and 5 mmol of potassium chloride. Aliquots were taken for protein deteJ;mination. Platelets were then frozen at 70° C until they were used for the binding assay. Platelet [3H]imipramine binding was conducted with a modification of the procedure of Nemeroff et a!.l2 In each sample, total binding was assayed at seven different concentrations (0.25 nmol to 8 nmol) of pH]imipramine (specific activity 45.4 Ciimmol); the dissociation constant (Kd) is approximately 1 nmo!. Each assay tube (all concentrations were measured in triplicate) contained approximately 100 f.Lg of platelet protein, and the final assay volume was 250 mm'. Tubes were incubated with the radiolabel for 60 minutes on ice. The reaction was terminated by addition of 5 ml of ice-cold assay buffer, and bound radioactivity was collected by filtration under vacuum over glass fiber filters presoaked for 3 hours in buffer. The filters were then washed three times with 5 ml of ice-cold assay buffer, and bound radioactivity was collected by filtration under vacuum over glass fiber filters presoaked for 3 hours in buffer. The filters were then washed three times with 5 ml of ice-cold assay buffer. Nonspecific binding was defined as binding persisting in the presence of 100 f.Lmol of desipramine. Specific binding represented 80% of total binding, which was less than 5% of the total counts added. Scatchard analysis was used to obtain the maximum binding and Kd for ['H]imipramine. Linear regression of Scatchard plots was used to determine the correlation coefficient r. Assays with r < 0.90 were excluded. The interassay coefficients of variation were 10% for maximum binding and 5% for Kd. Comparison of the first and second assays of imipramine binding sites in each laboratory session was performed by paired Student t test. The Student-Newman-Keuls test was used to compare pooled [3H]imipramine level results from controls with those from subjects with late luteal phase dysphoric disorder. Repeated measures analysis of variance was used to measure group (control vs late luteal phase dysphoric disorder) x time (follicular vs luteal) interactions. Analyses of variance with Duncan's one-tailed t tests were used to compare results from both groups within one
170 Steege et al.
July 1992 Am J Obstet Gynecol
c:
1600
'(j)
.- 1400 0
.... a..
-
1200
E "0 E
800
0
' 0)
00
•
1000
600 >< ro E 400
III
o
o
200
o
o
Follicular
Luteal PMS (N=9)
Follicular Luteal Controls (N= 12)
Fig. 1. ['H]Imipramine binding sites in controls and women with premenstrual syndrome by menstrual cycle phase. Bars, means; shaded areas, ± I SEM. Only paired data points were used for statistical analysis.
cycle phase. Pearson's r was used to measure the correlation between luteal platelet binding site density and Beck Depression Inventory scores from baseline cycles. The p < 0.05 level of significance was used. Results
Control subjects had, as expected, normal Beck Depression Inventory scores in the follicular (1.0 ± 1.2, mean ± SD) and luteal (2.4 ± 3.2, mean:±: SD) phases of a baseline charting cycle. PMS subjects were normal in the follicular phase but were significantly depressed in the luteal phase; this was demonstrated by increased Beck Depression Inventory scores (2.1 ± 2.90 vs 17.1 ± 10.0, P< 0.001). Four of the nine PMS subjects had luteal phase Beck scores in the clinically depressed range (> 16). The number of [3H]imipramine binding sites in samples obtained before and after the laboratory sessions did not differ significantly. This attests to the stability of platelet binding site characteristics over short time intervals, i.e., test-retest reliability. The mean value for each day was therefore used for subsequent analyses. For the analyses of variance, one subject was dropped from each group as a result of our inability to obtain samples in both menstrual cycle phases. When follicular and luteal phase values within each subject group were combined, platelet [3H]imipramine binding site density for PMS women was significantly lower than that of controls (800.2 ± 237.8 vs 981.4 ± 258.5 fmol/gm protein [F(l,39) = 5.13, P = 0.03]). Repeated measures analysis of variance revealed a time effect at the trend level (F[l,17] = 2.57, P = 0.12), but no time X group effect. When platelet binding site number by cycle phase across subject groups was considered, (Fig. 1) women
with PMS had lower maximum binding values than did controls in the follicular phase (744.8 ± 247.6 vs 963.0 ± 240.9 fmoll gm protein) by analysis of variance (F[I,20] = 4.12, P = 0.057) and Duncan's one-tailed t test (p < 0.05), but the two groups were not different in the luteal phase (862.4 ± 225.4 vs 1006.3 ± 284.2 fmol/gm protein; F[l,18], p = 0.27, and p > 0.05, respectively). The increase in the number of [3H]imipramine binding sites in platelets in PMS women from the follicular to the luteal phase was statistically significant (750.5 ± 264.1 vs 862.4 ± 225.4 fmol/gm protein, p < 0.05). Luteal phase Beck Depression Inventory scores were not correlated with either follicular or luteal platelet [3H]imipramine binding site density in PMS subjects. Comment
This study documents lower numbers of platelet [3H]imipramine binding sites in women with well-documented PMS who, in the follicular phase of the menstrual cycle, were clearly free of depression by stringent clinical and psychometric criteria. In results consistent with those of previous reports,'" 14 asymptomatic women showed no influence of menstrual cycle phase on platelet [3H]imipramine binding. Subjects in the present study with late luteal phase dysphoric disorder had lower values in the follicular as opposed to the luteal phase. Rojansky et al. 15 also noted differences in [3H]imipramine binding in controls as opposed to women with "premenstrual changes," but they noted lower maximum binding values in early versus late luteal phase samples from symptomatic women. However, they observed a significant cycle phase effect on maximum binding in controls, a change that was not observed in this study or in previous reports. 13.14 Their
[3Hjlmipramine binding sites in PMS
Volume 167 Number 1
subject population included some women with a lifetime diagnosis of depression, but this study excluded those women. The large degree of variability in the assay results and the limited number of times of sampling may further account for these apparent discrepanCIes. The density of [3H]imipramine binding sites has been reported to be reduced in depressed patients in the majority of studies 8 and in scattered reports of a few possibly related conditions such as anorexia,16 agoraphobia,17 and obsessive-compulsive disorder. 16 The latter two disorders are among those overrepresented in an epidemiologic study of psychiatric disease and premenstrual symptoms (Duke Epidemiologic Catchment Area Study. Unpublished data). The number of platelet [3H]imipramine binding sites was about 18% lower in women with PMS than in controls; this was a smaller difference than the average 35% reduction seen in studies of depressed patients, which suggests that PMS may be a variant of depression modulated by the menstrual cycle. The observation of a strong trend toward reduced binding sites to the serotonin transporter in women with PMS during their asymptomatic times is of interest. Platelets have a mean life-span of 21 days. Because [3H]imipramine binding sites are cell membrane characteristics of the anucleate platelet, they may be seen as evidence of synthetic events that occurred weeks earlier in the megakaryocyte. Low-follicular-phase density of the binding sites and the emotional state of the preceding luteal phase may both be the results of luteal phase biosynthetic events. Although previous studies have shown that platelet [3H]imipramine binding sites return to normal after the clinical resolution of depression, most researchers have measured this change 1 to 2 months after clinical improvement has occurred. 7 . 18. 19 PMS may indeed provide a viable model for the study of affective state change that may resemble an acute depressive episode. Sex hormones are well known to affect serotonergic systems. Pharmacologic doses of estrogen in rats· o and physiologic replacement doses (10 mg of estradiol valerate monthly) in surgically menopausal women 21 increase the number of these binding sites approximately 20%. The time course of these changes is uncertain. In the latter study,21 women given estrogen showed improved mood com pared with that of their preoperative state, but controls failed to show any decline in mood while deprived of estrogen replacement. The precise relationship of estrogen levels to mood thus remains enigmatic with respect to both premenstrual and postmenopausal states. Previous studies of serotonergic mechanisms in PMS have yielded contradictory findings. Whole blood serotonin levels··' 23 and 5-hydroxytryptamine 24 uptake are reported to be lower in premenstrual women with
171
premenstrual syndrome, and the maximum volume of platelet serotonin uptake is reportedly decreased premenstrually as well. 22 Malmgren et al.,25 in a study limited by the absence of daily symptom ratings, compared women who had PMS with control subjects and found no effect of subject group or cycle phase on platelet serotonin uptake. The lack of correlation of a state measure of depression (Beck Depression Inventory score) with the density of binding sites is consistent with reports of depressed patients in which psychometric measures of mood failed to correlate with this measure of presynaptic serotonergic function. 12 The high level of variance in binding site data in general makes it difficult to show such associations without studying large numbers of subjects and diminishes the potential clinical diagnostic value of this laboratory measure. This study provides impetus for the further investigation of the dynamics of this presynaptic serotonergic binding site as a marker of affective state across the menstrual cycle and the relationship between premenstrual emotional changes and more continuous depressive disorders. REFERENCES
I. Diagnostic and statistical manual of mental disorders. 3rd ed, revised. Washington: American Psychiatric Association,1987. 2. Halbreich U, Endicott J. Relationship of dysphoric premenstrual changes to depressive disorders. Acta Psychiatr Scand 1985;71:331. 3. Haskett R, Steiner M, Carroll BS. Dexamethasone suppression tests and the menstrual cycle. J Affective Disord 1984;6:191. 4. Roy-Byrne PP, Rubinow DR, Hoban MC, et al. TSH and prolactin responses to TRH in patients with premenstrual syndrome. Am J Psychiatry 1987; 144:480. 5. Lewis DA, McChesney C. Tritiated imipramine binding distinguishes among subtypes of depression. Arch Gen Psychiatry 1985;42:485-8. 6. Braddock LE, Cowen PJ, ElliottJM, Fraser S, Stump K. Changes in the binding to platelets of ['HJ-imipramine and fHJ-yohimbine in normal subjects taking amitriptyline. Neuropharmacology 1984;23:285-6. 7. Langer SZ, Sechter D, Loo H, Raisman R, Zarifian E. Electroconvulsive shock therapy and maximum binding of platelet ['Hj-imipramine in depression. Arch Gen Psychiatry 1986;43:949-52. 8. Langer SZ, Galzin AM, Poirier MF, Loo H, Sechter D, Zarifian E. Association of ['HJ-paroxetine binding with the 5HT transporter in brain and platelets: relevance to studies in depression. J Recept Res 1987;7:499-521. 9. Halbreich U, Endicott J, Schacht S, Nee J. The diversity of premenstrual changes as reflected and the Premenstrual Assessment Form. Acta Psychiatr Scand 1982; 65:46-65. 10. Spitzer RL, WilliamsJBW, Gibbon M. Instruction manual for the structured clinical interview for the DSM-IlI-R. April, 1987 revision. New York: Biometrics Research Departments, New York State Psychiatric Institute, 1987. 11. Beck AT, Rush AJ, Shaw BF, et al. Cognitive therapy of depression. New York: Guilford Press, 1979:398-9. 12. NemeroffCB, Knight DL, Krishnan RR, Slotkin TA, Bissette G, Melville ML, Blazer DG. Marked reduction in the number of platelet tritiated imipramine binding sites in
Steege et al.
13.
14. 15.
16.
17. 18.
19.
geriatric depression. Arch Gen Psychiatry 1988;45:91923. Poirier MF, Benkelfat C, Galzin AM, Langer SZ. Platelet 3H-imipramine binding and steroid hormones serum concentrations during the menstrual cycle. Psychopharmacology (Berl) 1986;88:86-9. Stowell LI, McIntosh Cj, Cooke R, Ellis PM. Adrenoreceptor and imipramine receptor binding during the menstrual cycle. Acta Psychiatr Scand 1988;78:366-8. Rojansky N, Halbreich U, Zanden K, Barkai A, Goldstein S. Imipramine receptor binding and serotonin uptake in platelets of women with premenstrual changes. Gynecol Obstet Invest 1991 ;31: 146-52. Weizman R, Carmi M, Tyano S, Rehavi M. Reduced 3H_ imipramine binding but unaltered 3H-serotonin uptake in platelets of adolescent enuretics. Psychiatry Res 1986;19:37-42. Lewis DA, Noyes R, Coryell W, Clancy J. Tritiated imipramine binding to platelets is decreased in patients with agoraphobia. Psychiatry Res 1985;16:1-9. Suranyi-Cadotte BE, Wood PL, Nouir NPV, Schwartz G. Normalization of platelets (3H) imipramine binding in depressed patients during remission. Eur J Pharmacol 1982;85:357 -8. Suranyi-Cadotte BE, Quirion R, McQuade P, Nair NPV,
July 1992 Am J ObstetGynecol
Schwartz G, Mosticyan S, Wood PL. Platelet 'H-imipramine binding: a state-dependent marker in depression. Prog Neuropsychopharmacol Bioi Psychiatry 1984;8:73741.
20. Rehavi M, Sepcuti H, Weizman A. Up-regulation of imipramine binding and serotonin uptake by estradiol in female rat brain. Brain Res 1987;410:135-9. 21. Sherwin BB, Suranyi-Cadotte BE. Up-regulatory effect of estrogen on platelet 3H-imipramine binding sites in surgically menopausal women. Bioi Psychiatry 1990;28;33948. 22. Taylor DL, Mathew Rj, Mo BT, Weineman ML. Serotonin levels and platelet uptake during premenstrual tension. Neuropsychobiology 1984; 12: 16-18. 23. Rapkin Aj, Edelmuth E, Chang L, Reading AE, McGuire MT, Su TP. Whole blood serotonin in premenstrual syndrome. Obstet Gynecol 1987;70:533-7. 24. Ashby CR, Carr LA, Cook CL, Steptoe MM, Franks DD. Alteration of platelet serotonergic mechanisms and monoamine oxidase activity in premenstrual syndrome. BioI Psychiatry 1988;24:225-33. 25. Malmgren R, Collins A, Nilsson CG. Platelet serotonin uptake and effects of vitamin B6 treatment in premenstrual tension. Neuropsychobiology 1987;18:83-8.
The development of ultradian rhythms in the human fetus Mary Pillai, MB, David K. James, MD, and Michael Parker, MB Bristol, Canada OBJECTIVE: Our objective was to determine the normal sequence of neurobehavioral development in the human fetus between 14 weeks' gestation and delivery. STUDY DESIGN: The study was performed by longitudinal ultrasonographic observation of 45 low-risk singleton fetuses. RESULTS: During the first half of the midtrimester there was a high rate of spontaneous movement that appeared randomly distributed. By the end of that trimester an increase in the duration of intervals of quiescence resulted in activity appearing cyclically distributed, with the duration of quiet cycles progressively increasing to term. Fetal mouthing and breathing were linked with cyclic behavior from the time of their emergence. Fetal heart rate pattern A could be recognized from around 32 weeks, due to a reduction in baseline variability in quiet cycles after 30 weeks, whereas pattern B could be recognized several weeks earlier. From the time cyclic behavior and heart rate patterns could be recognized, intervals of coincidence of the fetal behavioral state variables increased progressively with advancing gestation. (AM J OBSTET GVNECOL 1992;167:172-7.)
Key words: Fetal behavior, ultradian rhythms From the University Department of Obstetrics and Gynaecology, Bristol Maternity Hospital. Supported by Action Research for the Crippled Child. Received for publication August 15, 1991; revised Decembel' 20, 1991; accepted December 30, 1991. Reprint requests: Mary Pillai, Women's Hospital, 735 Notre Dame Ave., Winnipeg, Manitoba, Canada R3E OL8. 611 /35916 172
The first half of pregnancy is characterized by rapid emergence of a wide variety of fetal movements of relatively high frequency but little neurologic organization." 2 During the second half of pregnancy these movements become organized. Diurnal rhythms are not demonstrable at 13 weeks but have been reported
